0% found this document useful (0 votes)

81 views

Base Proc 6982

base proc

Uploaded by

rpkn541

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

81 views

Base Proc 6982

base proc

Uploaded by

rpkn541

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 1861

Base SAS 9.

1 Procedures Guide

The correct bibliographic citation for this manual is as follows: SAS Institute Inc. 2004.
Base SAS 9.1 Procedures Guide. Cary, NC: SAS Institute Inc.
Base SAS 9.1 Procedures Guide
Copyright 2004 by SAS Institute Inc., Cary, NC, USA
ISBN 1-59047-204-7
All rights reserved. Produced in the United States of America. No part of this publication
may be reproduced, stored in a retrieval system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, or otherwise, without the prior written
permission of the publisher, SAS Institute Inc.
U.S. Government Restricted Rights Notice. Use, duplication, or disclosure of this
software and related documentation by the U.S. government is subject to the Agreement
with SAS Institute and the restrictions set forth in FAR 52.22719 Commercial Computer
Software-Restricted Rights (June 1987).
SAS Institute Inc., SAS Campus Drive, Cary, North Carolina 27513.
1st printing, January 2004
SAS Publishing provides a complete selection of books and electronic products to help
customers use SAS software to its fullest potential. For more information about our
e-books, e-learning products, CDs, and hard-copy books, visit the SAS Publishing Web site
at support.sas.com/pubs or call 1-800-727-3228.
SAS and all other SAS Institute Inc. product or service names are registered trademarks
or trademarks of SAS Institute Inc. in the USA and other countries. indicates USA
registration.
Other brand and product names are registered trademarks or trademarks of their
respective companies.

Contents
Whats New
xi
Overview
xi
Details
xi

PART

Concepts
Chapter 1

4 Choosing the Right Procedure

Chapter 2

4 Fundamental Concepts for Using Base SAS Procedures

Language Concepts
16
Procedure Concepts
19
Output Delivery System

Chapter 3

4 Statements with the Same Function in Multiple Procedures

Overview
57
Statements
58

PART

Procedures
Chapter 4

4 The APPEND Procedure

Overview: APPEND Procedure 75

Syntax: APPEND Procedure 75

Chapter 5

4 The CALENDAR Procedure

Overview: CALENDAR Procedure

79
Syntax: CALENDAR Procedure
84
Concepts: CALENDAR Procedure
101
Results: CALENDAR Procedure
112
Examples: CALENDAR Procedure 114

Chapter 6

4 The CATALOG Procedure

153

Overview: CATALOG Procedure

153
Syntax: CATALOG Procedure 154
Concepts: CATALOG Procedure 165
Results: CATALOG Procedure
169
Examples: CATALOG Procedure
170

Chapter 7

4 The CHART Procedure

179

Overview: CHART Procedure 179

Syntax: CHART Procedure 184
Concepts: CHART Procedure
197
Results: CHART Procedure
197
Examples: CHART Procedure
198
References
213

Chapter 8

4 The CIMPORT Procedure

215

4 The COMPARE Procedure

225

Overview: CIMPORT Procedure

215
Syntax: CIMPORT Procedure
216
Results: CIMPORT Procedure 221
Examples: CIMPORT Procedure
222

Chapter 9

Overview: COMPARE Procedure

226
Syntax: COMPARE Procedure
229
Concepts: COMPARE Procedure 240
Results: COMPARE Procedure
244
Examples: COMPARE Procedure
256

Chapter 10

4 The CONTENTS Procedure

275

Overview: CONTENTS Procedure 275

Syntax: CONTENTS Procedure
275

Chapter 11

4 The COPY Procedure

277

4 The CORR Procedure

283

Overview: COPY Procedure 277

Syntax: COPY Procedure 277
Concepts: COPY Procedure
278
Example: COPY Procedure
279

Chapter 12

Information about the CORR Procedure

Chapter 13

4 The CPORT Procedure

283

285

Overview: CPORT Procedure

285
Syntax: CPORT Procedure
286
Concepts: CPORT Procedure
294
Results: CPORT Procedure 294
Examples: CPORT Procedure
295

Chapter 14

4 The CV2VIEW Procedure

301

Information about the CV2VIEW Procedure

Chapter 15

4 The DATASETS Procedure

Overview: DATASETS Procedure 304

Syntax: DATASETS Procedure
307
Concepts: DATASETS Procedure
357
Results: DATASETS Procedure
363

301

303

Examples: DATASETS Procedure

Chapter 16

376

4 The DBCSTAB Procedure

393

Information about the DBCSTAB Procedure

Chapter 17

4 The DISPLAY Procedure

393

395

Overview: DISPLAY Procedure

395
Syntax: DISPLAY Procedure
395
Example: DISPLAY Procedure
396

Chapter 18

4 The DOCUMENT Procedure

399

Information about the DOCUMENT Procedure

Chapter 19

4 The EXPLODE Procedure

401

Information about the EXPLODE Procedure

Chapter 20

399

4 The EXPORT Procedure

401

403

Overview: EXPORT Procedure

403
Syntax: EXPORT Procedure
404
Examples: EXPORT Procedure
411

Chapter 21

4 The FONTREG Procedure

419

Overview: FONTREG Procedure

419
Syntax: FONTREG Procedure
419
Concepts: FONTREG Procedure
423
Examples: FONTREG Procedure
425

Chapter 22

4 The FORMAT Procedure

429

Overview: FORMAT Procedure 430

Syntax: FORMAT Procedure
431
Informat and Format Options
451
Specifying Values or Ranges
453
Concepts: FORMAT Procedure
455
Results: FORMAT Procedure
458
Examples: FORMAT Procedure 463

Chapter 23

4 The FORMS Procedure

485

Information about the FORMS Procedure

Chapter 24

4 The FREQ Procedure

487

Information about the FREQ Procedure

Chapter 25

4 The FSLIST Procedure

485

487

489

Overview: FSLIST Procedure

489
Syntax: FSLIST Procedure 489
Using the FSLIST Window
494

Chapter 26

4 The IMPORT Procedure

Overview: IMPORT Procedure

501

Syntax: IMPORT Procedure

502
Examples: IMPORT Procedure
514

Chapter 27

4 The MEANS Procedure

Overview: MEANS Procedure

523

524

Syntax: MEANS Procedure

526

Concepts: MEANS Procedure

550

Statistical Computations: MEANS Procedure

553

Results: MEANS Procedure

556
Examples: MEANS Procedure
558
References

Chapter 28

587

4 The MIGRATE Procedure

589

Information about the MIGRATE Procedure

Chapter 29

4 The OPTIONS Procedure

Overview: OPTIONS Procedure

591

Syntax: OPTIONS Procedure

595

Results: OPTIONS Procedure

596

Examples: OPTIONS Procedure

Chapter 30

597

4 The OPTLOAD Procedure

601

4 The OPTSAVE Procedure

603

Overview: OPTLOAD Procedure

601
Syntax: OPTLOAD Procedure
601

Chapter 31

Overview: OPTSAVE Procedure

603

Syntax: OPTSAVE Procedure

Chapter 32

603

4 The PLOT Procedure

Overview: PLOT Procedure

Syntax: PLOT Procedure

605

606
608

Concepts: PLOT Procedure

624

Results: PLOT Procedure

629
Examples: PLOT Procedure 631

Chapter 33

4 The PMENU Procedure

Overview: PMENU Procedure

665

Syntax: PMENU Procedure

666

Concepts: PMENU Procedure

679

Examples: PMENU Procedure

Chapter 34

682

4 The PRINT Procedure

Overview: PRINT Procedure

Syntax: PRINT Procedure

665

703
705

Results: Print Procedure 720

Examples: PRINT Procedure
723

703

589

vii

Chapter 35

4 The PRINTTO Procedure

771

Overview: PRINTTO Procedure

771
Syntax: PRINTTO Procedure
772
Concepts: PRINTTO Procedure 775
Examples: PRINTTO Procedure
776

Chapter 36

4 The PROTO Procedure

787

Information about the PROTO Procedure

Chapter 37

787

4 The PRTDEF Procedure

789

4 The PRTEXP Procedure

803

Overview: PRTDEF Procedure

789
Syntax: PRTDEF Procedure
789
Input Data Set: PRTDEF Procedure
791
Examples: PRTDEF Procedure
796

Chapter 38

Overview: PRTEXP Procedure

803
Syntax: PRTEXP Procedure
803
Concepts: PRTEXP Procedure 805
Examples: PRTEXP Procedure
805

Chapter 39

4 The PWENCODE Procedure

807

Overview: PWENCODE Procedure

807
Syntax: PWENCODE Procedure
807
Concepts: PWENCODE Procedure
808
Examples: PWENCODE Procedure
809

Chapter 40

4 The RANK Procedure

813

Overview: RANK Procedure

813
Syntax: RANK Procedure
815
Concepts: RANK Procedure 820

Results: RANK Procedure

821
Examples: RANK Procedure
822
References
829

Chapter 41

4 The REGISTRY Procedure

831

Overview: REGISTRY Procedure

831
Syntax: REGISTRY Procedure
831
Creating Registry Files with the REGISTRY Procedure
Examples: REGISTRY Procedure 839

Chapter 42

4 The REPORT Procedure

Overview: REPORT Procedure 847

Concepts: REPORT Procedure
852
Syntax: REPORT Procedure
869
REPORT Procedure Windows
912
How PROC REPORT Builds a Report

845

936

836

viii

Examples: REPORT Procedure

Chapter 43

948

4 The SORT Procedure

1003

Overview: SORT Procedure

1003
Syntax: SORT Procedure 1005
Concepts: SORT Procedure 1013
Integrity Constraints: SORT Procedure
Results: SORT Procedure
1016
Examples: SORT Procedure
1016

Chapter 44

4 The SQL Procedure

1015

1027

Overview: SQL Procedure

1029
Syntax: SQL Procedure 1031
SQL Procedure Component Dictionary 1071
Concepts: SQL Procedure 1114
PROC SQL and the ANSI Standard 1122
Examples: SQL Procedure
1125

Chapter 45

4 The STANDARD Procedure

1163

Overview: STANDARD Procedure

1163
Syntax: STANDARD Procedure
1165
Results: STANDARD Procedure
1170
Statistical Computations: STANDARD Procedure
Examples: STANDARD Procedure
1171

Chapter 46

4 The SUMMARY Procedure

1177

4 The TABULATE Procedure

1179

4 The TEMPLATE Procedure

1171

1285

Overview: SUMMARY Procedure

1177
Syntax: SUMMARY Procedure
1177

Chapter 47

Overview: TABULATE Procedure 1180

Terminology: TABULATE Procedure
1183
Syntax: TABULATE Procedure 1186
Concepts: TABULATE Procedure
1213
Results: TABULATE Procedure
1222
Examples: TABULATE Procedure 1232
References
1283

Chapter 48

Information about the TEMPLATE Procedure

Chapter 49

4 The TIMEPLOT Procedure

1285

1287

Overview: TIMEPLOT Procedure

1287
Syntax: TIMEPLOT Procedure 1289
Results: TIMEPLOT Procedure
1297
Examples: TIMEPLOT Procedure
1299

Chapter 50

4 The TRANSPOSE Procedure

1311

Overview: TRANSPOSE Procedure

1311
Syntax: TRANSPOSE Procedure
1314
Results: TRANSPOSE Procedure
1320
Examples: TRANSPOSE Procedure
1321

Chapter 51

4 The TRANTAB Procedure

1333

Information about the TRANTAB Procedure

Chapter 52

4 The UNIVARIATE Procedure

1333

1335

Information about the UNIVARIATE Procedure

PART

Appendices
Appendix 1

1335

1337

4 SAS Elementary Statistics Procedures

1339

Overview
1339
Keywords and Formulas
1340
Statistical Background 1348
References
1373

Appendix 2

4 Operating Environment-Specic Procedures

Descriptions of Operating Environment-Specic Procedures

Appendix 3

4 Raw Data and DATA Steps

1377

Overview
1377
CENSUS 1377
CHARITY
1378
CUSTOMER_RESPONSE 1380
DJIA
1383
EDUCATION 1384
EMPDATA 1385
ENERGY
1387
GROC
1388
MATCH_11 1388
PROCLIB.DELAY
1390
PROCLIB.EMP95 1391
PROCLIB.EMP96 1392
PROCLIB.INTERNAT 1393
PROCLIB.LAKES
1393
PROCLIB.MARCH
1394
PROCLIB.PAYLIST2
1395
PROCLIB.PAYROLL 1395
PROCLIB.PAYROLL2 1398
PROCLIB.SCHEDULE
1399
PROCLIB.STAFF 1402
PROCLIB.SUPERV
1405
RADIO
1405

Appendix 4

4 Recommended Reading

xii Whats New

System (ODS). PROC CONTENTS output now displays the data representation of a le
by reporting the native platform for each le, rather than just telling you whether the
data representation is native or foreign. Also, PROC CONTENTS output now provides
the encoding value, whether a character variable is transcoded if required, and whether
the data set is part of a generation group. A new example was added that shows how to
get PROC CONTENTS output into an ODS output data set for processing.
The ORDER= option was added to the CONTENTS statement to enable you to print
a list of variables in alphabetical order even if they include mixed-case names.

The COPY Procedure

The following options are new or enhanced in the COPY procedure and the COPY
statement in PROC DATASETS:
3 The FORCE option enables you to use the MOVE option for a SAS data set that
has an audit trail.
3 The CLONE option now copies the data representation and encoding data set
attributes.

The CORR Procedure

The CORR procedure has the following new features:
3 The FISHER option in the PROC CORR statement requests condence limits and
pvalues for Pearson and Spearman correlation coefcients based on Fishers z
transformation. With the FISHER option, you can specify an alpha value and a
null hypothesis value. You can also specify the type of condence limit (upper,
lower, or two-sided) and whether the bias adjustment is to be used for the
condence limits.
3 The PLOTS=MATRIX option in the PROC CORR statement uses ODS graphics to
produce either a rectangular matrix plot (if you also specify a WITH statement) or
a symmetric matrix plot (if you do not specify a WITH statement) for variables.
3 The PLOTS=SCATTER option in the PROC CORR statement uses ODS graphics
to produce scatter plots for variables. By default, the scatter plot also includes a
95% prediction ellipse. You can use the ELLIPSE= option with the
PLOTS=SCATTER option to include prediction ellipses for new observations,
condence ellipses for the mean, or no ellipses.

The DATASETS Procedure

Directory listings from the DATASETS procedure provide a new look for its output,
which improves the format for the Output Delivery System (ODS).
The following statements are enhanced in the DATASETS procedure:
3 The AUDIT statement provides the AUDIT_ALL= option in order to specify
whether logging can be suspended and whether audit settings can be changed. In
addition, the LOG option in the AUDIT statement now enables you to control the
logging of administrative events to the audit le with ADMIN_IMAGE=.
3 The ICCREATE statement now enables you to create overlapping constraints.
This means that variables in a SAS data set are part of both a primary key
denition and a foreign key denition.
3 The MODIFY statement provides the CORRECTENCODING= option to change
the encoding indicator, which is stamped in the les descriptor information, in
order to match the actual encoding of the les data.

Whats New xiii

The DOCUMENT Procedure

The new DOCUMENT procedure enables you to customize or modify your output
hierarchy and replay your output to different destinations without rerunning the PROC
or DATA step. For complete information, see SAS Output Delivery System: Users Guide.

The EXPORT Procedure

The EXPORT procedure now enables you to

3 export to Microsoft Excel 2002 spreadsheets and Microsoft Access 2002 tables. The
new data sources are available for the Windows operating environment on 32-bit
platforms if your site has a license for SAS/ACCESS Interface to PC File Formats.

3 specify SAS data set options in the DATA= argument when you are exporting to
all data sources except for delimited, comma-separated, and tab-delimited external
les. For example, if the data set that you are exporting has an assigned
password, use the ALTER=, PW=, READ=, or WRITE= data set option. To export
only data that meets a specied condition, use the WHERE= data set option.

3 identify a specic spreadsheet in a workbook by specifying the SHEET= option.

Exporting to multiple sheets is available for Microsoft Excel 97, 2000, and 2002
spreadsheets for the Windows operating environment on 32-bit platforms if your
site has a license for the SAS/ACCESS Interface to PC File Formats.

The FCMP Procedure (Experimental)

The new FCMP procedure enables you to create, test, and store SAS functions and
subroutines for use by other SAS procedures.
For more information about the FCMP procedure, go to http://support.sas.com/
documentation/onlinedoc. Select Base SAS from the Product-Specic Documentation
list.

The FONTREG Procedure

The new FONTREG procedure enables you to add system fonts to the SAS registry.

The FORMAT Procedure

3 The maximum length for character format names is now 31. The maximum length
for numeric format names is now 32.

3 The maximum length for character informat names is now 30. The maximum
length for numeric informat names is now 31.

The FREQ Procedure

In the PROC FREQ statement, the new NLEVELS option displays a table that shows
the number of levels for each variable that is named in the TABLES statement(s).
The WEIGHT statement now has the ZEROS option to include observations with zero
weight values. The frequency and crosstabulation tables will display any levels that
correspond to observations with 0 weights. PROC FREQ includes levels with 0 weights

xiv

Whats New

in the chi-square goodness-of-t test for one-way tables, in the binomial computations
for one-way tables, and in the computation of kappa statistics for two-way tables.
The following new options are available in the TABLES statement:
3 The CONTENTS= option enables you to specify the text for the HTML contents
le links to crosstabulation tables.
3 The BDT option enables you to request Tarones adjustment in the Breslow-Day
test for homogeneity of odds ratios when you use the CMH option to compute the
Breslow-Day test for stratied 222 tables.
3 The NOWARN option suppresses the log warning message that the asymptotic
chi-square test might not be valid when more than 20% of the table cells have
expected frequencies less than 5.
3 The CROSSLIST option displays crosstabulation tables in ODS column format.
This option creates a table that has a table denition that you can customize with
the TEMPLATE procedure.
Additionally, the FREQ procedure now produces exact condence limits for the
common odds ratio and related tests.

The IMPORT Procedure

The IMPORT procedure now enables you to
3 import Microsoft Excel 2002 spreadsheets and Microsoft Access 2002 tables. The
new data sources are available for the Windows operating environment on 32-bit
platforms if your site has a license for SAS/ACCESS Interface to PC File Formats.
3 specify SAS data set options in the OUT= argument when you are importing from
all data sources except for delimited, comma-separated, and tab-delimited external
les. For example, in order to assign a password for a resulting SAS data set, use
the ALTER=, PW=, READ=, or WRITE= data set option. To import only data that
meets a specied condition, use the WHERE= data set option.

The MEANS and SUMMARY Procedures

The new THREADS|NOTHREADS option enables or prevents the activation of
multi-threaded processing.
When you format class variables with user-dened formats that are created with the
MULTILABEL and NOTSORTED options, specifying the MLF, PRELOADFMT, and
ORDER=DATA options together in the CLASS statement now orders the procedure
output according to the label order in the format denition.

The MIGRATE Procedure

The new MIGRATE procedure is available specically for migrating a SAS data
library from a previous release to the most recent release. For migration, PROC
MIGRATE offers benets that PROC COPY does not. For PROC MIGRATE
documentation, see the Migration Community at http://support.sas.com/rnd/
migration.

The PROTO Procedure

The PROTO procedure, which has been available in SAS Risk Dimensions software,
is now a Base SAS procedure. The PROTO procedure enables you to register, in batch,

Whats New xv

external functions that are written in the C or C++ programming languages for use in
SAS programs and C-language structures and types. For PROC PROTO documentation,
go to http://support.sas.com/documentation/onlinedoc. Select Base SAS from
the Product-Specic Documentation list.

The PRTDEF Procedure

There are 15 new variables to control the default printer settings.

The PRTEXP Procedure

The new PRTEXP procedure enables you to write attributes, which are used by
PROC PRTDEF to dene a printer, either to a SAS data set or to the SAS log. With
this capability you can replicate and modify those attributes easily.

The PWENCODE Procedure

The new PWENCODE procedure enables you to encode a password. You can use the
encoded password in place of plain-text passwords in SAS programs that access
relational database management systems (RDBMSs) and SAS servers (such as the SAS
Metadata Server).

The REGISTRY Procedure

The REGISTRY procedure has three new options:

3 The LISTREG option lists the contents of the registry in the log.
3 The COMPAREREG1 and COMPAREREG2 options are used together to compare
two registries. The results appear in the log.

The REPORT Procedure

The REPORT procedure has the following new features:
3 The new THREADS|NOTHREADS option enables or prevents the activation of
multi-threaded processing.

3 Numeric class variables that do not have a format assigned to them are
automatically formatted with the BEST12. format.

3 PROC REPORT now writes the value _PAGE_ for the _BREAK_ variable in the
output data set for observations that are derived from a COMPUTE BEFORE
_PAGE_ or COMPUTE AFTER _PAGE_ statement.

The SORT Procedure

The SORT procedure has the following new options:

3 The DATECOPY option copies to the output data set the SAS internal date and
time when the input data set was created, and the SAS internal date and time
when it was last modied prior to the sort.

3 The DUPOUT= option species an output data set that contains duplicate
observations.

xvi

Whats New

3 The OVERWRITE option deletes the input data set before the replacement output
data set is populated with observations.

3 The THREADS|NOTHREADS option enables or prevents the activation of

multi-threaded sorting.

The SQL Procedure

The SQL procedure has the following new features:

3 The PROC SQL statement now has a THREADS | NOTHREADS option.

THREADS enables PROC SQL to take advantage of the new parallel processing
capabilities in SAS when performing sorting operations.

3 There are new DICTIONARY tables, new columns in existing DICTIONARY

tables, and SASHELP views of the new tables. For DICTIONARY.TABLES and
SASHELP.VTABLE, if a table is read-protected with a password, the only
information that is listed for that table is the library name, member name, member
type, and type of password protection; all other information is set to missing.

3 You can now reference a permanent SAS data set by its physical lename.
3 When using the INTO clause to assign values to a range of macro variables, you
can now specify leading zeroes in the macro variable names.

3 PROC SQL now supports TRANSCODE=YES|NO as a column modier.

The SYLK Procedure (Experimental)

The new SYLK procedure enables you to read an external SYLK-formatted
spreadsheet into SAS, including data, formulas, and formats. You can also use PROC
SYLK as a batch spreadsheet, using programming statements to manipulate data,
perform calculations, generate summaries, and format the output.
For more information about the SYLK procedure, go to http://support.sas.com/
documentation/onlinedoc. Select Base SAS from the Product-Specic Documentation
list.

The TABULATE Procedure

The TABULATE procedure has the following new features:

3 The new THREADS|NOTHREADS option enables or prevents the activation of

multi-threaded processing.

3 Available statistics include upper and lower condence limits, skewness, and
kurtosis. PROC TABULATE now supports the ALPHA= option, which enables you
to specify a condence level.

3 Numeric class variables that do not have a format assigned to them are
automatically formatted with the BEST12. format.

3 The new FORMAT_PRECEDENCE and STYLE_PRECEDENCE options in the

TABLE statement enable you to specify which formats and styles (dened for the
column, row, or page dimensions) are applied.
Additionally, when you format class variables with user-dened formats that are
created with the MULTILABEL and NOTSORTED options, specifying the MLF,
PRELOADFMT, and ORDER=DATA options together in the CLASS statement now
orders the procedure output according to the label order in the format denition.

Whats New xvii

The TEMPLATE Procedure

The TEMPLATE procedure now enables you to customize or create your own markup
language for your output. For complete information, see SAS Output Delivery System:
Users Guide.

The TIMEPLOT Procedure

The TIMEPLOT procedure now supports the SPLIT= option, which enables you to
specify a character at which labels will be split into multiple lines.

The UNIVARIATE Procedure

The UNIVARIATE procedure has the following new features:

3 The LOWER= and NOUPPER= suboptions in the KERNEL option in the

HISTOGRAM statement specify the lower and upper bounds for tted kernel
density curves.

3 The FRONTREF option in the HISTOGRAM statement draws reference lines in

front of the histogram bars instead of behind them.

xviii Whats New

P A R T

Concepts
Chapter

1. . . . . . . . . . Choosing the Right Procedure

Chapter

2 . . . . . . . . . . Fundamental Concepts for Using Base SAS Procedures

Chapter

3 . . . . . . . . . . Statements with the Same Function in Multiple

Procedures 57

3
15

CHAPTER

1
Choosing the Right Procedure
Functional Categories of Base SAS Procedures 3
Report Writing 3
Statistics 3
Utilities 4
Report-Writing Procedures 4
Statistical Procedures 6
Available Statistical Procedures 6
Efciency Issues 7
Quantiles 7
Computing Statistics for Groups of Observations 7
Additional Information about the Statistical Procedures 7
Utility Procedures 7
Brief Descriptions of Base SAS Procedures 10

Functional Categories of Base SAS Procedures

Report Writing
These procedures display useful information, such as data listings (detail reports),
summary reports, calendars, letters, labels, multipanel reports, and graphical reports:
CALENDAR

PLOT

SUMMARY*

CHART*

TABULATE*

FREQ*

REPORT*

TIMEPLOT

MEANS

SQL

* These procedures produce reports and compute statistics.

Statistics
These procedures compute elementary statistical measures that include descriptive
statistics based on moments, quantiles, condence intervals, frequency counts,

Utilities

Chapter 1

cross-tabulations, correlations, and distribution tests. They also rank and standardize
data:
CHART

RANK

SUMMARY

CORR

REPORT

TABULATE

FREQ

SQL

UNIVARIATE

MEANS

STANDARD

Utilities
These procedures perform basic utility operations. They create, edit, sort, and
transpose data sets, create and restore transport data sets, create user-dened formats,
and provide basic le maintenance such as to copy, append, and compare data sets:
APPEND

EXPORT

PWENCODE

FONTREG

PRTEXP

CATALOG

FORMAT

REGISTRY

CIMPORT

FSLIST

RELEASE*

COMPARE

IMPORT

SORT

CONTENTS

OPTIONS

SOURCE*

CONVERT*

OPTLOAD

SQL

COPY

OPTSAVE

TAPECOPY*

CPORT

PDS*

TAPELABEL*

CV2VIEW@

PDSCOPY*

TEMPLATE+

DATASETS

PMENU

TRANSPOSE

PRINTTO

TRANTAB#

BMDP

DBCSTAB

DOCUMENT+

*
+
@
#

PRTDEF

See the SAS documentation for your operating environment for a description of these procedures.
See SAS Output Delivery System: Users Guide for a description of these procedures.
See SAS/ACCESS for Relational Databases: Reference for a description of this procedure.
See SAS National Language Support (NLS): Users Guide for a description of these procedures.

Report-Writing Procedures
Table 1.1 on page 5 lists report-writing procedures according to the type of report.

Choosing the Right Procedure

Report-Writing Procedures

Table 1.1 Report-Writing Procedures by Task

To produce

Use this procedure

Which

Detail reports

produces data listings quickly; can supply titles,

footnotes, and column sums.

REPORT

offers more control and customization than PROC

PRINT; can produce both column and row sums; has
DATA step computation abilities.

SQL

combines Structured Query Language and SAS

features such as formats; can manipulate data and
create a SAS data set in the same step that creates the
report; can produce column and row statistics; does not
offer as much control over output as PROC PRINT and
PROC REPORT.

MEANS or
SUMMARY

computes descriptive statistics for numeric variables;

can produce a printed report and create an output data
set.

produces only one summary report: can sum the BY

variables.

REPORT

combines features of the PRINT, MEANS, and

TABULATE procedures with features of the DATA step
in a single report writing tool that can produce a
variety of reports; can also create an output data set.

SQL

computes descriptive statistics for one or more SAS

data sets or DBMS tables; can produce a printed
report or create a SAS data set.

TABULATE

produces descriptive statistics in a tabular format; can

produce stub-and-banner reports (multidimensional
tables with descriptive statistics); can also create an
output data set.

Summary reports

Miscellaneous highly formatted reports

Calendars

CALENDAR

produces schedule and summary calendars; can

schedule tasks around nonwork periods and holidays,
weekly work schedules, and daily work shifts.

Multipanel reports
(telephone book listings)

REPORT

produces multipanel reports.

CHART

produces bar charts, histograms, block charts, pie

charts, and star charts that display frequencies and
other statistics.

PLOT

produces scatter diagrams that plot one variable

against another.

TIMEPLOT

produces plots of one or more variables over time

intervals.

Low-resolution graphical reports*

* These reports quickly produce a simple graphical picture of the data. To produce high-resolution graphical
reports, use SAS/GRAPH software.

Statistical Procedures

Chapter 1

Statistical Procedures
Available Statistical Procedures
Table 1.2 on page 6 lists statistical procedures according to task. Table A1.1 on page
1341 lists the most common statistics and the procedures that compute them.
Table 1.2 Elementary Statistical Procedures by Task
To produce

Use this procedure

Which

Descriptive statistics

CORR

and partial correlations; also computes Hoeffdings D and
Cronbachs coefcient alpha.

Distribution analysis

UNIVARIATE

computes tests for location and tests for normality.

FREQ

computes a test for the binomial proportion for one-way

tables; computes a goodness-of-t test for one-way tables;
computes a chi-square test of equal distribution for two-way
tables.

UNIVARIATE

computes robust estimates of scale, trimmed means, and

Winsorized means.

RANK

computes ranks for one or more numeric variables across

the observations of a SAS data set and creates an output
data set; can produce normal scores or other rank scores.

Frequency and
cross-tabulation tables

Robust estimation
Data transformation
Computing ranks

Choosing the Right Procedure

To produce

Utility Procedures

Use this procedure

Which

STANDARD

creates an output data set that contains variables that are

standardized to a given mean and standard deviation.

CHART

produces a graphical report that can show one of the

following statistics for the chart variable: frequency counts,
percentages, cumulative frequencies, cumulative
percentages, totals, or averages.

UNIVARIATE

Standardizing data

produces descriptive plots such as stem and leaf, box plot,

and normal probability plot.

Low-resolution graphics*

* To produce high-resolution graphical reports, use SAS/GRAPH software.

Efciency Issues

Quantiles
For a large sample size n, the calculation of quantiles, including the median, requires
computing time proportional to nlog(n). Therefore, a procedure, such as UNIVARIATE,
that automatically calculates quantiles may require more time than other data
summarization procedures. Furthermore, because data is held in memory, the procedure
also requires more storage space to perform the computations. By default, the report
procedures PROC MEANS, PROC SUMMARY, and PROC TABULATE require less
memory because they do not automatically compute quantiles. These procedures also
provide an option to use a new xed-memory quantiles estimation method that is
usually less memory intense. See Quantiles on page 555 for more information.

Computing Statistics for Groups of Observations

To compute statistics for several groups of observations, you can use any of the
previous procedures with a BY statement to specify BY-group variables. However,
BY-group processing requires that you previously sort or index the data set, which for
very large data sets may require substantial computer resources. A more efcient way
to compute statistics within groups without sorting is to use a CLASS statement with
one of the following procedures: MEANS, SUMMARY, or TABULATE.

Additional Information about the Statistical Procedures

Appendix 1, SAS Elementary Statistics Procedures, on page 1339 lists standard
keywords, statistical notation, and formulas for the statistics that base SAS procedures
compute frequently. The individual statistical procedures discuss the statistical
concepts that are useful to interpret the output of a procedure.

Utility Procedures
Table 1.3 on page 8 groups utility procedures according to task.

Utility Procedures

Chapter 1

Table 1.3 Utility Procedures by Task

To perform these utility

tasks

Use this procedure

Which

Supply information

COMPARE

compares the contents of two SAS data sets.

CONTENTS

describes the contents of a SAS data library or specic

library members.

OPTIONS

lists the current values of all SAS system options.

SQL

supplies information through dictionary tables on an

individual SAS data set as well as all SAS les active in
the current SAS session. Dictionary tables can also
provide information about macros, titles, indexes,
external les, or SAS system options.

OPTIONS

lists the current values of all SAS system options.

OPTLOAD

reads SAS system option settings that are stored in the

SAS registry or a SAS data set.

OPTSAVE

saves SAS system option settings to the SAS registry or a

SAS data set.

DOCUMENT+

manipulates procedure output that is stored in ODS

documents.

FONTREG

adds system fonts to the SAS registry.

FORMAT

creates user-dened formats to display and print data.

PRINTTO

routes procedure output to a le, a SAS catalog entry, or

a printer; can also redirect the SAS log to a le.

PRTDEF

creates printer denitions.

Manage SAS system options

Affect printing and Output

Delivery System output

PRTEXP
TEMPLATE
Create, browse, and edit
data

exports printer denition attributes to a SAS data set.

customizes ODS output.

creates SAS data sets using Structured Query Language

and SAS features.

DBCSTAB#

produces conversion tables for the double-byte character

sets that SAS supports.

FORMAT

creates user-dened informats to read data and

user-dened formats to display data.

SORT

sorts SAS data sets by one or more variables.

SQL

sorts SAS data sets by one or more variables.

TRANSPOSE

transforms SAS data sets so that observations become

variables and variables become observations.

TRANTAB#
Manage SAS les

browses external les such as les that contain SAS

source lines or SAS procedure output.

SQL
Transform data

FSLIST

creates, edits, and displays customized translation tables.

APPEND

appends one SAS data set to the end of another.

BMDP*

invokes a BMDP program to analyze data in a SAS data

set.

Choosing the Right Procedure

To perform these utility

tasks

Utility Procedures

Use this procedure

Which

CATALOG

manages SAS catalog entries.

CIMPORT

restores a transport sequential le that PROC CPORT

creates (usually under another operating environment) to
its original form as a SAS catalog, a SAS data set, or a
SAS library.

CONVERT*

converts BMDP system les, OSIRIS system les, and

SPSS portable les to SAS data sets.

COPY

copies a SAS data library or specic members of the

library.

CPORT

converts a SAS catalog, a SAS data set, or a SAS library

to a transport sequential le that PROC CIMPORT can
restore (usually under another operating environment) to
its original form.

CV2VIEW@

converts SAS/ACCESS view descriptors to PROC SQL

views.

DATASETS

manages SAS les.

EXPORT

reads data from a SAS data set and writes them to an

external data source.

IMPORT

reads data from an external data source and writes them

to a SAS data set.

PDS*

lists, deletes, and renames the members of a partitioned

data set.

PDSCOPY*

copies partitioned data sets from disk to tape, disk to

disk, tape to tape, or tape to disk.

REGISTRY

imports registry information to the USER portion of the

SAS registry.

RELEASE*

releases unused space at the end of a disk data set under

the z/OS environment.

SOURCE*

provides an easy way to back up and process source

library data sets.

SQL
TAPECOPY

concatenates SAS data sets.

copies an entire tape volume or les from one or more

tape volumes to one output tape volume.

TAPELABEL*

lists the label information of an IBM standard-labeled

tape volume under the z/OS environment.

Control windows

PMENU

creates customized pull-down menus for SAS applications.

Miscellaneous

PWENCODE

encodes passwords for use in SAS programs.

*
+
@
#

See
See
See
See

the SAS documentation for your operating environment for a description of these procedures.
SAS Output Delivery System: Users Guide for a description of these procedures.
SAS/ACCESS for Relational Databases: Reference for a description of this procedure.
SAS National Language Support (NLS): Users Guide for a description of these procedures.

Brief Descriptions of Base SAS Procedures

Chapter 1

Brief Descriptions of Base SAS Procedures

APPEND procedure
adds observations from one SAS data set to the end of another SAS data set.
BMDP procedure
invokes a BMDP program to analyze data in a SAS data set. See the SAS
documentation for your operating environment for more information.
CALENDAR procedure
displays data from a SAS data set in a monthly calendar format. PROC
CALENDAR can display holidays in the month, schedule tasks, and process data
for multiple calendars with work schedules that vary.
CATALOG procedure
manages entries in SAS catalogs. PROC CATALOG is an interactive,
nonwindowing procedure that enables you to display the contents of a catalog,
copy an entire catalog or specic entries in a catalog, and rename, exchange, or
delete entries in a catalog.
CHART procedure
produces vertical and horizontal bar charts, block charts, pie charts, and star
charts. These charts provide a quick visual representation of the values of a single
variable or several variables. PROC CHART can also display a statistic associated
with the values.
CIMPORT procedure
restores a transport le created by the CPORT procedure to its original form (a
SAS data library, catalog, or data set) in the format appropriate to the operating
environment. Coupled with the CPORT procedure, PROC CIMPORT enables you
to move SAS data libraries, catalogs, and data sets from one operating
environment to another.
COMPARE procedure
compares the contents of two SAS data sets. You can also use PROC COMPARE to
compare the values of different variables within a single data set. PROC
COMPARE produces a variety of reports on the comparisons that it performs.
CONTENTS procedure
prints descriptions of the contents of one or more les in a SAS data library.
CONVERT procedure
converts BMDP system les, OSIRIS system les, and SPSS portable les to SAS
data sets. See the SAS documentation for your operating environment for more
information.
COPY procedure
copies an entire SAS data library or specic members of the library. You can limit
processing to specic types of library members.
CORR procedure
computes Pearson product-moment and weighted product-moment correlation
coefcients between variables and descriptive statistics for these variables. In
addition, PROC CORR can compute three nonparametric measures of association
(Spearmans rank-order correlation, Kendalls tau-b, and Hoeffdings measure of
dependence, D), partial correlations (Pearsons partial correlation, Spearmans
partial rank-order correlation, and Kendalls partial tau-b), and Cronbachs
coefcient alpha.

Choosing the Right Procedure

Brief Descriptions of Base SAS Procedures

CPORT procedure
writes SAS data libraries, data sets, and catalogs in a special format called a
transport le. Coupled with the CIMPORT procedure, PROC CPORT enables you
to move SAS libraries, data sets, and catalogs from one operating environment to
another.
CV2VIEW procedure
converts SAS/ACCESS view descriptors to PROC SQL views. Starting in SAS
System 9, conversion of SAS/ACCESS view descriptors to PROC SQL views is
recommended because PROC SQL views are platform independent and enable you
to use the LIBNAME statement. See SAS/ACCESS for Relational Databases:
Reference for details.
DATASETS procedure
lists, copies, renames, and deletes SAS les and SAS generation groups, manages
indexes, and appends SAS data sets in a SAS data library. The procedure provides
all the capabilities of the APPEND, CONTENTS, and COPY procedures. You can
also modify variables within data sets, manage data set attributes, such as labels
and passwords, or create and delete integrity constraints.
DBCSTAB procedure
produces conversion tables for the double-byte character sets that SAS supports.
DOCUMENT procedure
manipulates procedure output that is stored in ODS documents. PROC
DOCUMENT enables a user to browse and edit output objects and hierarchies,
and to replay them to any supported ODS output format. See SAS Output Delivery
System: Users Guide for details.
EXPORT procedure
reads data from a SAS data set and writes it to an external data source.
FONTREG procedure
adds system fonts to the SAS registry.
FORMAT procedure
creates user-dened informats and formats for character or numeric variables.
PROC FORMAT also prints the contents of a format library, creates a control data
set to write other informats or formats, and reads a control data set to create
informats or formats.
FREQ procedure
produces one-way to n-way frequency tables and reports frequency counts. PROC
FREQ can compute chi-square tests for one-way to n-way tables, tests and
measures of association and of agreement for two-way to n-way cross-tabulation
tables, risks and risk difference for 222 tables, trends tests, and
Cochran-Mantel-Haenszel statistics. You can also create output data sets.
FSLIST procedure
displays the contents of an external le or copies text from an external le to the
SAS Text Editor.
IMPORT procedure
reads data from an external data source and writes them to a SAS data set.
MEANS procedure
computes descriptive statistics for numeric variables across all observations and
within groups of observations. You can also create an output data set that contains
specic statistics and identies minimum and maximum values for groups of
observations.

Brief Descriptions of Base SAS Procedures

Chapter 1

OPTIONS procedure
lists the current values of all SAS system options.
OPTLOAD procedure
reads SAS system option settings from the SAS registry or a SAS data set, and
puts them into effect.
OPTSAVE procedure
saves SAS system option settings to the SAS registry or a SAS data set.
PDS procedure
lists, deletes, and renames the members of a partitioned data set. See the SAS
documentation for your operating environment for more information.
PDSCOPY procedure
copies partitioned data sets from disk to tape, disk to disk, tape to tape, or tape to
disk. See the SAS documentation for your operating environment for more
information.
PLOT procedure
produces scatter plots that graph one variable against another. The coordinates of
each point on the plot correspond to the two variables values in one or more
observations of the input data set.
PMENU procedure
denes menus that you can use in DATA step windows, macro windows, and
SAS/AF windows, or in any SAS application that enables you to specify customized
menus.
PRINT procedure
prints the observations in a SAS data set, using all or some of the variables.
PROC PRINT can also print totals and subtotals for numeric variables.
PRINTTO procedure
denes destinations for SAS procedure output and the SAS log.
PRTDEF procedure
creates printer denitions for individual SAS users or all SAS users.
PRTEXP procedure
exports printer denition attributes to a SAS data set so that they can be easily
replicated and modied.
PWENCODE procedure
encodes passwords for use in SAS programs.
RANK procedure
computes ranks for one or more numeric variables across the observations of a
SAS data set. The ranks are written to a new SAS data set. Alternatively, PROC
RANK produces normal scores or other rank scores.
REGISTRY procedure
imports registry information into the USER portion of the SAS registry.
RELEASE procedure
releases unused space at the end of a disk data set in the z/OS environment. See
the SAS documentation for this operating environment for more information.
REPORT procedure
combines features of the PRINT, MEANS, and TABULATE procedures with
features of the DATA step in a single report-writing tool that can produce both
detail and summary reports.

Choosing the Right Procedure

Brief Descriptions of Base SAS Procedures

SORT procedure
sorts observations in a SAS data set by one or more variables. PROC SORT stores
the resulting sorted observations in a new SAS data set or replaces the original
data set.
SOURCE procedure
provides an easy way to back up and process source library data sets. See the SAS
documentation for your operating environment for more information.
SQL procedure
implements a subset of the Structured Query Language (SQL) for use in SAS. SQL
is a standardized, widely used language that retrieves and updates data in SAS
data sets, SQL views, and DBMS tables, as well as views based on those tables.
PROC SQL can also create tables and views, summaries, statistics, and reports
and perform utility functions such as sorting and concatenating.
STANDARD procedure
standardizes some or all of the variables in a SAS data set to a given mean and
standard deviation and produces a new SAS data set that contains the
standardized values.
SUMMARY procedure
computes descriptive statistics for the variables in a SAS data across all
observations and within groups of observations and outputs the results to a new
SAS data set.
TABULATE procedure
displays descriptive statistics in tabular form. The value in each table cell is
calculated from the variables and statistics that dene the pages, rows, and
columns of the table. The statistic associated with each cell is calculated on values
from all observations in that category. You can write the results to a SAS data set.
TAPECOPY procedure
copies an entire tape volume or les from one or more tape volumes to one output
tape volume. See the SAS documentation for your operating environment for more
information.
TAPELABEL procedure
lists the label information of an IBM standard-labeled tape volume under the z/OS
environment. See the SAS documentation for this operating environment for more
information.
TEMPLATE procedure
customizes ODS output for an entire SAS job or a single ODS output object. See
SAS Output Delivery System: Users Guide for details.
TIMEPLOT procedure
produces plots of one or more variables over time intervals.
TRANSPOSE procedure
transposes a data set that changes observations into variables and vice versa.
TRANTAB procedure
creates, edits, and displays customized translation tables.
UNIVARIATE procedure
computes descriptive statistics (including quantiles), condence intervals, and
robust estimates for numeric variables. Provides detail on the distribution of
numeric variables, which include tests for normality, plots to illustrate the
distribution, frequency tables, and tests of location.

CHAPTER

2
Fundamental Concepts for Using
Base SAS Procedures
Language Concepts 16
Temporary and Permanent SAS Data Sets 16
Naming SAS Data Sets 16
USER Data Library 17
SAS System Options 17
Data Set Options 18
Global Statements 18
Procedure Concepts 19
Input Data Sets 19
RUN-Group Processing 19
Creating Titles That Contain BY-Group Information 20
BY-Group Processing 20
Suppressing the Default BY Line 20
Inserting BY-Group Information into a Title 20
Example: Inserting a Value from Each BY Variable into the Title 21
Example: Inserting the Name of a BY Variable into a Title 22
Example: Inserting the Complete BY Line into a Title 23
Error Processing of BY-Group Specications 24
Shortcuts for Specifying Lists of Variable Names 24
Formatted Values 25
Using Formatted Values 25
Example: Printing the Formatted Values for a Data Set 25
Example: Grouping or Classifying Formatted Data 27
Example: Temporarily Associating a Format with a Variable 28
Example: Temporarily Dissociating a Format from a Variable 29
Formats and BY-Group Processing 30
Formats and Error Checking 30
Processing All the Data Sets in a Library 30
Operating Environment-Specic Procedures 30
Statistic Descriptions 31
Computational Requirements for Statistics 32
Output Delivery System 32
What Is the Output Delivery System? 32
Gallery of ODS Samples 33
Introduction to the ODS Samples 33
Listing Output 33
PostScript Output 35
HTML Output 35
RTF Output 36
PDF Output 37
XML Output 38

Language Concepts

Chapter 2

Commonly Used ODS Terminology 40

How Does ODS Work? 41
Components of SAS Output 41
Features of ODS 42
What Are the ODS Destinations? 43
Overview of ODS Destination Categories 43
Denition of Destination-Independent Input 43
The SAS Formatted Destinations 44
The Third-Party Formatted Destinations 45
What Controls the Formatting Features of Third-Party Formats? 46
ODS Destinations and System Resources 47
What Are Table Denitions, Table Elements, and Table Attributes? 47
What Are Style Denitions, Style Elements, and Style Attributes? 47
What Style Denitions Are Shipped with SAS Software? 48
How Do I Use Style Denitions with Base SAS Procedures? 48
Changing SAS Registry Settings for ODS 49
Overview of ODS and the SAS Registry 49
Changing Your Default HTML Version Setting 50
Changing ODS Destination Default Settings 51
Customized ODS Output 51
SAS Output 51
Selection and Exclusion Lists 52
How Does ODS Determine the Destinations for an Output Object? 52
Customized Output for an Output Object 53
Summary of ODS 54

Language Concepts

Temporary and Permanent SAS Data Sets

Naming SAS Data Sets

SAS data sets can have a one-level name or a two-level name. Typically, names of
temporary SAS data sets have only one level and are stored in the WORK data library.
The WORK data library is dened automatically at the beginning of the SAS session
and is automatically deleted at the end of the SAS session. Procedures assume that SAS
data sets that are specied with a one-level name are to be read from or written to the
WORK data library, unless you specify a USER data library (see USER Data Library
on page 17). For example, the following PROC PRINT steps are equivalent. The second
PROC PRINT step assumes that the DEBATE data set is in the WORK data library:
proc print data=work.debate;
run;
proc print data=debate;
run;

Fundamental Concepts for Using Base SAS Procedures

SAS System Options

The SAS system options WORK=, WORKINIT, and WORKTERM affect how you
work with temporary and permanent libraries. See SAS Language Reference:
Dictionary for complete documentation.
Typically, two-level names represent permanent SAS data sets. A two-level name
takes the form libref.SAS-data-set. The libref is a name that is temporarily associated
with a SAS data library. A SAS data library is an external storage location that stores
SAS data sets in your operating environment. A LIBNAME statement associates the
libref with the SAS data library. In the following PROC PRINT step, PROCLIB is the
libref and EMP is the SAS data set within the library:
libname proclib SAS-data-library;
proc print data=proclib.emp;
run;

USER Data Library

You can use one-level names for permanent SAS data sets by specifying a USER data
library. You can assign a USER data library with a LIBNAME statement or with the
SAS system option USER=. After you specify a USER data library, the procedure
assumes that data sets with one-level names are in the USER data library instead of
the WORK data library. For example, the following PROC PRINT step assumes that
DEBATE is in the USER data library:
options user=SAS-data-library;
proc print data=debate;
run;

Note: If you have a USER data library dened, then you can still use the WORK
data library by specifying WORK.SAS-data-set.

SAS System Options

Some SAS system option settings affect procedure output. The following are the SAS
system options that you are most likely to use with SAS procedures:
BYLINE|NOBYLINE
DATE|NODATE
DETAILS|NODETAILS
FMTERR|NOFMTERR
FORMCHAR=
FORMDLIM=
LABEL|NOLABEL
LINESIZE=
NUMBER|NONUMBER
PAGENO=
PAGESIZE=
REPLACE|NOREPLACE
SOURCE|NOSOURCE
For a complete description of SAS system options, see SAS Language Reference:
Dictionary.

Data Set Options

Chapter 2

Data Set Options

Most of the procedures that read data sets or create output data sets accept data set
options. SAS data set options appear in parentheses after the data set specication.
Here is an example:
proc print data=stocks(obs=25 pw=green);

The individual procedure chapters contain reminders that you can use data set
options where it is appropriate.
SAS data set options are
ALTER=

OBS=

BUFNO=

OBSBUF=

BUFSIZE=

OUTREP=

CNTLLEV=

POINTOBS=

COMPRESS=

PW=

DLDMGACTION=

PWREQ=

DROP=

READ=

ENCODING=

RENAME=

ENCRYPT=

REPEMPTY=

FILECLOSE=

REPLACE=

FIRSTOBS=

REUSE=

GENMAX=

SORTEDBY=

GENNUM=

SORTSEQ=

IDXNAME=

SPILL=

IDXWHERE=

TOBSNO=

IN=

TYPE=

INDEX=

WHERE=

KEEP=

WHEREUP=

LABEL=

WRITE=

For a complete description of SAS data set options, see SAS Language Reference:
Dictionary.

Global Statements
You can use these global statements anywhere in SAS programs except after a
DATALINES, CARDS, or PARMCARDS statement:
comment

ODS

OPTIONS

ENDSAS

PAGE

Fundamental Concepts for Using Base SAS Procedures

FILENAME

SASFILE

LIBNAME

SKIP

%LIST

TITLE

LOCK

%RUN

%INCLUDE

RUN-Group Processing

RUN

FOOTNOTE

For information about all but the ODS statement, refer to SAS Language Reference:
Dictionary. For information about the ODS statement, refer to Output Delivery
System on page 32 and to SAS Output Delivery System: Users Guide.

Procedure Concepts

Input Data Sets

Many base procedures require an input SAS data set. You specify the input SAS data
set by using the DATA= option in the procedure statement, as in this example:
proc print data=emp;

If you omit the DATA= option, the procedure uses the value of the SAS system option
_LAST_=. The default of _LAST_= is the most recently created SAS data set in the
current SAS job or session. _LAST_= is described in detail in SAS Language Reference:
Dictionary.

RUN-Group Processing
RUN-group processing enables you to submit a PROC step with a RUN statement
without ending the procedure. You can continue to use the procedure without issuing
another PROC statement. To end the procedure, use a RUN CANCEL or a QUIT
statement. Several base SAS procedures support RUN-group processing:
CATALOG
DATASETS
PLOT
PMENU
TRANTAB
See the section on the individual procedure for more information.
Note: PROC SQL executes each query automatically. Neither the RUN nor RUN
CANCEL statement has any effect. 4

Creating Titles That Contain BY-Group Information

Chapter 2

Creating Titles That Contain BY-Group Information

BY-Group Processing
BY-group processing uses a BY statement to process observations that are ordered,
grouped, or indexed according to the values of one or more variables. By default, when
you use BY-group processing in a procedure step, a BY line identies each group. This
section explains how to create titles that serve as customized BY lines.

Suppressing the Default BY Line

When you insert BY-group processing information into a title, you usually want to
eliminate the default BY line. To suppress it, use the SAS system option NOBYLINE.
Note: You must use the NOBYLINE option if you insert BY-group information into
titles for the following base SAS procedures:
MEANS
PRINT
STANDARD
SUMMARY
If you use the BY statement with the NOBYLINE option, then these procedures always
start a new page for each BY group. This behavior prevents multiple BY groups from
appearing on a single page and ensures that the information in the titles matches the
report on the pages. 4

Inserting BY-Group Information into a Title

The general form for inserting BY-group information into a title is
#BY-specication<.sufx>
BY-specication
is one of the following:
BYVALn | BYVAL(BY-variable)
places the value of the specied BY variable in the title. You specify the BY
variable with one of the following:
n
is the nth BY variable in the BY statement.
BY-variable
is the name of the BY variable whose value you want to insert in the
title.
BYVARn | BYVAR(BY-variable)
places the label or the name (if no label exists) of the specied BY variable in
the title. You designate the BY variable with one of the following:
n
is the nth BY variable in the BY statement.
BY-variable
is the name of the BY variable whose name you want to insert in the
title.

Fundamental Concepts for Using Base SAS Procedures

Creating Titles That Contain BY-Group Information

BYLINE
inserts the complete default BY line into the title.
sufx
supplies text to place immediately after the BY-group information that you insert
in the title. No space appears between the BY-group information and the sufx.

Example: Inserting a Value from Each BY Variable into the Title

This example
1 creates a data set, GROC, that contains data for stores from four regions. Each

store has four departments. See GROC on page 1388 for the DATA step that
creates the data set.
2 sorts the data by Region and Department.
3 uses the SAS system option NOBYLINE to suppress the BY line that normally

appears in output that is produced with BY-group processing.

4 uses PROC CHART to chart sales by Region and Department. In the rst TITLE

statement, #BYVAL2 inserts the value of the second BY variable, Department, into
the title. In the second TITLE statement, #BYVAL(Region) inserts the value of
Region into the title. The rst period after Region indicates that a sufx follows.
The second period is the sufx.
5 uses the SAS system option BYLINE to return to the creation of the default BY

line with BY-group processing.

data groc; u
input Region $9. Manager $ Department $ Sales;
datalines;
Southeast
Hayes
Paper
250
Southeast
Hayes
Produce
100
Southeast
Hayes
Canned
120
Southeast
Hayes
Meat
80
...more lines of data...
Northeast
Fuller
Paper
200
Northeast
Fuller
Produce
300
Northeast
Fuller
Canned
420
Northeast
Fuller
Meat
125
;

proc sort data=groc; v

by region department;
run;
options nobyline nodate pageno=1
linesize=64 pagesize=20; w
proc chart data=groc; x
by region department;
vbar manager / type=sum sumvar=sales;
title1 This chart shows #byval2 sales;
title2 in the #byval(region)..;
run;
options byline;
y

Creating Titles That Contain BY-Group Information

Chapter 2

This partial output shows two BY groups with customized BY lines:

This chart shows Canned sales
in the Northwest.

Sales Sum
400 +
*****
*****
|
*****
*****
300 +
*****
*****
|
*****
*****
*****
200 +
*****
*****
*****
|
*****
*****
*****
100 +
*****
*****
*****
|
*****
*****
*****
-------------------------------------------Aikmann
Duncan
Jeffreys
Manager

This chart shows Meat sales

in the Northwest.

Sales Sum
75 +
*****
*****
|
*****
*****
60 +
*****
*****
|
*****
*****
45 +
*****
*****
|
*****
*****
30 +
*****
*****
*****
|
*****
*****
*****
15 +
*****
*****
*****
|
*****
*****
*****
-------------------------------------------Aikmann
Duncan
Jeffreys
Manager

Example: Inserting the Name of a BY Variable into a Title

This example inserts the name of a BY variable and the value of a BY variable into
the title. The program
1 uses the SAS system option NOBYLINE to suppress the BY line that normally

appears in output that is produced with BY-group processing.

2 uses PROC CHART to chart sales by Region. In the rst TITLE statement,

#BYVAR(Region) inserts the name of the variable Region into the title. (If Region
had a label, #BYVAR would use the label instead of the name.) The sufx al is
appended to the label. In the second TITLE statement, #BYVAL1 inserts the value
of the rst BY variable, Region, into the title.
3 uses the SAS system option BYLINE to return to the creation of the default BY

line with BY-group processing.

options nobyline nodate pageno=1

linesize=64 pagesize=20;
proc chart data=groc;
v
by region;

Fundamental Concepts for Using Base SAS Procedures

Creating Titles That Contain BY-Group Information

vbar manager / type=mean sumvar=sales;

title1 #byvar(region).al Analysis;
title2 for the #byval1;
run;
options byline; w

This partial output shows one BY group with a customized BY line:

Regional Analysis
for the Northwest

Sales Mean
300 +
*****
|
*****
200 +
*****
*****
100 +
*****
*****
*****
|
*****
*****
*****
-------------------------------------------Aikmann
Duncan
Jeffreys
Manager

Example: Inserting the Complete BY Line into a Title

This example inserts the complete BY line into the title. The program
1 uses the SAS system option NOBYLINE to suppress the BY line that normally
appears in output that is produced with BY-group processing.
2 uses PROC CHART to chart sales by Region and Department. In the TITLE
statement, #BYLINE inserts the complete BY line into the title.
3 uses the SAS system option BYLINE to return to the creation of the default BY
line with BY-group processing.

options nobyline nodate pageno=1

linesize=64 pagesize=20; u
proc chart data=groc; v
by region department;
vbar manager / type=sum sumvar=sales;
title Information for #byline;
run;
options byline; w

Shortcuts for Specifying Lists of Variable Names

Chapter 2

This partial output shows two BY groups with customized BY lines:

Information for Region=Northwest Department=Canned

Information for Region=Northwest Department=Meat

Error Processing of BY-Group Specications

SAS does not issue error or warning messages for incorrect #BYVAL, #BYVAR, or
#BYLINE specications. Instead, the text of the item simply becomes part of the title.

Shortcuts for Specifying Lists of Variable Names

Several statements in procedures allow multiple variable names. You can use these
shortcut notations instead of specifying each variable name:
Notation

Meaning

x1-xn

species variables X1 through Xn. The numbers must be

consecutive.

species all variables that begin with the letter X.

x--a

species all variables between X and A, inclusive. This

notation uses the position of the variables in the data set.

x-numeric-a

species all numeric variables between X and A, inclusive.

This notation uses the position of the variables in the data set.

Fundamental Concepts for Using Base SAS Procedures

Formatted Values

Notation

Meaning

x-character-a

species all character variables between X and A, inclusive.

This notation uses the position of the variables in the data set.

_numeric_

species all numeric variables.

_character_

species all character variables.

_all_

species all variables.

Note: You cannot use shortcuts to list variable names in the INDEX CREATE
statement in PROC DATASETS. 4
See SAS Language Reference: Concepts for complete documentation.

Formatted Values
Using Formatted Values
Typically, when you print or group variable values, base SAS procedures use the
formatted values. This section contains examples of how base procedures use formatted
values.

Example: Printing the Formatted Values for a Data Set

The following example prints the formatted values of the data set
PROCLIB.PAYROLL. (See PROCLIB.PAYROLL on page 1395 for the DATA step that
creates this data set.) In PROCLIB.PAYROLL, the variable Jobcode indicates the job
and level of the employee. For example, TA1 indicates that the employee is at the
beginning level for a ticket agent.
libname proclib SAS-data-library;

options nodate pageno=1

linesize=64 pagesize=40;
proc print data=proclib.payroll(obs=10)
noobs;
title PROCLIB.PAYROLL;
title2 First 10 Observations Only;
run;

Formatted Values

Chapter 2

This is a partial printing of PROCLIB.PAYROLL:

PROCLIB.PAYROLL
First 10 Observations Only
Id
Number
1919
1653
1400
1350
1401
1499
1101
1333
1402
1479

Gender
M
F
M
F
M
M
M
M
M
F

Jobcode
TA2
ME2
ME1
FA3
TA3
ME3
SCP
PT2
TA2
TA3

Salary

Birth

Hired

34376
35108
29769
32886
38822
43025
18723
88606
32615
38785

12SEP60
15OCT64
05NOV67
31AUG65
13DEC50
26APR54
06JUN62
30MAR61
17JAN63
22DEC68

04JUN87
09AUG90
16OCT90
29JUL90
17NOV85
07JUN80
01OCT90
10FEB81
02DEC90
05OCT89

The following PROC FORMAT step creates the format $JOBFMT., which assigns
descriptive names for each job:
proc format;
value $jobfmt
FA1=Flight Attendant Trainee
FA2=Junior Flight Attendant
FA3=Senior Flight Attendant
ME1=Mechanic Trainee
ME2=Junior Mechanic
ME3=Senior Mechanic
PT1=Pilot Trainee
PT2=Junior Pilot
PT3=Senior Pilot
TA1=Ticket Agent Trainee
TA2=Junior Ticket Agent
TA3=Senior Ticket Agent
NA1=Junior Navigator
NA2=Senior Navigator
BCK=Baggage Checker
SCP=Skycap;
run;

The FORMAT statement in this PROC MEANS step temporarily associates the
$JOBFMT. format with the variable Jobcode:
options nodate pageno=1
linesize=64 pagesize=60;
proc means data=proclib.payroll mean max;
class jobcode;
var salary;
format jobcode $jobfmt.;
title Summary Statistics for;
title2 Each Job Code;
run;

Fundamental Concepts for Using Base SAS Procedures

Formatted Values

PROC MEANS produces this output, which uses the $JOBFMT. format:
Summary Statistics for
Each Job Code

The MEANS Procedure

Analysis Variable : Salary
N
Jobcode
Obs
Mean
Maximum
--------------------------------------------------------------Baggage Checker
9
25794.22
26896.00
Flight Attendant Trainee

23039.36

23979.00

Junior Flight Attendant

27986.88

28978.00

Senior Flight Attendant

32933.86

33419.00

Mechanic Trainee

28500.25

29769.00

Junior Mechanic

35576.86

36925.00

Senior Mechanic

42410.71

43900.00

Junior Navigator

42032.20

43433.00

Senior Navigator

52383.00

53798.00

Pilot Trainee

67908.00

71349.00

Junior Pilot

87925.20

91908.00

Senior Pilot

10504.50

11379.00

Skycap

18308.86

18833.00

Ticket Agent Trainee

27721.33

28880.00

33574.95

34803.00

Junior Ticket Agent

Senior Ticket Agent

12
39679.58
40899.00
---------------------------------------------------------------

Note: Because formats are character strings, formats for numeric variables are
ignored when the values of the numeric variables are needed for mathematical
calculations. 4

Example: Grouping or Classifying Formatted Data

If you use a formatted variable to group or classify data, then the procedure uses the
formatted values. The following example creates and assigns a format, $CODEFMT.,
that groups the levels of each job code into one category. PROC MEANS calculates
statistics based on the groupings of the $CODEFMT. format.
proc format;
value $codefmt
FA1,FA2,FA3=Flight Attendant
ME1,ME2,ME3=Mechanic
PT1,PT2,PT3=Pilot
TA1,TA2,TA3=Ticket Agent
NA1,NA2=Navigator
BCK=Baggage Checker

Formatted Values

Chapter 2

SCP=Skycap;
run;
options nodate pageno=1
linesize=64 pagesize=40;
proc means data=proclib.payroll mean max;
class jobcode;
var salary;
format jobcode $codefmt.;
title Summary Statistics for Job Codes;
title2 (Using a Format that Groups the Job Codes);
run;

PROC MEANS produces this output:

Summary Statistics for Job Codes
(Using a Format that Groups the Job Codes)

The MEANS Procedure

Analysis Variable : Salary
N
Jobcode
Obs
Mean
Maximum
------------------------------------------------------Baggage Checker
9
25794.22
26896.00
Flight Attendant

27404.71

33419.00

Mechanic

35274.24

43900.00

45913.75

53798.00

72176.25

91908.00

18308.86

18833.00

Navigator
Pilot
Skycap

Ticket Agent
41
34076.73
40899.00
-------------------------------------------------------

Example: Temporarily Associating a Format with a Variable

If you want to associate a format with a variable temporarily, then you can use the
FORMAT statement. For example, the following PROC PRINT step associates the
DOLLAR8. format with the variable Salary for the duration of this PROC PRINT step
only:
options nodate pageno=1
linesize=64 pagesize=40;
proc print data=proclib.payroll(obs=10)
noobs;
format salary dollar8.;
title Temporarily Associating a Format;
title2 with the Variable Salary;
run;

Fundamental Concepts for Using Base SAS Procedures

Formatted Values

PROC PRINT produces this output:

Temporarily Associating a Format
with the Variable Salary
Id
Number
1919
1653
1400
1350
1401
1499
1101
1333
1402
1479

Gender
M
F
M
F
M
M
M
M
M
F

Jobcode
TA2
ME2
ME1
FA3
TA3
ME3
SCP
PT2
TA2
TA3

Salary

Birth

Hired

$34,376
$35,108
$29,769
$32,886
$38,822
$43,025
$18,723
$88,606
$32,615
$38,785

12SEP60
15OCT64
05NOV67
31AUG65
13DEC50
26APR54
06JUN62
30MAR61
17JAN63
22DEC68

04JUN87
09AUG90
16OCT90
29JUL90
17NOV85
07JUN80
01OCT90
10FEB81
02DEC90
05OCT89

Example: Temporarily Dissociating a Format from a Variable

If a variable has a permanent format that you do not want a procedure to use, then
temporarily dissociate the format from the variable by using a FORMAT statement.
In this example, the FORMAT statement in the DATA step permanently associates
the $YRFMT. variable with the variable Year. Thus, when you use the variable in a
PROC step, the procedure uses the formatted values. The PROC MEANS step, however,
contains a FORMAT statement that dissociates the $YRFMT. format from Year for this
PROC MEANS step only. PROC MEANS uses the stored value for Year in the output.
proc format;
value $yrfmt

1=Freshman
2=Sophomore
3=Junior
4=Senior;

run;
data debate;
input Name $ Gender $
format year $yrfmt.;
datalines;
Capiccio m 1 3.598 Tucker
Bagwell f 2 3.722 Berry
Metcalf m 2 3.342 Gold
Gray
f 3 3.177 Syme
Baglione f 4 4.000 Carr
Hall
m 4 3.574 Lewis
;

Year $

m
m
f
f
m
m

1
2
3
3
4
4

GPA

3.901
3.198
3.609
3.883
3.750
3.421

options nodate pageno=1

linesize=64 pagesize=40;
proc means data=debate mean maxdec=2;
class year;
format year;
title Average GPA;
run;

@@;

Processing All the Data Sets in a Library

Chapter 2

PROC MEANS produces this output, which does not use the YRFMT. format:
Average GPA

The MEANS Procedure

Analysis Variable : GPA
N
Year
Obs
Mean
------------------------------1
2
3.75
2

3.42

3.56

4
4
3.69
-------------------------------

Formats and BY-Group Processing

When a procedure processes a data set, it checks to see if a format is assigned to the
BY variable. If it is, then the procedure adds observations to the current BY groups
until the formatted value changes. If nonconsecutive internal values of the BY
variable(s) have the same formatted value, then the values are grouped into different
BY groups. This results in two BY groups with the same formatted value. Further, if
different and consecutive internal values of the BY variable(s) have the same formatted
value, then they are included in the same BY group.

Formats and Error Checking

If SAS cannot nd a format, then it stops processing and prints an error message in
the SAS log. You can suppress this behavior with the SAS system option NOFMTERR.
If you use NOFMTERR, and SAS cannot nd the format, then SAS uses a default
format and continues processing. Typically, for the default, SAS uses the BESTw.
format for numeric variables and the $w. format for character variables.
Note: To ensure that SAS can nd user-written formats, use the SAS system option
FMTSEARCH=. How to store formats is described in Storing Informats and Formats
on page 456. 4

Processing All the Data Sets in a Library

You can use the SAS Macro Facility to run the same procedure on every data set in a
library. The macro facility is part of base SAS software.
Example 9 on page 767 shows how to print all the data sets in a library. You can use
the same macro denition to perform any procedure on all the data sets in a library.
Simply replace the PROC PRINT piece of the program with the appropriate procedure
code.

Operating Environment-Specic Procedures

Several base SAS procedures are specic to one operating environment or one
release. Appendix 2, Operating Environment-Specic Procedures, on page 1375
contains a table with additional information. These procedures are described in more
detail in the SAS documentation for operating environments.

Fundamental Concepts for Using Base SAS Procedures

Statistic Descriptions

Statistic Descriptions
Table 2.1 on page 31 identies common descriptive statistics that are available in
several Base SAS procedures. See Keywords and Formulas on page 1340 for more
detailed information about available statistics and theoretical information.
Table 2.1 Common Descriptive Statistics That Base Procedures Calculate

Statistic

Description

condence intervals

Procedures
FREQ, MEANS/SUMMARY, TABULATE, UNIVARIATE

CSS

corrected sum of
squares

CORR, MEANS/SUMMARY, REPORT, SQL,

TABULATE, UNIVARIATE

coefcient of variation

MEANS/SUMMARY, REPORT, SQL, TABULATE,

UNIVARIATE

goodness-of-t tests

FREQ, UNIVARIATE

KURTOSIS

kurtosis

MEANS/SUMMARY, TABULATE, UNIVARIATE

MAX

largest (maximum)
value

CORR, MEANS/SUMMARY, REPORT, SQL,

TABULATE, UNIVARIATE

MEAN

mean

CORR, MEANS/SUMMARY, REPORT, SQL,

TABULATE, UNIVARIATE

MEDIAN

median (50th percentile)

CORR (for nonparametric correlation measures),

MEANS/SUMMARY, TABULATE, UNIVARIATE

MIN

smallest (minimum)
value

CORR, MEANS/SUMMARY, REPORT, SQL,

TABULATE, UNIVARIATE

TABULATE, UNIVARIATE

Computational Requirements for Statistics

Chapter 2

Statistic

Description

Procedures

robust statistics

trimmed means,
Winsorized means

UNIVARIATE

SKEWNESS

skewness

MEANS/SUMMARY, TABULATE, UNIVARIATE

Spearman correlation

CORR

STD

standard deviation

CORR, MEANS/SUMMARY, REPORT, SQL,

Output Delivery System

What Is the Output Delivery System?
The Output Delivery System (ODS) gives you greater exibility in generating,
storing, and reproducing SAS procedure and DATA step output, with a wide range of
formatting options. ODS provides formatting functionality that is not available from
individual procedures or from the DATA step alone. ODS overcomes these limitations
and enables you to format your output more easily.
Prior to Version 7, most SAS procedures generated output that was designed for a
traditional line-printer. This type of output has limitations that prevents you from
getting the most value from your results:

Fundamental Concepts for Using Base SAS Procedures

Gallery of ODS Samples

3 Traditional SAS output is limited to monospace fonts. With todays desktop

document editors and publishing systems, you need more versatility in printed
output.
3 Some commonly used procedures do not produce output data sets. Prior to ODS, if
you wanted to use output from one of these procedures as input to another
procedure, then you relied on PROC PRINTTO and the DATA step to retrieve
results.

Gallery of ODS Samples

Introduction to the ODS Samples

This section shows you samples of the different kinds of formatted output that you
can produce with ODS. The input le contains sales records for TruBlend Coffee
Makers, a company that distributes coffee machines.

Listing Output
Traditional SAS output is Listing output. You do not need to change your SAS
programs to create listing output. By default, you continue to create this kind of output
even if you also create a type of output that contains more formatting.

Gallery of ODS Samples

Output 2.1

Chapter 2

Listing Output
Average Quarterly Sales Amount by Each Sales Representative

--------------------------------- Quarter=1 ---------------------------------The MEANS Procedure

Analysis Variable : AmountSold
N
SalesRep
Obs
N
Mean
Std Dev
Minimum
Maximum
____________________________________________________________________________
Garcia
8
8
14752.5
22806.1
495.0
63333.7
Hollingsworth

11926.9

12165.2

774.3

31899.1

Jensen
5
5
10015.7
8009.5
3406.7
20904.8
____________________________________________________________________________

Average Quarterly Sales Amount by Each Sales Representative

--------------------------------- Quarter=2 ---------------------------------The MEANS Procedure

Analysis Variable : AmountSold
N
SalesRep
Obs
N
Mean
Std Dev
Minimum
Maximum
____________________________________________________________________________
Garcia
6
6
18143.3
20439.6
1238.8
53113.6
Hollingsworth

16026.8

14355.0

1237.5

34686.4

Jensen
6
6
12455.1
12713.7
1393.7
34376.7
____________________________________________________________________________

Average Quarterly Sales Amount by Each Sales Representative

--------------------------------- Quarter=3 ---------------------------------The MEANS Procedure

Analysis Variable : AmountSold
N
SalesRep
Obs
N
Mean
Std Dev
Minimum
Maximum
____________________________________________________________________________
Garcia
21
21
10729.8
11457.0
2787.3
38712.5
Hollingsworth

7313.6

7280.4

1485.0

30970.0

Jensen
21
21
10585.3
7361.7
2227.5
27129.7
____________________________________________________________________________

Average Quarterly Sales Amount by Each Sales Representative

--------------------------------- Quarter=4 ---------------------------------The MEANS Procedure

Analysis Variable : AmountSold
N
SalesRep
Obs
N
Mean
Std Dev
Minimum
Maximum
____________________________________________________________________________
Garcia
5
5
11973.0
10971.8
3716.4
30970.0
Hollingsworth

13624.4

12624.6

5419.8

38093.1

Jensen
6
6
19010.4
15441.0
1703.4
38836.4
____________________________________________________________________________

Fundamental Concepts for Using Base SAS Procedures

Fundamental Concepts for Using Base SAS Procedures

Output 2.2 XML Output le

<?xml version="1.0" encoding="windows-1252"?>
<odsxml>
<head>
<meta operator="user"/>
</head>
<body>
<proc name="Print">
<label name="IDX"/>
<title class="SystemTitle" toc-level="1">US Census of Population and Housing</title>
<branch name="Print" label="The Print Procedure" class="ContentProcName" toc-level="1">
<leaf name="Print" label="Data Set SASHELP.CLASS" class="ContentItem" toc-level="2">
<output name="Print" label="Data Set SASHELP.CLASS" clabel="Data Set SASHELP.CLASS">
<output-object type="table" class="Table">
<style>
<border spacing="1" padding="7" rules="groups" frame="box"/>
</style>
<colspecs columns="6">
<colgroup>
<colspec name="1" width="2" align="right" type="int"/>
</colgroup>
<colgroup>
<colspec name="2" width="7" type="string"/>
<colspec name="3" width="1" type="string"/>
<colspec name="4" width="2" align="decimal" type="double"/>
<colspec name="5" width="4" align="decimal" type="double"/>
<colspec name="6" width="5" align="decimal" type="double"/>
</colgroup>
</colspecs>
<output-head>
<row>
<header type="string" class="Header" row="1" column="1">
<value>Obs</value>
</header>
<header type="string" class="Header" row="1" column="2">
<value>Name</value>
</header>
<header type="string" class="Header" row="1" column="3">
<value>Sex</value>
</header>
<header type="string" class="Header" row="1" column="4">
<value>Age</value>
</header>
<header type="string" class="Header" row="1" column="5">
<value>Height</value>
</header>
<header type="string" class="Header" row="1" column="6">
<value>Weight</value>
</header>
</row>
</output-head>
<output-body>
<row>
<header type="double" class="RowHeader" row="2" column="1">
<value> 1</value>
</header>
<data type="string" class="Data" row="2" column="2">
<value>Alfred</value>
</data>
... more xml tagged output...
<
/odsxml>

Gallery of ODS Samples

Commonly Used ODS Terminology

Chapter 2

Commonly Used ODS Terminology

data component
is a form, similar to a SAS data set, that contains the results (numbers and
characters) of a DATA step or PROC step that supports ODS.
table denition
is a set of instructions that describes how to format the data. This description
includes but is not limited to

3 the order of the columns

3 text and order of column headings
3 formats for data
3 font sizes and font faces.
output object
is an object that contains both the results of a DATA step or PROC step and
information about how to format the results. An output object has a name, label,
and path. For example, the Basic Statistical Measurement table generated from
the UNIVARIATE procedure is an output object. It contains the data component
and formatted presentation of the mean, median, mode, standard deviation,
variance, range, and interquartile range.
Note: Although many output objects include formatting instructions, not all of
them do. In some cases the output object consists of only the data component. 4
ODS destinations
are designations that produce specic types of output. ODS supports a number of
destinations, including the following:
LISTING
produces traditional SAS output (monospace format).
Markup Languages
produce SAS output that is formatted using one of many different markup
languages such as HTML (Hypertext Markup Language), XML (Extensible
Markup Language), and LaTeX that you can access with a web browser. SAS
supplies many markup languages for you to use ranging from DOCBOOK to
TROFF. You can specify a markup language that SAS supplies or create one
of your own and store it as a user-dened markup language.
DOCUMENT
produces a hierarchy of output objects that enables you to produce multiple
ODS output formats without rerunning a PROC or DATA step and gives you
more control over the structure of the output.
OUTPUT
produces a SAS data set.
Printer Family
produces output that is formatted for a high-resolution printer such as a
PostScript (PS), PDF, or PCL le.
RTF
produces output that is formatted for use with Microsoft Word.

Fundamental Concepts for Using Base SAS Procedures

How Does ODS Work?

ODS output
ODS output consists of formatted output from any of the ODS destinations. For
example, the OUTPUT destination produces SAS data sets; the LISTING
destination produces listing output; the HTML destination produces output that is
formatted in Hypertext Markup Language.

How Does ODS Work?

Components of SAS Output

The PROC or DATA step supplies raw data and the name of the table denition that
contains the formatting instructions, and ODS formats the output. You can use the
Output Delivery System to format output from individual procedures and from the
DATA step in many different forms other than the default SAS listing output.
The following gure shows how SAS produces ODS output.

Figure 2.1 ODS Processing: What Goes in and What Comes Out

ODS Processing: What Goes In and What Comes Out

Data
Component

Table
Definition
+

Output
Object

DOCUMENT LISTING OUTPUT

Document
Output

Listing
Output

SAS
Data Set

MARKUP

HTML

HTML3.2
Output

RTF

MS PS PCL PDF RTF

Windows
Output
Printers

SAS
TAGSETS*

SAS Formatted Destinations

PRINTER

User-defined
TAGSETS

Third-Party Formatted Destinations

ODS
Destinations

ODS
Outputs

How Does ODS Work?

Chapter 2

* List of Tagsets that SAS Supplies and Supports

Table 2.2 * List of Tagsets that SAS Supplies and Supports
CHTML

HTML4

SASIOXML

SASXMOH

CSVALL

HTMLCSS

SASREPORT

SASXMOIM

DEFAULT

IMODE

SASXML

SASXMOR

DOCBOOK

PHTML

SASXMOG

WML

EVENT_MAP

* List of Tagsets that SAS Supplies but Does Not Support

Table 2.3 Additional Tagsets that SAS Supplies but Does Not Support
COLORLATEX

LATEX

SHORT_MAP

TPL_STYLE_MAP

CSV

LATEX2

STYLE_DISPLAY

TROFF

CSVBYLINE

NAMEDHTML

STYLE_POPUP

WMLOLIST

GRAPH

ODSSTYLE

TEXT_MAP

GTABLEAPPLET

PYX

TPL_STYLE_LIST

CAUTION:

These tagsets are experimental tagsets. Do not use these tagsets in production jobs.

Features of ODS
ODS is designed to overcome the limitations of traditional SAS output and to make it
easy to access and create the new formatting options. ODS provides a method of
delivering output in a variety of formats, and makes the formatted output easy to access.
Important features of ODS include the following:
3 ODS combines raw data with one or more table denitions to produce one or more
output objects. These objects can be sent to any or all ODS destinations. You
control the specic type of output from ODS by selecting an ODS destination. The
currently available ODS destinations can produce
3 traditional monospace output
3 an output data set
3 an ODS document that contains a hierarchy le of the output objects
3 output that is formatted for a high-resolution printer such as PostScript and
PDF
3 output that is formatted in various markup languages such as HTML
3 RTF output that is formatted for use with Microsoft Word.

3 ODS provides table denitions that dene the structure of the output from SAS
procedures and from the DATA step. You can customize the output by modifying
these denitions, or by creating your own.
3 ODS provides a way for you to choose individual output objects to send to ODS
destinations. For example, PROC UNIVARIATE produces ve output objects. You
can easily create HTML output, an output data set, traditional listing output, or
printer output from any or all of these output objects. You can send different
output objects to different destinations.

Fundamental Concepts for Using Base SAS Procedures

What Are the ODS Destinations?

3 In the SAS windowing environment, ODS stores a link to each output object in the
Results folder in the Results window.

3 Because formatting is now centralized in ODS, the addition of a new ODS

destination does not affect any procedures or the DATA step. As future
destinations are added to ODS, they will automatically become available to the
DATA step and all procedures that support ODS.
3 With ODS, you can produce output for numerous destinations from a single source,
but you do not need to maintain separate sources for each destination. This
feature saves you time and system resources by enabling you to produce multiple
kinds of output with a single run of your procedure or data query.

What Are the ODS Destinations?

Overview of ODS Destination Categories
ODS enables you to produce SAS procedure and DATA step output to many different
destinations. ODS destinations are organized into two categories.
SAS Formatted
destinations

produce output that is controlled and interpreted by SAS, such as a

SAS data set, SAS output listing, or an ODS document.

Third-Party
Formatted
destinations

produce output which enables you to apply styles, markup

languages, or enables you to print to physical printers using page
description languages. For example, you can produce output in
PostScript, HTML, XML, or a style or markup language that you
created.

The following table lists the ODS destination categories, the destination that each
category includes, and the formatted output that results from each destination.
Table 2.4 Destination Category Table
Category

Destinations

Results

SAS Formatted

DOCUMENT

ODS document

LISTING

SAS output listing

OUTPUT

SAS data set

HTML

HTML le for online viewing

MARKUP

markup language tagsets

PRINTER

printable output in one of three

different formats: PCL, PDF,
or PS (PostScript)

RTF

output written in Rich Text

Format for use with Microsoft
Word 2000

Third-Party Formatted

As future destinations are added to ODS, they automatically will become available to
the DATA step and to all procedures that support ODS.

Denition of Destination-Independent Input

Destination-independent input means that one destination can support a feature
even though another destination does not support it. In this case, the request is ignored

What Are the ODS Destinations?

Chapter 2

by the destination that does not support it. Otherwise, ODS would support a small
subset of features that are only common to all destinations. If this was true, then it
would be difcult to move your reports from one output format to another output
format. ODS provides many output formatting options, so that you can use the
appropriate format for the output that you want. It is best to use the appropriate
destination suited for your purpose.

The SAS Formatted Destinations

The SAS formatted destinations create SAS entities such as a SAS data set, a SAS
output listing, or an ODS document. The statements in the ODS SAS Formatted
category create the SAS entities.
The three SAS formatted destinations are:
DOCUMENT Destination
The DOCUMENT destination enables you to restructure, navigate, and replay
your data in different ways and to different destinations as you like without
needing to rerun your analysis or repeat your database query. The DOCUMENT
destination makes your entire output stream available in "raw" form and
accessible to you to customize. The output is kept in the original internal
representation as a data component plus a table denition. When the output is in
a DOCUMENT form, it is possible to rearrange, restructure, and reformat without
rerunning your analysis. Unlike other ODS destinations, the DOCUMENT
destination has a GUI interface. However, everything that you can do through the
GUI, you can also do with batch commands using the ODS DOCUMENT
statement and the DOCUMENT procedure.
Prior to SAS 9, each procedure or DATA step produced output that was sent to
each destination that you specied. While you could always send your output to as
many destinations as you wanted, you needed to rerun your procedure or data
query if you decided to use a destination that you had not originally designated.
The DOCUMENT destination eliminates the need to rerun procedures or repeat
data queries by enabling you to store your output objects and replay them to
different destinations.
LISTING Destination
The LISTING destination produces output that looks the same as the traditional
SAS output. The LISTING destination is the default destination that opens when
you start your SAS session. Thus ODS is always being used, even when you do not
explicitly invoke ODS.
The LISTING destination enables you to produce traditional SAS output with
the same look and presentation as it had in previous versions of SAS.
Because most procedures share some of the same table denitions, the output is
more consistent. For example, if you have two different procedures producing an
ANOVA table, they will both produce it in the same way because each procedure
uses the same template to describe the table. However, there are four procedures
that do not use a default table denition to produce their output: PRINT
procedure, REPORT procedure, TABULATE procedure, and FREQ procedures
n-way tables. These procedures use the structure that you specied in your
program code to dene their tables.
OUTPUT Destination
The OUTPUT destination produces SAS output data sets. Because ODS already
knows the logical structure of the data and its native form, ODS can output a SAS
data set that represents exactly the same resulting data set that the procedure
worked with internally. The output data sets can be used for further analysis, or
for sophisticated reports in which you want to combine similar statistics across

Fundamental Concepts for Using Base SAS Procedures

What Are the ODS Destinations?

different data sets into a single table. You can easily access and process your
output data sets using all of the SAS data set features. For example, you can
access your output data using variable names and perform WHERE-expression
processing just as you would process data from any other SAS data set.

The Third-Party Formatted Destinations

The third-party formatted destinations enable you to apply styles to the output
objects that are used by applications other than SAS. For example, these destinations
support attributes such as "font" and "color."
Note: For a list of style elements and valid values, see the style elements table in
the SAS Output Delivery System: Users Guide. 4
The four categories of third-party formatted destinations are:

3 HTML (Hypertext Markup Language )

The HTML destination produces HTML 3.2-compatible output. You can,
however, produce (HTML 4 stylesheet) output using the HTML4 tagsets.
The HTML destination can create some or all of the following:

3 an HTML le (called the body le) that contains the results from the
procedure

3 a table of contents that links to the body le

3 a table of pages that links to the body le
3 a frame that displays the table of contents, the table of pages, and the body
le.
The body le is required with all ODS HTML output. If you do not want to link
to your output, then you do not have to create a table of contents, a table of pages,
or a frame le. However, if your output is very large, you might want to create a
table of contents and a table of pages for easier reading and transversing through
your le.
The HTML destination is intended only for on-line use, not for printing. To
print hard-copies of the output objects, use the PRINTER destination.

3 Markup Languages (MARKUP) Family

Just as table denitions describe how to lay out a table, and style attributes
describe the style of the output, tagsets describe how to produce a markup
language output. You can use a tagset that SAS supplies or you can create your
own using the TEMPLATE procedure. Like table denitions and style attributes,
tagsets enable you to modify your markup language output. For example, each
variety of XML can be specied as a new tagset. SAS supplies you with a
collection of XML tagsets and enables you to produce a customized variety of XML.
The important point is that you can implement a tagset that SAS supplies or a
customized tagset that you created without having to wait for the next release of
SAS. With the addition of modifying and creating your own tagsets by using PROC
TEMPLATE, now you have greater exibility in customizing your output.
Because the MARKUP destination is so exible, you can use either the SAS
tagsets or a tagset that you created. For a complete listing of the markup
language tagsets that SAS supplies, see the section on listing tagset names in the
SAS Output Delivery System: Users Guide. To learn how to dene your own
tagsets, see the section on methods to create your own tagsets in the SAS Output
Delivery System: Users Guide.
The MARKUP destination cannot replace ODS PRINTER or ODS RTF
destinations because it cannot do text measurement. Therefore, it cannot produce

What Are the ODS Destinations?

Chapter 2

output for a page description language or a hybrid language like RTF which
requires all of the text to be measured and placed at a specic position on the page.

3 PRINTER Family
The PRINTER destination produces output for

3 printing to physical printers such as Windows printers under Windows, PCL,

and PostScript printers on other operating systems

3 producing portable PostScript, PCL, and PDF les.

The PRINTER destinations produce ODS output that contain page description
languages: they describe precise positions where each line of text, each rule, and
each graphical element are to be placed on the page. In general, you cannot edit or
alter these formats. Therefore, the output from ODS PRINTER is intended to be
the nal form of the report.

3 Rich Text Format (RTF)

RTF produces output for Microsoft Word. While there are other applications
that can read RTF les, the RTF output might not work successfully with them.
The RTF destination enables you to view and edit the RTF output. ODS does
not dene the vertical measurement," meaning that SAS does not determine the
optimal place to position each item on the page. For example, page breaks are not
always xed, so when you edit your text, you do not want your RTF output tables
to split at inappropriate places. Your tables can remain whole and intact on one
page or can have logical breaks where you specied.
However, because Microsoft Word needs to know the widths of table columns
and it cannot adjust tables if they are too wide for the page, ODS measures the
width of the text and tables (horizontal measurement). Therefore, all the column
widths can be set properly by SAS and the table can be divided into panels if it is
too wide to t on a single page.
In short, when producing RTF output for input to Microsoft Word, SAS
determines the horizontal measurement and Microsoft Word controls the vertical
measurement. Because Microsoft Word can determine how much room there is on
the page, your tables will display consistently as you specied even after you
modied your RTF le.

What Controls the Formatting Features of Third-Party Formats?

All of the formatting features that control the appearance of the third-party
formatted destinations beyond what the LISTING destination can do are controlled by
two mechanisms:

3 ODS statement options

3 ODS style attributes
The ODS statement options control three features:
1 Features that are specic to a given destination, such as stylesheets for HTML.
2 Features that are global to the document, such as AUTHOR and table of contents

generation.
3 Features that we expect users to change on each document, such as the output le

name.
The ODS style attributes control the way that individual elements are created.
Attributes are aspects of a given style, such as type face, weight, font size, and color.
The values of the attributes collectively determine the appearance of each part of the
document to which the style is applied. With style attributes, it is unnecessary to insert

Fundamental Concepts for Using Base SAS Procedures

What Are Style Denitions, Style Elements, and Style Attributes?

destination-specic code (such as raw HTML) into the document. Each output
destination will interpret the attributes that are necessary to generate the presentation
of the document. Because not all destinations are the same, not all attributes can be
interpreted by all destinations. Style attributes that are incompatible with a selected
destination are ignored. For example, PostScript does not support active links, so the
URL= attribute is ignored when producing PostScript output.

ODS Destinations and System Resources

ODS destinations can be open or closed. You open and close a destination with the
appropriate ODS statement. When a destination is open, ODS sends the output objects
to it. An open destination uses system resources even if you use the selection and
exclusion features of ODS to select or exclude all objects from the destination.
Therefore, to conserve resources, close unnecessary destinations. For more information
about using each destination, see the topic on ODS statements in the SAS Output
Delivery System: Users Guide.
By default, the LISTING destination is open and all other destinations are closed.
Consequently, if you do nothing, your SAS programs run and produce listing output
looking just as they did in previous releases of SAS before ODS was available.

What Are Table Denitions, Table Elements, and Table Attributes?

A table denition describes how to generate the output for a tabular output object.
(Most ODS output is tabular.) A table denition determines the order of column
headers and the order of variables, as well the overall look of the output object that
uses it. For information about customizing the table denition, see the topic on the
TEMPLATE procedure in the SAS Output Delivery System: Users Guide.
In addition to the parts of the table denition that order the headers and columns,
each table denition contains or references table elements. A table element is a
collection of table attributes that apply to a particular header, footer, or column.
Typically, a table attribute species something about the data rather than about its
presentation. For example, FORMAT species the SAS format, such as the number of
decimal places. However, some table attributes describe presentation aspects of the
data, such as how many blank characters to place between columns.
Note: The attributes of table denitions that control the presentation of the data
have no effect on output objects that go to the LISTING or OUTPUT destination.
However, the attributes that control the structure of the table and the data values do
affect listing output. 4
For information on table attributes, see the section on table attributes in the SAS
Output Delivery System: Users Guide.

What Are Style Denitions, Style Elements, and Style Attributes?

To customize the output at the level of your entire output stream in a SAS session,
you specify a style denition. A style denition describes how to generate the
presentation aspects (color, font face, font size, and so on) of the entire SAS output. A
style denition determines the overall look of the documents that use it.
Each style denition is composed of style elements. A style element is a collection of
style attributes that apply to a particular part of the output. For example, a style
element may contain instructions for the presentation of column headers, or for the
presentation of the data inside the cells. Style elements may also specify default colors
and fonts for output that uses the style denition.

What Are Style Denitions, Style Elements, and Style Attributes?

Chapter 2

Each style attribute species a value for one aspect of the presentation. For example,
the BACKGROUND= attribute species the color for the background of an HTML table
or for a colored table in printed output. The FONT_STYLE= attribute species whether
to use a Roman or an italic font. For information on style attributes, see the section on
style attributes in the SAS Output Delivery System: Users Guide.
Note: Because style denitions control the presentation of the data, they have no
effect on output objects that go to the LISTING or OUTPUT destination. 4

What Style Denitions Are Shipped with SAS Software?

Base SAS software is shipped with many style denitions. To see a list of these
styles, you can view them in the SAS Explorer Window, use the TEMPLATE procedure,
or use the SQL procedure.
3 SAS Explorer Window:
To display a list of the available styles using the SAS Explorer Window, follow
these steps:
1 From any window in an interactive SAS session, select
View

Results

2 In the Results window, select

View

Templates

3 In the Templates window, select and open Sashelp.tmplmst.

4 Select and open the Styles folder, which contains a list of available style

denitions. If you want to view the underlying SAS code for a style
denition, then select the style and open it.
Operating Environment Information: For information on navigating in the
Explorer window without a mouse, see the section on Window Controls and
General Navigation in the SAS documentation for your operating
environment. 4

3 TEMPLATE Procedure:
You can also display a list of the available styles by submitting the following
PROC TEMPLATE statements:
proc template;
list styles;
run;

3 SQL Procedure:
You can also display a list of the available styles by submitting the following
PROC SQL statements:
proc sql;
select * from styles.style-name;

The stylename is the name of any style from the template store (for example,
styles.default or styles.beige).

For more information on how ODS destinations use styles and how you can
customize styles, see the section on the DEFINE STYLE statement in the SAS Output
Delivery System: Users Guide.

How Do I Use Style Denitions with Base SAS Procedures?

3 Most Base SAS Procedures

Fundamental Concepts for Using Base SAS Procedures

Changing SAS Registry Settings for ODS

Most Base SAS procedures that support ODS use one or more table denitions
to produce output objects. These table denitions include denitions for table
elements: columns, headers, and footers. Each table element can specify the use of
one or more style elements for various parts of the output. These style elements
cannot be specied within the syntax of the procedure, but you can use customized
styles for the ODS destinations that you use. For more information about
customizing tables and styles, see the TEMPLATE procedure in the SAS Output
Delivery System: Users Guide.

3 The PRINT, REPORT and TABULATE Procedures

The PRINT, REPORT and TABULATE procedures provide a way for you to
access table elements from the procedure step itself. Accessing the table elements
enables you to do such things as specify background colors for specic cells, change
the font face for column headers, and more. The PRINT, REPORT, and
TABULATE procedures provide a way for you to customize the markup language
and printed output directly from the procedure statements that create the report.
For more information about customizing the styles for these procedures, see the
Base SAS Procedures Guide.

Changing SAS Registry Settings for ODS

Overview of ODS and the SAS Registry
The SAS registry is the central storage area for conguration data that ODS uses.
This conguration data is stored in a hierarchical form, which works in a similar
manner to the way directory-based le structures work under UNIX, Windows, VMS,
and the z/OS UNIX system. However, the SAS registry uses keys and subkeys as the
basis for its structure, instead of using directories and subdirectories, like similar le
systems in DOS or UNIX. A key is a word or a text string that refers to a particular
aspect of SAS. Each key may be a place holder without values or subkeys associated
with it, or it may have many subkeys with associated values. For example, the ODS key
has DESTINATIONS, GUI, ICONS, and PREFERENCES subkeys. A subkey is a key
inside another key. For example, PRINTER is a subkey of the DESTINATIONS subkey.
Display 2.5 SAS Registry of ODS Subkeys

Changing SAS Registry Settings for ODS

Chapter 2

Changing Your Default HTML Version Setting

By default, the SAS registry is congured to generate HTML4 output when you
specify the ODS HTML statement. To permanently change the default HTML version,
you can change the setting of the HTML version in the SAS registry.
CAUTION:

If you make a mistake when you modify the SAS registry, then your system might become
unstable or unusable. You will not be warned if an entry is incorrect. Incorrect entries
can cause errors, and can even prevent you from bringing up a SAS session. For
more information about how to congure the SAS registry, see the SAS registry
section in SAS Language Reference: Concepts. 4
To change the default setting of the HTML version in the SAS registry:
1 Select

Solutions

Accessories

Registry Editor

or
Issue the command REGEDIT.
2 Select

ODS

Default HMTL Version

Modify

3 Select

Edit

or
Click the right mouse button and select MODIFY. The Edit String Value window
appears.
4 Type the HTML version in the Value Data text box and select OK.
Display 2.6 SAS Registry Showing HTML Version Setting

Fundamental Concepts for Using Base SAS Procedures

Customized ODS Output

Changing ODS Destination Default Settings

ODS destination subkeys are stored in the SAS registry. To change the values for
these destinations subkeys:
1 Select

ODS

Destinations

2 Select a destination subkey

3 Select a subkey in the Contents of window
4 Select

Edit

Modify

or
Click the right mouse button and select MODIFY.
5 Type in the Value Data entry into the Edit Value String or Edit Signed Integer

Value window and select OK.

Display 2.7 Registry Editor Window

Customized ODS Output

SAS Output
By default, ODS output is formatted according to instructions that a PROC step or
DATA step denes. However, ODS provides ways for you to customize the output. You
can customize the output for an entire SAS job, or you can customize the output for a
single output object.

Customized ODS Output

Chapter 2

Selection and Exclusion Lists

You can specify which output objects that you want to produce by selecting or
excluding them in a list. For each ODS destination, ODS maintains either a selection
list or an exclusion list. A selection list is a list of output objects that are sent to the
destination. An exclusion list is a list of output objects that are excluded from the
destination. ODS also maintains an overall selection list or an overall exclusion list. You
can use these lists to control which output objects go to the specied ODS destinations.
To see the contents of the lists use the ODS SHOW statement. The lists are written
to the SAS log. The following table shows the default lists:

Table 2.5 Default List for Each ODS Destination

ODS Destination

Default List

OUTPUT

EXCLUDE ALL

All others

SELECT ALL

How Does ODS Determine the Destinations for an Output Object?

To specify an output object, you need to know which output objects your SAS
program produces. The ODS TRACE statement writes to the SAS log a trace record
that includes the path, the label, and other information about each output object that is
produced. For more information, about the ODS TRACE statement see SAS Output
Delivery System: Users Guide. You can specify an output object as any of the following:

3 a full path. For example,

Univariate.City_Pop_90.TestsForLocation

is the full path of the output object.

3 a partial path. A partial path consists of any part of the full path that begins
immediately after a period (.) and continues to the end of the full path. For
example, if the full path is
Univariate.City_Pop_90.TestsForLocation

then the partial paths are:

City_Pop_90.TestsForLocation
TestsForLocation

3 a label that is enclosed in quotation marks.

For example,
"Tests For Location"

3 a label path. For example, the label path for the output object is
"The UNIVARIATE Procedure"."CityPop_90"."Tests For Location"
Note: The trace record shows the label path only if you specify the LABEL
option in the ODS TRACE statement. 4

3 a partial label path. A partial label path consists of any part of the label that
begins immediately after a period (.) and continues to the end of the label. For
example, if the label path is
"The UNIVARIATE Procedure"."CityPop_90"."Tests For Location"

Fundamental Concepts for Using Base SAS Procedures

Customized ODS Output

then the partial label paths are:

"CityPop_90"."Tests For Location"
"Tests For Location"

3 a mixture of labels and paths.

3 any of the partial path specications, followed by a pound sign (#) and a number.
For example, TestsForLocation#3 refers to the third output object that is named
TestsForLocation.
As each output object is produced, ODS uses the selection and exclusion lists to
determine which destination or destinations the output object will be sent to. The
following gure illustrates this process:

Figure 2.2 Directing an Output Object to a Destination

For each destination, ODS rst asks if the list for that destination includes the object. If it does
not, ODS does not send the output object to that destination. If the list for that destination does
include the object, ODS reads the overall list. If the overall list includes the object, ODS sends
it to the destination. If the overall list does not include the object, ODS does not send it to the
destination.

Does the destination list

include the output object
to the destination?
yes
no

Does the overall list

include the object ?
no
yes
ODS passes the object
to the destination

ODS doesn't pass the

object to the destination

Note: Although you can maintain a selection list for one destination and an
exclusion list for another, it is easier to understand the results if you maintain the same
types of lists for all the destinations where you route output. 4

Customized Output for an Output Object

For a procedure, the name of the table denition that is used for an output object
comes from the procedure code. The DATA step uses a default table denition unless
you specify an alternative with the TEMPLATE= suboption in the ODS option in the
FILE statement. For more information, see the section on the TEMPLATE= suboption
in the SAS Output Delivery System: Users Guide.
To nd out which table denitions a procedure or the DATA step uses for the output
objects, you must look at a trace record. To produce a trace record in your SAS log,
submit the following SAS statements:

Summary of ODS

Chapter 2

ods trace on;

your-proc-or-DATA-step
ods trace off;

Remember that not all procedures use table denitions. If you produce a trace record
for one of these procedures, no denition appears in the trace record. Conversely, some
procedures use multiple table denitions to produce their output. If you produce a trace
record for one of these procedures, more than one denition appears in the trace record.
The trace record refers to the table denition as a template. For a detailed
explanation of the trace record, see the section on the ODS TRACE statement in the
SAS Output Delivery System: Users Guide.
You can use PROC TEMPLATE to modify an entire table denition. When a
procedure or DATA step uses a table denition, it uses the elements that are dened or
referenced in its table denition. In general, you cannot directly specify a table element
for your procedure or DATA step to use without modifying the denition itself.
Note: Three Base SAS procedures, PROC PRINT, PROC REPORT and PROC
TABULATE, do provide a way for you to access table elements from the procedure step
itself. Accessing the table elements enables you to customize your report. For more
information about these procedures, see the Base SAS Procedures Guide 4

Summary of ODS
In the past, the term output has generally referred to the outcome of a SAS
procedure and DATA step. With the advent of the Output Delivery System, output
takes on a much broader meaning. ODS is designed to optimize output from SAS
procedures and the DATA step. It provides a wide range of formatting options and
greater exibility in generating, storing, and reproducing SAS output.
Important features of ODS include the following:

3 ODS combines raw data with one or more table denitions to produce one or more
output objects. An output object tells ODS how to format the results of a procedure
or DATA step.

3 ODS provides a way for you to choose individual output objects to send to ODS
destinations.

3 ODS stores a link to each output object in the Results folder for easy retrieval and
access.

3 As future destinations are added to ODS, they will automatically become available
to the DATA step and all procedures that support ODS.
One of the main goals of ODS is to enable you to produce output for numerous
destinations from a single source, without requiring separate sources for each
destination. ODS supports many destinations:
DOCUMENT
enables you to capture output objects from single run of the analysis and produce
multiple reports in various formats whenever you want without re-running your
SAS programs.
LISTING
produces output that looks the same as the traditional SAS output.

Fundamental Concepts for Using Base SAS Procedures

Summary of ODS

HTML
produces output for online viewing.
MARKUP
produces output for markup language tagsets.
OUTPUT
produces SAS output data sets, thereby eliminating the need to parse PROC
PRINTTO output.
PRINTER
produces presentation-ready printed reports.
RTF
produces output suitable for Microsoft Word reports.
By default, ODS output is formatted according to instructions that the procedure or
DATA step denes. However, ODS provides ways for you to customize the presentation
of your output. You can customize the presentation of your SAS output, or you can
customize the look of a single output object. ODS gives you greater exibility in
generating, storing, and reproducing SAS procedure and DATA step output with a wide
range of formatting options.

CHAPTER

3
Statements with the Same
Function in Multiple Procedures
Overview 57
Statements 58
BY 58
FREQ 61
QUIT 63
WEIGHT 63
WHERE 68

Overview
Several statements are available and have the same function in a number of base
SAS procedures. Some of the statements are fully documented in SAS Language
Reference: Dictionary, and others are documented in this section. The following list
shows you where to nd more information about each statement:
ATTRIB
affects the procedure output and the output data set. The ATTRIB statement does
not permanently alter the variables in the input data set. The LENGTH= option
has no effect. See SAS Language Reference: Dictionary for complete
documentation.
BY
orders the output according to the BY groups. See BY on page 58.
FORMAT
affects the procedure output and the output data set. The FORMAT statement does
not permanently alter the variables in the input data set. The DEFAULT= option
is not valid. See SAS Language Reference: Dictionary for complete documentation.
FREQ
treats observations as if they appear multiple times in the input data set. See
FREQ on page 61.
LABEL
affects the procedure output and the output data set. The LABEL statement does
not permanently alter the variables in the input data set except when it is used
with the MODIFY statement in PROC DATASETS. See SAS Language Reference:
Dictionary for complete documentation.
QUIT
executes any statements that have not executed and ends the procedure. See
QUIT on page 63.

Statements

Chapter 3

WEIGHT
species weights for analysis variables in the statistical calculations. See
WEIGHT on page 63.
WHERE
subsets the input data set by specifying certain conditions that each observation
must meet before it is available for processing. See WHERE on page 68.

Statements

BY
Orders the output according to the BY groups.
See also: Creating Titles That Contain BY-Group Information on page 20

BY < DESCENDING> variable-1

< <DESCENDING> variable-n>
<NOTSORTED>;

Required Arguments
variable

species the variable that the procedure uses to form BY groups. You can specify
more than one variable. If you do not use the NOTSORTED option in the BY
statement, then the observations in the data set must either be sorted by all the
variables that you specify, or they must be indexed appropriately. Variables in a BY
statement are called BY variables.

Options
DESCENDING

species that the observations are sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The observations are grouped in another way, for example, chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations with the same values for the BY variables
are not contiguous, then the procedure treats each contiguous set as a separate BY
group.
Note:

You cannot use the NOTSORTED option in a PROC SORT step.

Statements with the Same Function in Multiple Procedures

Note: You cannot use the GROUPFORMAT option, which is available in the BY
statement in a DATA step, in a BY statement in any PROC step. 4

BY-Group Processing
Procedures create output for each BY group. For example, the elementary statistics
procedures and the scoring procedures perform separate analyses for each BY group.
The reporting procedures produce a report for each BY group.
Note: All base SAS procedures except PROC PRINT process BY groups
independently. PROC PRINT can report the number of observations in each BY group
as well as the number of observations in all BY groups. Similarly, PROC PRINT can
sum numeric variables in each BY group and across all BY groups. 4
You can use only one BY statement in each PROC step. When you use a BY
statement, the procedure expects an input data set that is sorted by the order of the BY
variables or one that has an appropriate index. If your input data set does not meet
these criteria, then an error occurs. Either sort it with the SORT procedure or create an
appropriate index on the BY variables.
Depending on the order of your data, you may need to use the NOTSORTED or
DESCENDING option in the BY statement in the PROC step.
For more information on

3 the BY statement, see SAS Language Reference: Dictionary.

3 PROC SORT, see Chapter 43, The SORT Procedure, on page 1003.
3 creating indexes, see INDEX CREATE Statement on page 345.

Formatting BY-Variable Values

When a procedure is submitted with a BY statement, the following actions are taken
with respect to processing of BY groups:
1 The procedure determines whether the data is sorted by the internal

(unformatted) values of the BY variable(s).

2 The procedure determines whether a format has been applied to the BY

variable(s). If the BY variable is numeric and has no user-applied format, then the
BEST12. format is applied for the purpose of BY-group processing.
3 The procedure continues adding observations to the current BY group until both

the internal and the formatted values of the BY variable(s) change.

This process can have unexpected results if, for instance, nonconsecutive internal BY
values share the same formatted value. In this case, the formatted value is represented
in different BY groups. Alternatively, if different consecutive internal BY values share
the same formatted value, then these observations are grouped into the same BY group.

Chapter 3

Base SAS Procedures That Support the BY Statement

CALENDAR

REPORT (nonwindowing environment only)

CHART

SORT (required)

COMPARE

STANDARD

CORR

SUMMARY

FREQ

TABULATE

MEANS

TIMEPLOT

PLOT

TRANSPOSE

UNIVARIATE

RANK

Note: In the SORT procedure, the BY statement species how to sort the data. With
the other procedures, the BY statement species how the data are currently sorted. 4

Example
This example uses a BY statement in a PROC PRINT step. There is output for each
value of the BY variable, Year. The DEBATE data set is created in Example:
Temporarily Dissociating a Format from a Variable on page 29.
options nodate pageno=1 linesize=64
pagesize=40;
proc print data=debate noobs;
by year;
title Printing of Team Members;
title2 by Year;
run;

Statements with the Same Function in Multiple Procedures

Printing of Team Members

by Year

FREQ

------------------------ Year=Freshman ------------------------Name

Gender

Capiccio
Tucker

m
m

GPA
3.598
3.901

------------------------ Year=Sophomore -----------------------Name

Gender

GPA

f
m
m

3.722
3.198
3.342

Bagwell
Berry
Metcalf

------------------------- Year=Junior -------------------------Name

Gold
Gray
Syme

Gender
f
f
f

GPA
3.609
3.177
3.883

------------------------- Year=Senior -------------------------Name

Baglione
Carr
Hall
Lewis

Gender
f
m
m
m

GPA
4.000
3.750
3.574
3.421

FREQ
Treats observations as if they appear multiple times in the input data set.
Tip: You can use a WEIGHT statement and a FREQ statement in the same step of any
procedure that supports both statements.

FREQ variable;

Required Arguments
variable

species a numeric variable whose value represents the frequency of the observation.
If you use the FREQ statement, then the procedure assumes that each observation

FREQ

Chapter 3

represents n observations, where n is the value of variable. If variable is not an

integer, then SAS truncates it. If variable is less than 1 or is missing, then the
procedure does not use that observation to calculate statistics. If a FREQ statement
does not appear, then each observation has a default frequency of 1.
The sum of the frequency variable represents the total number of observations.

Procedures That Support the FREQ Statement

3
3
3
3
3
3

CORR
MEANS/SUMMARY
REPORT
STANDARD
TABULATE
UNIVARIATE

Example
The data in this example represent a ships course and speed (in nautical miles per
hour), recorded every hour. The frequency variable, Hours, represents the number of
hours that the ship maintained the same course and speed. Each of the following PROC
MEANS steps calculates average course and speed. The different results demonstrate
the effect of using Hours as a frequency variable.
The following PROC MEANS step does not use a frequency variable:
options nodate pageno=1 linesize=64 pagesize=40;
data track;
input Course Speed Hours @@;
datalines;
30 4 8 50 7 20
75 10 30 30 8 10
80 9 22 20 8 25
83 11 6 20 6 20
;

proc means data=track maxdec=2 n mean;

var course speed;
title Average Course and Speed;
run;

Without a frequency variable, each observation has a frequency of 1, and the total
number of observations is 8.

Average Course and Speed

The MEANS Procedure
Variable
N
Mean
----------------------------Course
8
48.50
Speed
8
7.88
-----------------------------

Statements with the Same Function in Multiple Procedures

WEIGHT

The second PROC MEANS step uses Hours as a frequency variable:

proc means data=track maxdec=2 n mean;
var course speed;
freq hours;
title Average Course and Speed;
run;

When you use Hours as a frequency variable, the frequency of each observation is the
value of Hours, and the total number of observations is 141 (the sum of the values of
the frequency variable).

Average Course and Speed

The MEANS Procedure

Variable
N
Mean
---------------------------------------Course
141
49.28
Speed
141
8.06
----------------------------------------

QUIT
Executes any statements that have not executed and ends the procedure.
QUIT;

Procedures That Support the QUIT Statement

3
3
3
3
3

CATALOG
DATASETS
PLOT
PMENU
SQL

WEIGHT
Species weights for analysis variables in the statistical calculations.
Tip: You can use a WEIGHT statement and a FREQ statement in the same step of any
procedure that supports both statements.

WEIGHT variable;

WEIGHT

Chapter 3

Required Arguments
variable

species a numeric variable whose values weight the values of the analysis variables.
The values of the variable do not have to be integers. The behavior of the procedure
when it encounters a nonpositive weight variable value is as follows:
Weight value
0
less than 0
missing

The procedure
counts the observation in the total number of observations
converts the weight value to zero and counts the observation in
the total number of observations
excludes the observation from the analysis

Different behavior for nonpositive values is discussed in the WEIGHT statement

syntax under the individual procedure.
Prior to Version 7 of SAS, no base SAS procedure excluded the observations with
missing weights from the analysis. Most SAS/STAT procedures, such as PROC GLM,
have always excluded not only missing weights but also negative and zero weights
from the analysis. You can achieve this same behavior in a base SAS procedure that
supports the WEIGHT statement by using the EXCLNPWGT option in the PROC
statement.
The procedure substitutes the value of the WEIGHT variable for i , which
appears in Keywords and Formulas on page 1340.

Procedures That Support the WEIGHT Statement

3
3
3
3
3
3
3

CORR
FREQ
MEANS/SUMMARY
REPORT
STANDARD
TABULATE
UNIVARIATE

Note: In PROC FREQ, the value of the variable in the WEIGHT statement
represents the frequency of occurrence for each observation. See the PROC FREQ
documentation in Volume 3 of this book for more information. 4

Calculating Weighted Statistics

The procedures that support the WEIGHT statement also support the VARDEF=
option, which lets you specify a divisor to use in the calculation of the variance and
standard deviation.
By using a WEIGHT statement to compute moments, you assume that the ith
observation has a variance that is equal to 2 i . When you specify VARDEF=DF (the
default), the computed variance is a weighted least squares estimate of 2 . Similarly,
the computed standard deviation is an estimate of . Note that the computed variance

Statements with the Same Function in Multiple Procedures

WEIGHT

is not an estimate of the variance of the ith observation, because this variance involves
the observations weight which varies from observation to observation.
If the values of your variable are counts that represent the number of occurrences of
each observation, then use this variable in the FREQ statement rather than in the
WEIGHT statement. In this case, because the values are counts, they should be
integers. (The FREQ statement truncates any noninteger values.) The variance that is
computed with a FREQ variable is an estimate of the common variance, 2 , of the
observations.
Note: If your data come from a stratied sample where the weights wi represent
the strata weights, then neither the WEIGHT statement nor the FREQ statement
provides appropriate stratied estimates of the mean, variance, or variance of the
mean. To perform the appropriate analysis, consider using PROC SURVEYMEANS,
which is a SAS/STAT procedure that is documented in the SAS/STAT Users Guide. 4

Weighted Statistics Example

As an example of the WEIGHT statement, suppose 20 people are asked to estimate
the size of an object 30 cm wide. Each person is placed at a different distance from the
object. As the distance from the object increases, the estimates should become less
precise.
The SAS data set SIZE contains the estimate (ObjectSize) in centimeters at each
distance (Distance) in meters and the precision (Precision) for each estimate. Notice
that the largest deviation (an overestimate by 20 cm) came at the greatest distance (7.5
meters from the object). As a measure of precision, 1/Distance, gives more weight to
estimates that were made closer to the object and less weight to estimates that were
made at greater distances.
The following statements create the data set SIZE:
options nodate pageno=1 linesize=64 pagesize=60;
data size;
input Distance ObjectSize @@;
Precision=1/distance;
datalines;
1.5 30 1.5 20 1.5 30 1.5 25
3
43 3
33 3
25 3
30
4.5 25 4.5 36 4.5 48 4.5 33
6
43 6
36 6
23 6
48
7.5 30 7.5 25 7.5 50 7.5 38
;

The following PROC MEANS step computes the average estimate of the object size
while ignoring the weights. Without a WEIGHT variable, PROC MEANS uses the
default weight of 1 for every observation. Thus, the estimates of object size at all
distances are given equal weight. The average estimate of the object size exceeds the
actual size by 3.55 cm.
proc means data=size maxdec=3 n mean var stddev;
var objectsize;
title1 Unweighted Analysis of the SIZE Data Set;
run;

WEIGHT

Chapter 3

Unweighted Analysis of the SIZE Data Set

The MEANS Procedure

Analysis Variable : ObjectSize
N
Mean
Variance
Std Dev
-------------------------------------------------20
33.550
80.892
8.994
--------------------------------------------------

The next two PROC MEANS steps use the precision measure (Precision) in the
WEIGHT statement and show the effect of using different values of the VARDEF=
option. The rst PROC step creates an output data set that contains the variance and
standard deviation. If you reduce the weighting of the estimates that are made at
greater distances, the weighted average estimate of the object size is closer to the actual
size.
proc means data=size maxdec=3 n mean var stddev;
weight precision;
var objectsize;
output out=wtstats var=Est_SigmaSq std=Est_Sigma;
title1 Weighted Analysis Using Default VARDEF=DF;
run;
proc means data=size maxdec=3 n mean var std
vardef=weight;
weight precision;
var objectsize;
title1 Weighted Analysis Using VARDEF=WEIGHT;
run;

In the rst PROC MEANS step, the variance is an estimate of 2 , where the
variance of the ith observation is assumed to be var (xi ) = 2 =wi and wi is the weight
for the ith observation. In the second PROC MEANS step, the computed variance is an
estimate of (n 1=n) 2 =w, where w is the average weight. For large n, this is an
approximate estimate of the variance of an observation with average weight.

Weighted Analysis Using Default VARDEF=DF

The MEANS Procedure
Analysis Variable : ObjectSize
N
Mean
Variance
Std Dev
-------------------------------------------------20
31.088
20.678
4.547
--------------------------------------------------

Statements with the Same Function in Multiple Procedures

Weighted Analysis Using VARDEF=WEIGHT

Est_
Sigma

Est_
VarObs

Est_
StdObs

20.678
20.678
20.678
20.678

4.547
4.547
4.547
4.547

155.09
155.09
155.09
155.09

12.453
12.453
12.453
12.453

WHERE
Subsets the input data set by specifying certain conditions that each observation must meet before
it is available for processing.
WHERE where-expression;

Statements with the Same Function in Multiple Procedures

WHERE

Required Arguments
where-expression

is a valid arithmetic or logical expression that generally consists of a sequence of

operands and operators. See SAS Language Reference: Dictionary for more
information on where processing.

Procedures That Support the WHERE Statement

You can use the WHERE statement with any of the following base SAS procedures
that read a SAS data set:
CALENDAR

RANK

CHART

REPORT

COMPARE

SORT

CORR

SQL

DATASETS (APPEND statement)

STANDARD

FREQ

TABULATE

MEANS/SUMMARY

TIMEPLOT

PLOT

TRANSPOSE

UNIVARIATE

Details
3 The CALENDAR and COMPARE procedures and the APPEND statement in
PROC DATASETS accept more than one input data set. See the documentation for
the specic procedure for more information.
3 To subset the output data set, use the WHERE= data set option:
proc report data=debate nowd
out=onlyfr(where=(year=1));
run;

For more information on WHERE=, see SAS Language Reference: Dictionary.

Example
In this example, PROC PRINT prints only those observations that meet the condition
of the WHERE expression. The DEBATE data set is created in Example: Temporarily
Dissociating a Format from a Variable on page 29.
options nodate pageno=1 linesize=64
pagesize=40;

proc print data=debate noobs;

where gpa>3.5;
title Team Members with a GPA;
title2 Greater than 3.5;

WHERE

Chapter 3

run;

Team Members with a GPA

Greater than 3.5
Name
Capiccio
Tucker
Bagwell
Gold
Syme
Baglione
Carr
Hall

Gender
m
m
f
f
f
f
m
m

Year
Freshman
Freshman
Sophomore
Junior
Junior
Senior
Senior
Senior

GPA
3.598
3.901
3.722
3.609
3.883
4.000
3.750
3.574

P A R T

9 . . . . . . . . . . The COMPARE Procedure

225

Chapter

10. . . . . . . . .The CONTENTS Procedure

275

Chapter

11. . . . . . . . .The COPY Procedure

277

Chapter

12. . . . . . . . .The CORR Procedure

283

589

Chapter

29. . . . . . . . .The OPTIONS Procedure

591

Chapter

30. . . . . . . . .The OPTLOAD Procedure

601

Chapter

31. . . . . . . . .The OPTSAVE Procedure

603

789

Chapter

38. . . . . . . . .The PRTEXP Procedure

803

403
419
429
485

487
489
501
523

605
665
703
771
787

807

813
831
845

1003
1027

Chapter

51. . . . . . . . .The TRANTAB Procedure

Chapter

52. . . . . . . . .The UNIVARIATE Procedure

1311

1333
1335

CHAPTER

4
The APPEND Procedure
Overview: APPEND Procedure 75
Syntax: APPEND Procedure 75

Overview: APPEND Procedure

The APPEND procedure adds the observations from one SAS data set to the end of
another SAS data set.
Generally, the APPEND procedure functions the same as the APPEND statement in
the DATASETS procedure. The only difference between the APPEND procedure and
the APPEND statement in PROC DATASETS is the default for libref in the BASE= and
DATA= arguments. For PROC APPEND, the default is either WORK or USER. For the
APPEND statement, the default is the libref of the procedure input library.

Syntax: APPEND Procedure

Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See

Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.
Reminder: You can use data set options with the BASE= and DATA= options. See Data

Set Options on page 18 for a list.

Reminder: Complete documentation for the APPEND statement and the APPEND

procedure is in APPEND Statement on page 313 .

PROC APPEND BASE=< libref.>SAS-data-set <DATA=< libref.>SAS-data-set>

<FORCE> <APPENDVER=V6>;

CHAPTER

5
The CALENDAR Procedure
Overview: CALENDAR Procedure 79
What Does the CALENDAR Procedure Do? 79
What Types of Calendars Can PROC CALENDAR Produce? 79
Advanced Scheduling and Project Management Tasks 83
Syntax: CALENDAR Procedure 84
PROC CALENDAR Statement 85
BY Statement 91
CALID Statement 92
DUR Statement 93
FIN Statement 94
HOLIDUR Statement 95
HOLIFIN Statement 95
HOLISTART Statement 96
HOLIVAR Statement 97
MEAN Statement 97
OUTDUR Statement 98
OUTFIN Statement 98
OUTSTART Statement 99
START Statement 99
SUM Statement 100
VAR Statement 101
Concepts: CALENDAR Procedure 101
Type of Calendars 101
Schedule Calendar 102
Denition 102
Required Statements 102
Examples 102
Summary Calendar 102
Denition 102
Required Statements 103
Multiple Events on a Single Day 103
Examples 103
The Default Calendars 103
Description 103
When You Unexpectedly Produce a Default Calendar 103
Examples 104
Calendars and Multiple Calendars 104
Denitions 104
Why Create Multiple Calendars 104
How to Identify Multiple Calendars 104
Using Holidays or Calendar Data Sets with Multiple Calendars

105

Contents

Chapter 5

Types of Reports That Contain Multiple Calendars 105

How to Identify Calendars with the CALID Statement and the Special Variable _CAL_
When You Use Holidays or Calendar Data Sets 106
Examples 106
Input Data Sets 106
Activities Data Set 107
Purpose 107
Requirements and Restrictions 107
Structure 107
Multiple Activities per Day in Summary Calendars 107
Examples 108
Holidays Data Set 108
Purpose 108
Structure 108
No Sorting Needed 108
Using SAS Date Versus SAS Datetime Values 108
Create a Generic Holidays Data Set 109
Holidays and Nonwork Periods 109
Examples 109
Calendar Data Set 109
Purpose 109
Structure 109
Using Default Workshifts Instead of a Workdays Data Set 110
Examples 110
Workdays Data Set 110
Purpose 110
Use Default Work Shifts or Create Your Own? 110
Structure 111
How Missing Values Are Treated 111
Examples 111
Missing Values in Input Data Sets 111
Results: CALENDAR Procedure 112
What Affects the Quantity of PROC CALENDAR Output 112
How Size Affects the Format of PROC CALENDAR Output 112
What Affects the Lines That Show Activity Duration 113
Customizing the Calendar Appearance 113
Portability of ODS Output with PROC CALENDAR 113
Examples: CALENDAR Procedure 114
Example 1: Schedule Calendar with Holidays: 5-Day Default 114
Example 2: Schedule Calendar Containing Multiple Calendars 118
Example 3: Multiple Schedule Calendars with Atypical Workshifts (Separated Output) 122
Example 4: Multiple Schedule Calendars with Atypical Workshifts (Combined and Mixed
Output) 127
Example 5: Schedule Calendar, Blank or with Holidays 134
Example 6: Calculating a Schedule Based on Completion of Predecessor Tasks 137
Example 7: Summary Calendar with MEAN Values By Observation 143
Example 8: Multiple Summary Calendars with Atypical Workshifts (Separated Output) 147

106

The CALENDAR Procedure

What Types of Calendars Can PROC CALENDAR Produce?

Overview: CALENDAR Procedure

What Does the CALENDAR Procedure Do?
The CALENDAR procedure displays data from a SAS data set in a monthly calendar
format. You can produce a schedule calendar, which schedules events around holidays
and nonwork periods, or you can produce a summary calendar, which summarizes data
and displays only one-day events and holidays. When you use PROC CALENDAR you
can

3 schedule work around holidays and other nonwork periods

3 display holidays
3 process data about multiple calendars in a single step and print them in a
separate, mixed, or combined format
3 apply different holidays, weekly work schedules, and daily work shifts to multiple
calendars in a single PROC step
3 produce a mean and a sum for variables based on either the number of days in a
month or the number of observations.
PROC CALENDAR also contains features that are specically designed to work with
PROC CPM in SAS/OR software, a project management scheduling tool.

What Types of Calendars Can PROC CALENDAR Produce?

Simple Schedule Calendar

Output 5.1 illustrates the simplest kind of schedule calendar that you can produce.
This calendar output displays activities that are planned by a banking executive. The
following statements produce Output 5.1.
options nodate pageno=1 linesize=132 pagesize=60;
proc calendar data=allacty;
start date;
dur long;
run;

For the activities data set shown that is in this calendar, see Example 1 on page 114.

What Types of Calendars Can PROC CALENDAR Produce?

Chapter 5

Output 5.1 Simple Schedule Calendar

This calendar uses one of the two default calendars, the 24-hour-day, 7-day-week calendar.

The SAS System

------------------------------------------------------------------------------------------------------------------------------|
|
|
July 1996
|
|
|
|-----------------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
|
1
|
2
|
3
|
4
|
5
|
6
|
|
|
|

|
|
|

|
|
|

|
|
|

|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
7
|
8
|
9
|
10
|
11
|
12
|
13
|
|
|
|
|
|
|
|
|
|
|

|
|

|
|
|

|
|
|
|+Planning Counci+|+=Seminar/White=+|
|+==================Trade Show/Knox==================+|+====Mgrs. Meeting/District 7=====+|
|+================================Sales Drive/District 6=================================+|

|
|
|

|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
14
|
15
|
16
|
17
|
18
|
19
|
20
|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

|
|
|

|
|

|
|

|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
21
|
22
|
23
|
24
|
25
|
26
|
27
|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

|
|
|
|+=Birthday/Mary=+|+======Close Sale/WYGIX Co.=======+|
|
|
|+===============Inventors Show/Melvin===============+|+Planning Counci+|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
|
|
|
|

|
|
|
|
|

|
|
|

|
|

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
-------------------------------------------------------------------------------------------------------------------------------

Advanced Schedule Calendar

Output 5.2 is an advanced schedule calendar produced by PROC CALENDAR. The
statements that create this calendar
3 schedule activities around holidays
3 identify separate calendars
3 print multiple calendars in the same report
3 apply different holidays to different calendars
3 apply different work patterns to different calendars.

The CALENDAR Procedure

What Types of Calendars Can PROC CALENDAR Produce?

For an explanation of the program that produces this calendar, see Example 4 on
page 127.

Output 5.2 Advanced Schedule Calendar

Well Drilling Work Schedule: Combined Calendars

|
|
|

|+===================Build Pump House/$2,000.00====================+|

|
|<=====================Assemble Tank/$1,000.00=====================+|
|
|<===Lay Power Line/$2,000.00====+|
|+===Pour Foundation/$1,500.00===>|

|
|
|

|.........|................|................|................|................|................|................|................|
| CAL2
|
|<Excavate/$3,50>|****Vacation****|<Excavate/$3,50+|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
14
|
15
|
16
|
17
|
18
|
19
|
20
|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|

|
|
|

|+===============================Install Pump/$500.00===============================+|
|<===========Pour Foundation/$1,500.00============+|
|+Install Pipe/$>|
|
|
|
|
|
|

|
|
|

|+==============================Erect Tower/$2,500.00===============================>|
|<====Install Pipe/$1,000.00=====+|
|
|
|
|
|
|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
----------------------------------------------------------------------------------------------------------------------------------

Simple Summary Calendar

Output 5.3 shows a simple summary calendar that displays the number of meals
served daily in a hospital cafeteria:
options nodate pageno=1 linesize=132 pagesize=60;
proc calendar data=meals;
start date;

What Types of Calendars Can PROC CALENDAR Produce?

Chapter 5

sum brkfst lunch dinner;

mean brkfst lunch dinner;
run;

In a summary calendar, each piece of information for a given day is the value of a
variable for that day. The variables can be either numeric or character, and you can
format them as necessary. You can use the SUM and MEAN options to calculate sums
and means for any numeric variables. These statistics appear in a box below the
calendar, as shown in Output 5.3. The data set that is shown in this calendar is created
in Example 7 on page 143.

The CALENDAR Procedure

Advanced Scheduling and Project Management Tasks

Output 5.3 Simple Summary Calendar

The SAS System

---------------------------------------------------------------------------------------------------------|
|
|

December

|
|
|

1996

|--------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
| Wednesday
|
Thursday
|
Friday
|
Saturday
|
|--------------+--------------+--------------+--------------+--------------+--------------+--------------|
|
1
|
2
|
3
|
4
|
5
|
6
|
7
|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

123 |
234 |
238 |

188 |
188 |
198 |

123 |
183 |
176 |

200 |
267 |
243 |

176 |
165 |
177 |

|
|
|

|--------------+--------------+--------------+--------------+--------------+--------------+--------------|
|
8
|
9
|
10
|
11
|
12
|
13
|
14
|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

178 |
198 |
187 |

165 |
176 |
187 |

187 |
176 |
231 |

176 |
187 |
222 |

187 |
187 |
123 |

|
|
|

|--------------+--------------+--------------+--------------+--------------+--------------+--------------|
|
15
|
16
|
17
|
18
|
19
|
20
|
21
|
|
|
|

|
|
|

|
176 |
165 |

|
156 |
. |

|
198 |
143 |

|
178 |
198 |

|
165 |
176 |

|
|
|

|
|
177 |
167 |
167 |
187 |
187 |
|
|--------------+--------------+--------------+--------------+--------------+--------------+--------------|
|
22
|
23
|
24
|
25
|
26
|
27
|
28
|
|
|
|

|
|
|

|
187 |
187 |

|
|
|

|
|
123 |
|
|
|
|
|
|--------------+--------------+--------------+--------------+--------------+--------------+--------------|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
----------------------------------------------------------------------------------------------------------------------------------------------------|
|
Sum
|
Mean
|
|
| Brkfst
| Lunch

|
|
|

|
2763 |
2830 |

|
172.688 |
188.667 |

| Dinner
|
2990 |
186.875 |
--------------------------------------------

Advanced Scheduling and Project Management Tasks

For more complex scheduling tasks, consider using the CPM procedure in SAS/OR
software. PROC CALENDAR requires that you specify the starting date of each
activity. When the beginning of one task depends on the completion of others and a
date slips in a schedule, recalculating the schedule can be time-consuming. Instead of
manually recalculating dates, you can use PROC CPM to calculate dates for project
activities based on an initial starting date, activity durations, and which tasks are
identied as successors to others. For an example, see Example 6 on page 137.

Syntax: CALENDAR Procedure

Chapter 5

Syntax: CALENDAR Procedure

You must use a START statement.
Required: For schedule calendars, you must also use a DUR or a FIN statement.
Tip: If you use a DUR or FIN statement, then PROC CALENDAR produces a schedule
calendar.
Tip: Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Name: Calendar
Reminder: You can use the FORMAT, LABEL, and WHERE statements as well as any
global statements.
Required:

PROC CALENDAR <option(s)>;

START variable;
BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
CALID variable
</ OUTPUT=COMBINE|MIX|SEPARATE>;
DUR variable;
FIN variable;
HOLISTART variable;
HOLIDUR variable;
HOLIFIN variable;
HOLIVAR variable;
MEAN variable(s) < / FORMAT=format-name>;
OUTSTART day-of-week;
OUTDUR number-of-days;
OUTFIN day-of-week;
SUM variable(s) </ FORMAT=format-name>;
VAR variable(s);

The following table lists the statements and options available in the CALENDAR
procedure according to function.
To do this

Use this statement

Create summary calendar

MEAN
SUM

Create schedule calendar

DUR or FIN

Create multiple calendars

CALID

Specify holidays

HOLISTART
HOLIDUR
HOLIFIN
HOLIVAR

The CALENDAR Procedure

PROC CALENDAR Statement

To do this

Use this statement

Control display

OUTSTART
OUTDUR
OUTFIN

Specify grouping

BY
CALID

PROC CALENDAR Statement

PROC CALENDAR <option(s)>;

To do this

Use this option

Specify data sets containing

weekly work schedules

CALEDATA=

activities

DATA=

holidays

HOLIDATA=

unique shift patterns

WORKDATA=

Control printing
display all months, even if no activities exist

FILL

dene characters used for outlines, dividers, and so on

FORMCHAR=

specify the type of heading for months

HEADER=

display month and weekday names in local language

(experimental)

LOCALE

specify how to show missing values

MISSING

suppress the display of Saturdays and Sundays

WEEKDAYS

Specify time or duration

specify that START and FIN variables are in DATETIME format

DATETIME

specify the number of hours in a standard work day

DAYLENGTH=

specify the units of the DUR and HOLIDUR variables

INTERVAL=

Control summary information

identify variables in the calendar

LEGEND

specify the type of mean to calculate

MEANTYPE=

Options

PROC CALENDAR Statement

Chapter 5

CALEDATA=SAS-data-set

species the calendar data set, a SAS data set that contains weekly work schedules
for multiple calendars.
Default: If you omit the CALEDATA= option, then PROC CALENDAR uses a
default work schedule, as described in The Default Calendars on page 103.
Tip: A calendar data set is useful if you are using multiple calendars or a
nonstandard work schedule.
See also: Calendar Data Set on page 109
Featured in: Example 3 on page 122
DATA=SAS-data-set

species the activities data set, a SAS data set that contains starting dates for all
activities and variables to display for each activity. Activities must be sorted or
indexed by starting date.
Default: If you omit the DATA= option, then the most recently created SAS data set
is used.
See also: Activities Data Set on page 107
Featured in: All examples. See Examples: CALENDAR Procedure on page 114
DATETIME

species that START and FIN variables contain values in DATETIME. format.
Default: If you omit the DATETIME option, then PROC CALENDAR assumes that
the START and FIN values are in the DATE. format.
Featured in: Example 3 on page 122
DAYLENGTH=hours

gives the number of hours in a standard working day. The hour value must be a SAS
TIME value.
Default: 24 if INTERVAL=DAY (the default), 8 if INTERVAL=WORKDAY.
Restriction: DAYLENGTH= applies only to schedule calendars.
Interaction: If you specify the DAYLENGTH= option and the calendar data set
contains a D_LENGTH variable, then PROC CALENDAR uses the DAYLENGTH=
value only when the D_LENGTH value is missing.
Interaction: When INTERVAL=DAY and you have no CALEDATA= data set,
specifying a DAYLENGTH= value has no effect.
Tip: The DAYLENGTH= option is useful when you use the DUR statement and
your work schedule contains days of varying lengths, for example, a 5 half-day
work week. In a work week with varying day lengths, you need to set a standard
day length to use in calculating duration times. For example, an activity with a
duration of 3.0 workdays lasts 24 hours if DAYLENGTH=8:00 or 30 hours if
DAYLENGTH=10:00.
Tip: Instead of specifying the DAYLENGTH= option, you can specify the length of
the working day by using a D_LENGTH variable in the CALEDATA= data set. If
you use this method, then you can specify different standard day lengths for
different calendars.
See also: Calendar Data Set on page 109 for more information on setting the
length of the standard workday
FILL

displays all months between the rst and last activity, start and nish dates
inclusive, including months that contain no activities.
Default: If you do not specify FILL, then PROC CALENDAR prints only months
that contain activities. (Months that contain only holidays are not printed.)

The CALENDAR Procedure

Featured in:

PROC CALENDAR Statement

Example 5 on page 134

FORMCHAR <(position(s))>=formatting-character(s)

denes the characters to use for constructing the outlines and dividers for the cells in
the calendar as well as all identifying markers (such as asterisks and arrows) used to
indicate holidays or continuation of activities in PROC CALENDAR output.
position(s)
identies the position of one or more characters in the SAS formatting-character
string. A space or a comma separates the positions.
Default: Omitting (position(s)) is the same as specifying all 20 possible system
formatting characters, in order.
Range: PROC CALENDAR uses 17 of the 20 formatting characters that SAS
provides. Table 5.1 on page 87 shows the formatting characters that PROC
CALENDAR uses. Figure 5.1 on page 88 illustrates their use in PROC
CALENDAR output.
formatting-character(s)
lists the characters to use for the specied positions. PROC CALENDAR assigns
characters in formatting-character(s) to position(s), in the order that they are listed.
For instance, the following option assigns an asterisk (*) to the twelfth position,
assigns a single dash (-) to the thirteenth, and does not alter remaining characters:
formchar(12 13)=*-

These new settings change the activity line from this:

+=================ACTIVITY===============+

to this:
*------------------ACTIVITY--------------*

Interaction: The SAS system option FORMCHAR= species the default formatting

characters. The SAS system option denes the entire string of formatting
characters. The FORMCHAR= option in a procedure can redene selected
characters.
Tip: You can use any character in formatting-characters, including hexadecimal
characters. If you use hexadecimal characters, then you must put an x after the
closing quotation mark. For instance, the following option assigns the hexadecimal
character 2D to the third formatting character, the hexadecimal character 7C to
the seventh character, and does not alter the remaining characters:

formchar(3,7)=2D7Cx

See also: For information on which hexadecimal codes to use for which characters,

consult the documentation for your hardware.

Table 5.1 Formatting Characters Used by PROC CALENDAR

Position

Default

Used to draw

vertical bar

horizontal bar

cell: upper left corner

cell: upper middle intersection

PROC CALENDAR Statement

Chapter 5

Position

Default

Used to draw

cell: upper right corner

cell: middle left cell side

cell: middle middle intersection

cell: middle right cell side

cell: lower left corner

cell: lower middle intersection

cell: lower right corner

activity start and nish

activity line

activity separator

activity continuation from

activity continuation to

holiday marker

Figure 5.1 Formatting Characters in PROC CALENDAR Output

The CALENDAR Procedure

PROC CALENDAR Statement

HEADER=SMALL | MEDIUM | LARGE

species the type of heading to use in printing the name of the month.
SMALL
prints the month and year on one line.
MEDIUM
prints the month and year in a box four lines high.
LARGE
prints the month seven lines high using asterisks (*). The year is included if space
is available.
Default: MEDIUM
HOLIDATA=SAS-data-set

species the holidays data set, a SAS data set that contains the holidays you want to
display in the output. One variable must contain the holiday names and another
must contain the starting dates for each holiday. PROC CALENDAR marks holidays
in the calendar output with asterisks (*) when space permits.
Interaction: Displaying holidays on a calendar requires a holidays data set and a
HOLISTART statement. A HOLIVAR statement is recommended for naming
holidays. HOLIDUR is required if any holiday lasts longer than one day.
Tip: The holidays data set does not require sorting.
See also: Holidays Data Set on page 108
Featured in: All examples. See Examples: CALENDAR Procedure on page 114
INTERVAL=DAY | WORKDAY

species the units of the DUR and HOLIDUR variables to one of two default
daylengths:
DAY
species the values of the DUR and HOLIDUR variables in units of 24-hour days
and species the default 7-day calendar. For instance, a DUR value of 3.0 is
treated as 72 hours. The default calendar work schedule consists of seven working
days, all starting at 00:00 with a length of 24:00.
WORKDAY
species the values of the DUR and HOLIDUR variables in units of 8-hour days
and species that the default calendar contains ve days a week, Monday through
Friday, all starting at 09:00 with a length of 08:00. When WORKDAY is specied,
PROC CALENDAR treats the values of the DUR and HOLIDUR variables in units
of working days, as dened in the DAYLENGTH= option, the CALEDATA= data
set, or the default calendar. For example, if the working day is 8 hours long, then
a DUR value of 3.0 is treated as 24 hours.
Default: DAY
Interaction: In the absence of a CALEDATA= data set, PROC CALENDAR uses
the work schedule dened in a default calendar.
Interaction: The WEEKDAYS option automatically sets the INTERVAL= value to
WORKDAY.
See also: Calendars and Multiple Calendars on page 104 and Calendar Data Set
on page 109 for more information on the INTERVAL= option and the specication
of working days; The Default Calendars on page 103
Featured in: Example 5 on page 134
LEGEND

prints the names of the variables whose values appear in the calendar. This
identifying text, or legend box, appears at the bottom of the page for each month if

PROC CALENDAR Statement

Chapter 5

space permits; otherwise, it is printed on the following page. PROC CALENDAR

identies each variable by name or by label if one exists. The order of variables in
the legend matches their order in the calendar.
Restriction: LEGEND applies only to summary calendars.
Interaction: If you use the SUM and MEAN statements, then the legend box also

contains SUM and MEAN values.

Featured in:

Example 8 on page 147

LOCALE (Experimental)

prints the names of months and weekdays in the language that is indicated by the
value of the LOCALE= SAS system option. The LOCALE option in PROC
CALENDAR does not change the starting day of the week.
Default: If LOCALE is not specied, then names of months and weekdays are

printed in English.
CAUTION:

LOCALE is an experimental option that is available in SAS 9.1. Do not use this option
in production jobs. 4
MEANTYPE=NOBS | NDAYS

species the type of mean to calculate for each month.

NOBS
calculates the mean over the number of observations displayed in the month.
NDAYS
calculates the mean over the number of days displayed in the month.
Default: NOBS
Restriction: MEANTYPE= applies only to summary calendars.
Interaction: Normally, PROC CALENDAR displays all days for each month.

However, it may omit some days if you use the OUTSTART statement with the
OUTDUR or OUTFIN statement.
Featured in:

Example 7 on page 143

MISSING

determines how missing values are treated, based on the type of calendar.
Summary Calendar
If there is a day without an activity scheduled, then PROC CALENDAR prints the
values of variables for that day by using the SAS or user-dened that is format
specied for missing values.
Default: If you omit MISSING, then days without activities contain no values.

Schedule Calendar
variables with missing values appear in the label of an activity, using the format
specied for missing values.
Default: If you do not specify MISSING, then PROC CALENDAR ignores missing

values in labeling activities.

See also: Missing Values in Input Data Sets on page 111 for more information on

missing values
WEEKDAYS

suppresses the display of Saturdays and Sundays in the output. It also species that
the value of the INTERVAL= option is WORKDAY.
Default: If you omit WEEKDAYS, then the calendar displays all seven days.

The CALENDAR Procedure

BY Statement

The WEEKDAYS option is an alternative to using the combination of

INTERVAL=WORKDAY and the OUTSTART and OUTFIN statements, as shown
here:

Tip:

Example Code 5.1

Illustration of Formatting Characters in PROC CALENDAR Output

proc calendar weekdays;

start date;
run;
proc calendar interval=workday;
start date;
outstart monday;
outfin friday;
run;

Featured in:

Example 1 on page 114

WORKDATA=SAS-data-set

species the workdays data set, a SAS data set that denes the work pattern during
a standard working day. Each numeric variable in the workdays data set denotes a
unique workshift pattern during one working day.
Tip:

The workdays data set is useful in conjunction with the calendar data set.

See also: Workdays Data Set on page 110 and Calendar Data Set on page 109
Featured in:

Example 3 on page 122

BY Statement
Processes activities separately for each BY group, producing a separate calendar for each value of
the BY variable.
Calendar type: Summary and schedule
Main discussion:

BY on page 58

Showing Multiple Calendars in Related Groups

When you use the CALID statement, you can process activities that apply to
different calendars, indicated by the value of the CALID variable. Because you can
specify only one CALID variable, however, you can create only one level of grouping.
For example, if you want a calendar report to show the activities of several departments
within a company, then you can identify each department with the value of the CALID
variable and produce calendar output that shows the calendars for all departments.
When you use a BY statement, however, you can further divide activities into related
groups. For example, you can print calendar output that groups departmental
calendars by division. The observations for activities must contain a variable that
identies which department an activity belongs to and a variable that identies the
division that a department resides in. Specify the variable that identies the
department with the CALID statement. Specify the variable that identies the division
with the BY statement.

CALID Statement
Processes activities in groups dened by the values of a calendar identier variable.
Calendar type: Summary and schedule
Tip: Useful for producing multiple schedule calendars and for use with SAS/OR
software.
See also: Calendar Data Set on page 109
Featured in:

Example 2 on page 118, Example 3 on page 122, and Example 6 on page 137

CALID variable
</ OUTPUT=COMBINE|MIX|SEPARATE>;

Required Arguments
variable

a character or numeric variable that identies which calendar an observation

contains data for.
If you specify the CALID variable, then both the activities and
holidays data sets must contain this variable. If either of these data sets does not
contain the CALID variable, then a default calendar is used.

Requirement:

The CALENDAR Procedure

DUR Statement

Interaction: SAS/OR software uses this variable to identify which calendar an

observation contains data for.

You do not need to use a CALID statement to create this variable. You can
include the default variable _CALID_ in the input data sets.

Tip:

the START variable or must contain an appropriate composite index.

Featured in: Example 3 on page 122 and Example 8 on page 147
Default: COMBINE

DUR Statement
Species the variable that contains the duration of each activity.
Alias:

DURATION

Calendar type: Schedule

Interaction:
Tip:

If you use both a DUR and a FIN statement, then DUR is ignored.

To produce a schedule calendar, you must use either a DUR or FIN statement.

Featured in:

All schedule calendars (see Examples: CALENDAR Procedure on page 114)

DUR variable;

FIN Statement

Chapter 5

Required Arguments
variable

contains the duration of each activity in a schedule calendar.

Range: The duration may be a real or integral value.
Restriction: This variable must be in the activities data set.
See also: For more information on activity durations, see Activities Data Set on
page 107 and Calendar Data Set on page 109

Duration
3 Duration is measured inclusively from the start of the activity (as given in the
START variable). In the output, any activity that lasts part of a day is displayed
as lasting a full day.
3 The INTERVAL= option in a PROC CALENDAR statement automatically sets the
unit of the duration variable, depending on its own value as follows:
If INTERVAL=

Then the default length of the duration unit is

DAY (the default)

24 hours

WORKDAY

8 hours

3 You can override the default length of a duration unit by using

3 the DAYLENGTH= option
3 a D_LENGTH variable in the CALEDATA= data set.

FIN Statement
Species the variable in the activities data set that contains the nishing date of each activity.
FINISH
Calendar type: Schedule
Interaction: If you use both a FIN and a DUR statement, then FIN is used.
Tip: To produce a schedule calendar, you must use either a FIN or DUR statement.
Featured in: Example 6 on page 137
Alias:

FIN variable;

Required Arguments
variable

contains the nishing date of each activity.

Restriction: The values of variable must be either SAS date or datetime values.

The CALENDAR Procedure

HOLIFIN Statement

Restriction: If the FIN variable contains datetime values, then you must specify

the DATETIME option in the PROC CALENDAR statement.

Restriction: Both the START and FIN variables must have matching formats. For

example, if one contains datetime values, then so must the other.

HOLIDUR Statement
Species the variable in the holidays data set that contains the duration of each holiday for a
schedule calendar.
Alias: HOLIDURATION
Calendar type: Schedule
Default: If you do not use a HOLIDUR or HOLIFIN statement, then all holidays last

one day.
Restriction: Cannot use with a HOLIFIN statement.
Featured in: Example 1 on page 114 through Example 5 on page 134

HOLIDUR variable;

Required Arguments
variable

contains the duration of each holiday.

Range: The duration may be a real or integral value.
Restriction: This variable must be in the holidays data set.
Featured in: Example 3 on page 122 and Example 8 on page 147

Holiday Duration
3 If you use both the HOLIFIN and HOLIDUR statements, then PROC CALENDAR
uses the HOLIFIN variable value to dene each holidays duration.
3 Set the unit of the holiday duration variable in the same way that you set the unit
of the duration variable; use either the INTERVAL= and DAYLENGTH= options
or the CALEDATA= data set.
3 Duration is measured inclusively from the start of the holiday (as given in the
HOLISTART variable). In the output, any holiday lasting at least half a day
appears as lasting a full day.

HOLIFIN Statement
Species the variable in the holidays data set that contains the nishing date of each holiday.
Alias: HOLIFINISH
Calendar type: Schedule
Default: If you do not use a HOLIFIN or HOLIDUR statement, then all holidays last

one day.

HOLISTART Statement

Chapter 5

HOLIFIN variable;

Required Arguments
variable

contains the nishing date of each holiday.

Restriction: This variable must be in the holidays data set.
Restriction: Values of variable must be in either SAS date or datetime values.
Restriction: If the HOLIFIN variable contains datetime values, then you must

specify the DATETIME option in the PROC CALENDAR statement.

Holiday Duration
If you use both the HOLIFIN and the HOLIDUR statements, then PROC
CALENDAR uses only the HOLIFIN variable.

HOLISTART Statement
Species a variable in the holidays data set that contains the starting date of each holiday.
Alias:

HOLISTA, HOLIDAY

Calendar type: Summary and schedule

Requirement:
Featured in:

When you use a holidays data set, HOLISTART is required.

Example 1 on page 114 through Example 5 on page 134

HOLISTART variable;

Required Arguments
variable

contains the starting date of each holiday.

Restriction: Values of variable must be in either SAS date or datetime values.
Restriction: If the HOLISTART variable contains datetime values, then specify the

DATETIME option in the PROC CALENDAR statement.

Details
3 The holidays data set need not be sorted.
3 All holidays last only one day, unless you use a HOLIFIN or HOLIDUR statement.
3 If two or more holidays occur on the same day, then PROC CALENDAR uses only
the rst observation.

The CALENDAR Procedure

MEAN Statement

HOLIVAR Statement
Species a variable in the holidays data set whose values are used to label the holidays.
HOLIVARIABLE, HOLINAME
Calendar type: Summary and schedule
Alias:

If you do not use a HOLIVAR statement, then PROC CALENDAR uses the
word DATE to identify holidays.

Default:

Featured in: Example 1 on page 114 through Example 5 on page 134

HOLIVAR variable;

Required Arguments
variable

a variable whose values are used to label the holidays. Typically, this variable
contains the names of the holidays.
Range: character or numeric.
Restriction: This variable must be in the holidays data set.
Tip:

You can format the HOLIVAR variable as you like.

MEAN Statement
Species numeric variables in the activities data set for which mean values are to be calculated
for each month.
Calendar type: Summary
Tip:

You can use multiple MEAN statements.

Featured in: Example 7 on page 143

MEAN variable(s) < / FORMAT=format-name>;

Required Arguments
variable(s)

numeric variable for which mean values are calculated for each month.
Restriction: This variable must be in the activities data set.

OUTDUR Statement

Chapter 5

Options
FORMAT=format-name

names a SAS or user-dened format to be used in displaying the means requested.

Alias: F=
Default: BEST. format
Featured in: Example 7 on page 143

What Is Displayed and How

3 The means appear at the bottom of the summary calendar page, if there is room;
otherwise they appear on the following page.
3 The means appear in the LEGEND box if you specify the LEGEND option.
3 PROC CALENDAR automatically displays variables named in a MEAN statement
in the calendar output, even if the variables are not named in the VAR statement.

OUTDUR Statement
Species in days the length of the week to be displayed.
OUTDURATION
Requirement: The OUTSTART statement is required.
Alias:

OUTDUR number-of-days;

Required Arguments
number-of-days

an integer that expresses the length in days of the week to be displayed.

Length of Week
Use either the OUTDUR or OUTFIN statement to supply the procedure with
information about the length of the week to display. If you use both, then PROC
CALENDAR ignores the OUTDUR statement.

OUTFIN Statement
Species the last day of the week to display in the calendar.
OUTFINISH
The OUTSTART statement is required.
Featured in: Example 3 on page 122 and Example 8 on page 147
Alias:

Requirement:

The CALENDAR Procedure

START Statement

OUTFIN day-of-week;

Required Arguments
day-of-week

the name of the last day of the week to display. For example,
outfin friday;

Length of Week
Use either the OUTFIN or OUTDUR statement to supply the procedure with
information about the length of the week to display. If you use both, then PROC
CALENDAR uses only the OUTFIN statement.

OUTSTART Statement
Species the starting day of the week to display in the calendar.
OUTSTA
If you do not use OUTSTART, then each calendar week begins with Sunday.
Featured in: Example 3 on page 122 and Example 8 on page 147
Alias:

Default:

OUTSTART day-of-week;

Required Arguments
day-of-week

the name of the starting day of the week for each week in the calendar. For example,
outstart monday;

Interaction with OUTDUR and OUTFIN

By default, a calendar displays all seven days in a week. Use OUTDUR or OUTFIN,
in conjunction with OUTSTART, to control how many days are displayed and which day
starts the week.

START Statement
Species the variable in the activities data set that contains the starting date of each activity.
STA, DATE, ID
Required: START is required for both summary and schedule calendars.
Featured in: All examples
Alias:

100

SUM Statement

Chapter 5

START variable;

Required Arguments
variable

contains the starting date of each activity.

Restriction: This variable must be in the activities data set.
Restriction: Values of variable must be in either SAS date or datetime values.
Restriction: If you use datetime values, then specify the DATETIME option in the

PROC CALENDAR statement.

Restriction: Both the START and FIN variables must have matching formats. For

example, if one contains datetime values, then so must the other.

SUM Statement
Species numeric variables in the activities data set to total for each month.
Calendar type: Summary
Tip: To apply different formats to variables that are being summed, use multiple SUM
statements.

Example 7 on page 143 and Example 8 on page 147

Featured in:

SUM variable(s) </ FORMAT=format-name>;

Required Arguments
variable(s)

species one or more numeric variables to total for each month.

Restriction: This variable must be in the activities data set.

Options
FORMAT=format-name

names a SAS or user-dened format to use in displaying the sums requested.

Alias:

Default: BEST. format

Featured in:

Example 7 on page 143 and Example 8 on page 147

What Is Displayed and How

3 The sum appears at the bottom of the calendar page, if there is room; otherwise, it
appears on the following page.

The CALENDAR Procedure

Type of Calendars

101

3 The sum appears in the LEGEND box if you specify the LEGEND option.
3 PROC CALENDAR automatically displays variables named in a SUM statement
in the calendar output, even if the variables are not named in the VAR statement.

VAR Statement
Species the variables that you want to display for each activity.
Alias:

VARIABLE

VAR variable(s);

Required Arguments
variable(s)

species one or more variables that you want to display in the calendar.
Range: The values of variable can be either character or numeric.
Restriction: These variables must be in the activities data set.
Tip:

You can apply a format to this variable.

Details
When VAR Is Not Used
If you do not use a VAR statement, then the procedure displays all variables in the
activities data set in the order in which they occur in the data set, except for the BY,
CALID, START, DUR, and FIN variables. However, not all variables are displayed if
the LINESIZE= and PAGESIZE= settings do not allow enough space in the calendar.

Display of Variables

3 PROC CALENDAR displays variables in the order that they appear in the VAR
statement. Not all variables are displayed, however, if the LINESIZE= and
PAGESIZE= settings do not allow enough space in the calendar.

3 PROC CALENDAR also displays any variable named in a SUM or MEAN

statement for each activity in the calendar output, even if you do not name that
variable in a VAR statement.

Concepts: CALENDAR Procedure

Type of Calendars
PROC CALENDAR can produce two kinds of calendars: schedule and summary.

102

Schedule Calendar

Chapter 5

Use a

if you want to

and can accept this

restriction

schedule calendar

schedule activities around holidays

and nonwork periods

cannot calculate sums and

means

schedule calendar

schedule activities that last more than

one day

summary calendar

calculate sums and means

activities can last only one

day

Note: PROC CALENDAR produces a summary calendar if you do not use a DUR or
FIN statement in the PROC step. 4

Schedule Calendar
Denition
A report in calendar format that shows when activities and holidays start and end.

Required Statements
You must supply a START statement and either a DUR or FIN statement.
Use this statement

to specify a variable whose value indicates the

START

starting date of an activity

DUR*

duration of an activity

FIN*

ending date of an activity

* Choose one of these. If you do not use a DUR or FIN statement, then PROC CALENDAR
assumes that you want to create a summary calendar report.

Examples
See Simple Schedule Calendar on page 79, Advanced Schedule Calendar on page
80, as well as Example 1 on page 114, Example 2 on page 118, Example 3 on page 122,
Example 4 on page 127, Example 5 on page 134, and Example 6 on page 137

Summary Calendar
Denition
A report in calendar format that displays activities and holidays that last only one
day and that can provide summary information in the form of sums and means.

The CALENDAR Procedure

The Default Calendars

103

Required Statements
You must supply a START statement. This statement identies the variable in the
activities data set that contains an activitys starting date.

Multiple Events on a Single Day

A summary calendar report can display only one activity on a given date. Therefore,
if more than one activity has the same START value, then only the last observation
that was read is used. In such situations, you might nd PROC SUMMARY useful in
collapsing your data set to contain one activity per starting date.

Examples
See Simple Summary Calendar on page 81, Example 7 on page 143, and Example 8
on page 147

The Default Calendars

Description
PROC CALENDAR provides two default calendars for simple applications. You can
produce calendars without having to specify detailed workshifts and weekly work
patterns if your application can use one of two simple work patterns. Consider using a
default calendar if

3 your application uses a 5-day work week with 8-hour days or a 7-day work week
with 24-hour days. See Table 5.2 on page 103.

3 you want to print all activities on the same calendar.

3 you do not need to identify separate calendars.

Table 5.2 Default Calendar Settings and Examples

If scheduled work days are

Then set
INTERVAL=

By default
DAYLENGTH=

So work periods are

Shown in
Example

7 (M-Sun)

DAY

24-hour days

5 (M-F)

WORKDAY

8-hour days

When You Unexpectedly Produce a Default Calendar

If you want to produce a specialized calendar but do not provide all the necessary
information, then PROC CALENDAR attempts to produce a default calendar. These
errors cause PROC CALENDAR to produce a calendar with default features:

3 If the activities data set does not contain a CALID variable, then PROC
CALENDAR produces a default calendar.

3 If both the holidays and calendar data sets do not contain a CALID variable, then
PROC CALENDAR produces a default calendar even if the activities data set
contains a CALID variable.

3 If the activities and calendar data sets contain the CALID variable, but the
holidays data set does not, then the default holidays are used.

104

Calendars and Multiple Calendars

Chapter 5

Examples
See the 7-day default calendar in Output 5.1 and the 5-day default calendar in
Example 1 on page 114

Calendars and Multiple Calendars

Denitions
calendar
a logical entity that represents a weekly work pattern, which consists of weekly
work schedules and daily shifts. PROC CALENDAR contains two default work
patterns: 5-day week with an 8-hour day or a 7-day week with a 24-hour day. You
can also dene your own work patterns by using CALENDAR and WORKDAYS
data sets.
calendar report
a report in calendar format that displays activities, holidays, and nonwork periods.
A calendar report can contain multiple calendars in one of three formats
separate
Each identied calendar prints on separate output pages.
combined
All identied calendars print on the same output pages and each is identied.
mixed
All identied calendars print on the same output pages but are not identied
as belonging to separate calendars.
multiple calendar
a logical entity that represents multiple weekly work patterns.

Why Create Multiple Calendars

Create a multiple calendar if you want to print a calendar report that shows
activities that follow different work schedules or different weekly work patterns. For
example, a construction project report might need to use different work schedules and
weekly work patterns for work crews on different parts of the project.
Another use for multiple calendars is to identify activities so that you can choose to
print them in the same calendar report. For example, if you identify activities as
belonging to separate departments within a division, then you can choose to print a
calendar report that shows all departmental activities on the same calendar.
Finally, using multiple calendars, you can produce separate calendar reports for each
calendar in a single step. For example, if activities are identied by department, then
you can produce a calendar report that prints the activities of each department on
separate pages.

How to Identify Multiple Calendars

Because PROC CALENDAR can process only one data set of each type (activities,
holidays, calendar, workdays) in a single PROC step, you must be able to identify for
PROC CALENDAR which calendar an activity, holiday, or weekly work pattern belongs

The CALENDAR Procedure

Calendars and Multiple Calendars

105

to. Use the CALID statement to specify the variable whose values identify the
appropriate calendar. This variable can be numeric or character.
You can use the special variable name _CAL_ or you can use another variable name.
PROC CALENDAR automatically looks for a variable named _CAL_ in the holiday and
calendar data sets, even when the activities data set uses a variable with another name
as the CALID variable. Therefore, if you use the name _CAL_ in your holiday and
calendar data sets, then you can more easily reuse these data sets in different calendar
applications.

Using Holidays or Calendar Data Sets with Multiple Calendars

When using a holidays or calendar data set with multiple calendars, PROC
CALENDAR treats the variable values in the following way:

3 Every value of the CALID variable that appears in either the holidays or calendar
data sets denes a calendar.

3 If a CALID value appears in the HOLIDATA= data set but not in the
CALEDATA= data set, then the work schedule of the default calendar is used.

3 If a CALID value appears in the CALEDATA= data set but not in the
HOLIDATA= data set, then the holidays of the default calendar are used.

3 If a CALID value does not appear in either the HOLIDATA= or CALEDATA= data
set, then the work schedule and holidays of the default calendar are used.

3 If the CALID variable is not found in the holiday or calendar data set, then PROC
CALENDAR looks for the default variable _CAL_ instead. If neither the CALID
variable nor a _CAL_ variable appears in a data set, then the observations in that
data set are applied to a default calendar.

Types of Reports That Contain Multiple Calendars

Because you can associate different observations with different calendars, you can
print a calendar report that shows activities that follow different work schedules or
different work shifts or that contain different holidays. You can

3 print separate calendars on the same page and identify each one.
3 print separate calendars on the same page without identifying them.
3 print separate pages for each identied calendar.
As an example, consider a calendar that shows the activities of all departments
within a division. Each department can have its own calendar identication value and,
if necessary, can have individual weekly work patterns, daily work shifts, and holidays.
If you place activities that are associated with different calendars in the same
activities data sets, then you use PROC CALENDAR to produce calendar reports that
print

3 the schedule and events for each department on a separate pages (separate output)
3 the schedule and events for the entire division, each identied by department
(combined output)

3 the schedule and events for the entire division, but not identied by department
(mixed output).
The multiple-calendar feature was added specically to enable PROC CALENDAR to
process the output of PROC CPM in SAS/OR software, a project management tool. See
Example 6 on page 137.

106

Input Data Sets

Chapter 5

How to Identify Calendars with the CALID Statement and the Special
Variable _CAL_
To identify multiple calendars, you must use the CALID statement to specify the
variable whose values identify which calendar an event belongs with. This variable can
be numeric or character.
You can use the special variable name _CAL_ or you can use another variable name.
PROC CALENDAR automatically looks for a variable named _CAL_ in the holiday and
calendar data sets, even when the activities data set uses a variable with another name
as the CALID variable. Therefore, if you use the name _CAL_ in your holiday and
calendar data sets, then you can more easily reuse these data sets in different calendar
applications.

When You Use Holidays or Calendar Data Sets

When you use a holidays or calendar data set with multiple calendars, PROC
CALENDAR treats the variable values in the following way:

3 Every value of the CALID variable that appears in either the holidays or calendar
data sets denes a calendar.

3 If a CALID value appears in the HOLIDATA= data set but not in the
CALEDATA= data set, then the work schedule of the default calendar is used.

3 If a CALID value appears in the CALEDATA= data set but not in the
HOLIDATA= data set, then the holidays of the default calendar are used.

3 If a CALID value does not appear in either the HOLIDATA= or CALEDATA= data
set, then the work schedule and holidays of the default calendar are used.

3 If the CALID variable is not found in the holiday or calendar data sets, then
PROC CALENDAR looks for the default variable _CAL_ instead. If neither the
CALID variable nor a _CAL_ variable appears in a data set, then the observations
in that data set are applied to a default calendar.

Examples
Example 2 on page 118, Example 3 on page 122, Example 4 on page 127, and
Example 8 on page 147

Input Data Sets

You may need several data sets to produce a calendar, depending on the complexity
of your application. PROC CALENDAR can process one of each of four data sets. See
Table 5.3 on page 106.
Table 5.3 Four Possible Input Data Sets for PROC CALENDAR
Data Set

Description

Specify with the

activities

Each observation contains information

about a single activity.

DATA= option

holidays

Each observation contains information

about a holiday

HOLIDATA= option

The CALENDAR Procedure

Activities Data Set

Data Set

Description

Specify with the

calendar

Each observation denes one weekly

work schedule.

CALEDATA= option

workdays

Each variable represents one daily

schedule of alternating work and
nonwork periods.

107

WORKDATA= option

Activities Data Set

Purpose
The activities data set, specied with the DATA= option, contains information about
the activities to be scheduled by PROC CALENDAR. Each observation describes a
single activity.

Requirements and Restrictions

3 An activities data set is required. (If you do not specify an activities data set with
the DATA= option, then PROC CALENDAR uses the _LAST_ data set.)

3 Only one activities data set is allowed.

3 The activities data set must always be sorted or indexed by the START variable.
3 If you use a CALID (calendar identier) variable and want to produce output that
shows multiple calendars on separate pages, then the activities data set must be
sorted by or indexed on the CALID variable and then the START variable.
3 If you use a BY statement, then the activities data set must be sorted by or
indexed on the BY variables.

Structure
Each observation in the activities data set contains information about one activity.
One variable must contain the starting date. If you are producing a schedule calendar,
then another variable must contain either the activity duration or nishing date. Other
variables can contain additional information about an activity.
If a variable contains an
activitys

Then specify it with the

starting date

START statement

For this type of

calendar
Schedule
Summary

duration

DUR statement

Schedule

nishing date

FIN statement

Schedule

Multiple Activities per Day in Summary Calendars

A summary calendar can display only one activity on a given date. Therefore, if more
than one activity has the same START value, then only the last observation that is read

108

Holidays Data Set

Chapter 5

is used. In such situations, you might nd PROC SUMMARY useful to collapse your
data set to contain one activity per starting date.

Examples
Every example in the Examples section uses an activities data set.

Holidays Data Set

Purpose
You can use a holidays data set, specied with the HOLIDATA= option, to

3 identify holidays on your calendar output

3 identify days that are not available for scheduling work. (In a schedule calendar,
PROC CALENDAR does not schedule activities on these days.)

Structure
Each observation in the holidays data set must contain at least the holiday starting
date. A holiday lasts only one day unless a duration or nishing date is specied.
Supplying a holiday name is recommended, though not required. If you do not specify
which variable contains the holiday name, then PROC CALENDAR uses the word DATE
to identify each holiday.
If a variable contains a holidays

Then specify it with this statement

starting date

HOLISTART

name

HOLIVAR

duration

HOLIDUR

nishing date

HOLIFIN

No Sorting Needed
You do not need to sort or index the holidays data set.

Using SAS Date Versus SAS Datetime Values

PROC CALENDAR calculates time using SAS datetime values. Even when your data
is in DATE. format, the procedure automatically calculates time in minutes and
seconds. Therefore, if you specify only date values, then PROC CALENDAR prints
messages similar to the following ones to the SAS log:
NOTE: All holidays are assumed to start at the
time/date specified for the holiday variable
and last one DTWRKDAY.
WARNING: The units of calculation are SAS datetime

The CALENDAR Procedure

Calendar Data Set

109

values while all the holiday variables are

not. All holidays are converted to SAS
datetime values.

Create a Generic Holidays Data Set

If you have many applications that require PROC CALENDAR output, then consider
creating a generic holidays data set that contains standard holidays. You can begin
with the generic holidays and add observations that contain holidays or nonwork events
specic to an application.

Holidays and Nonwork Periods

Do not schedule holidays during nonwork periods. Holidays that are dened in the
HOLIDATA= data set cannot occur during any nonwork periods that are dened in the
work schedule. For example, you cannot schedule Sunday as a vacation day if the work
week is dened as Monday through Friday. When such a conict occurs, the holiday is
rescheduled to the next available working period following the nonwork day.

Examples
Every example in the Examples section uses a holidays data set.

Calendar Data Set

Purpose
You can use a calendar data set, specied with the CALEDATA= option, to specify
work schedules for different calendars.

Structure
Each observation in the calendar data set denes one weekly work schedule. The
data set created in the DATA step shown below denes weekly work schedules for two
calendars, CALONE and CALTWO.
data cale;
input _sun_ $ _mon_ $ _tue_ $ _wed_ $ _thu_ $ /
_fri_ $ _sat_ $ _cal_ $ d_length time6.;
datalines;
holiday workday workday workday workday
workday holiday calone 8:00
holiday shift1 shift1 shift1 shift1
shift2 holiday caltwo 9:00
;

The variables in this calendar data set consist of

_SUN_ through _SAT_
the name of each day of the week that appears in the calendar. The values of
these variables contain the name of workshifts. Valid values for workshifts are
3 WORKDAY (the default workshift)
3 HOLIDAY (a nonwork period)
3 names of variables in the WORKDATA= data set (in this example, SHIFT1
and SHIFT2).

110

Workdays Data Set

Chapter 5

_CAL_
the CALID (calendar identier) variable. The values of this variable identify
different calendars. If this variable is not present, then the rst observation in
this data set denes the work schedule that is applied to all calendars in the
activities data set.
If the CALID variable contains a missing value, then the character or numeric
value for the default calendar (DEFAULT or 0) is used. See The Default Calendars
on page 103 for further details.
D_LENGTH
the daylength identier variable. Values of D_LENGTH indicate the length of the
standard workday to be used in calendar calculations. You can set the workday
length either by placing this variable in your calendar data set or by using the
DAYLENGTH= option.
Missing values for this variable default to the number of hours specied in the
DAYLENGTH= option; if the DAYLENGTH= option is not used, the day length
defaults to 24 hours if INTERVAL=DAY, or 8 hours if INTERVAL=WORKDAY.

Using Default Workshifts Instead of a Workdays Data Set

You can use a calendar data set with or without a workdays data set. Without a
workdays data set, WORKDAY in the calendar data set is equal to one of two standard
workdays, depending on the setting of the INTERVAL= option:
If INTERVAL=

Then the work-shift begins at

And the day length is

DAY

00:00

24 hours

WORKDAY

9:00

8 hours

You can reset the length of the standard workday with the DAYLENGTH= option or
a D_LENGTH variable in the calendar data set. You can dene other work shifts in a
workdays data set.

Examples
Example 3 on page 122, Example 4 on page 127, and Example 7 on page 143 feature
a calendar data set.

Workdays Data Set

Purpose
You can use a workdays data set, specied with the WORKDATA= option, to dene
the daily workshifts named in a CALEDATA= data set.

Use Default Work Shifts or Create Your Own?

You do not need a workdays data set if your application can use one of two default
work shifts:

The CALENDAR Procedure

Missing Values in Input Data Sets

If INTERVAL=

Then the work-shift begins at

And the day length is

DAY

00:00

24 hours

WORKDAY

9:00

111

8 hours

See the INTERVAL= option on page 89.

Structure
Each variable in the workdays data set contains one daily schedule of alternating
work and nonwork periods. For example, this DATA step creates a data set that
contains specications for two work shifts:
data work;
input shift1 time6. shift2 time6.;
datalines;
7:00 7:00
12:00 11:00
13:00
.
17:00
.
;

The variable SHIFT1 species a 10-hour workday, with one nonwork period (a lunch
hour); the variable SHIFT2 species a 4-hour workday with no nonwork periods.

How Missing Values Are Treated

The missing values default to 00:00 in the rst observation and to 24:00 in all other
observations. Two consecutive values of 24:00 dene a zero-length time period, which is
ignored.

Examples
See Example 3 on page 122

Missing Values in Input Data Sets

Table 5.4 on page 111 summarizes the treatment of missing values for variables in
the data sets used by PROC CALENDAR.
Table 5.4 Treatment of Missing Values in PROC CALENDAR
Data set

Variable

Treatment of missing values

Activities (DATA=)

CALID

default calendar value is used

START

observation is not used

DUR

1.0 is used

FIN

START value + daylength is used

112

Results: CALENDAR Procedure

Data set

Chapter 5

Variable
VAR

CALID

default calendar value is used

corresponding shift for default calendar is used

D_LENGTH

if available, DAYLENGTH= value is used; otherwise, if

INTERVAL=DAY, 24:00 is used; otherwise 8:00 is used

SUM, MEAN

CALID

all holidays apply to all calendars

HOLISTART

observation is not used

HOLIDUR

if available, HOLIFIN value is used instead of

HOLIDUR value; otherwise 1.0 is used

HOLIFIN

if available, HOLIDUR value is used instead of HOLIFIN

value; otherwise, HOLISTART value + day length is used

HOLIVAR
Workdays (WORKDATA=)

_SUN_ through _SAT_

Holiday (HOLIDATA=)

if a summary calendar or the MISSING option is

specied, then the missing value is used; otherwise, no
value is used

SUM, MEAN
Calendar (CALEDATA=)

Treatment of missing values

no value is used

any

for the rst observation, 00:00 is used; otherwise, 24:00

is used

Results: CALENDAR Procedure

What Affects the Quantity of PROC CALENDAR Output

The quantity of printed calendar output depends on
3 the range of dates in the activities data set
3 whether the FILL option is specied
3 the BY statement
3 the CALID statement.
PROC CALENDAR always prints one calendar for every month that contains any
activities. If you specify the FILL option, then the procedure prints every month
between the rst and last activities, including months that contain no activities. Using
the BY statement prints one set of output for each BY value. Using the CALID
statement with OUTPUT=SEPARATE prints one set of output for each value of the
CALID variable.

How Size Affects the Format of PROC CALENDAR Output

PROC CALENDAR always attempts to t the calendar within a single page, as
dened by the SAS system options PAGESIZE= and LINESIZE=. If the PAGESIZE=
and LINESIZE= values do not allow sufcient room, then PROC CALENDAR might
print the legend box on a separate page. If necessary, PROC CALENDAR truncates or

The CALENDAR Procedure

Portability of ODS Output with PROC CALENDAR

113

omits values to make the output t the page and prints messages to that effect in the
SAS log.

What Affects the Lines That Show Activity Duration

In a schedule calendar, the duration of an activity is shown by a continuous line
through each day of the activity. Values of variables for each activity are printed on the
same line, separated by slashes (/). Each activity begins and ends with a plus sign (+).
If an activity continues from one week to the next, then PROC CALENDAR displays
arrows (< >) at the points of continuation.
The length of the activity lines depends on the amount of horizontal space available.
You can increase this by specifying

3 a larger linesize with the LINESIZE= option in the OPTIONS statement

3 the WEEKDAYS option to suppress the printing of Saturday and Sunday, which
provides more space for Monday through Friday.

Customizing the Calendar Appearance

PROC CALENDAR uses 17 of the 20 SAS formatting characters to construct the
outline of the calendar and to print activity lines and to indicate holidays. You can use
the FORMCHAR= option to customize the appearance of your PROC CALENDAR
output by substituting your own characters for the default. See Table 5.1 on page 87
and Figure 5.1 on page 88.
If your printer supports an extended character set (one that includes graphics
characters in addition to the regular alphanumeric characters), then you can greatly
improve the appearance of your output by using the FORMCHAR= option to redene
formatting characters with hexadecimal characters. For information on which
hexadecimal codes to use for which characters, consult the documentation for your
hardware. For an example of assigning hexadecimal values, see FORMCHAR= on page
87.

Portability of ODS Output with PROC CALENDAR

Under certain circumstances, using PROC CALENDAR with the Output Delivery
System produces les that are not portable. If the SAS system option FORMCHAR= in
your SAS session uses nonstandard line-drawing characters, then the output might
include strange characters instead of lines in operating environments in which the SAS
Monospace font is not installed. To avoid this problem, specify the following OPTIONS
statement before executing PROC CALENDAR:
options formchar="|----|+|---+=|-/\<>*";

114

Examples: CALENDAR Procedure

Chapter 5

Examples: CALENDAR Procedure

Example 1: Schedule Calendar with Holidays: 5-Day Default

Procedure features:

PROC CALENDAR statement options:

DATA=
HOLIDATA=
WEEKDAYS
DUR statement
HOLISTART statement
HOLIVAR statement
HOLIDUR statement
START statement
Other features:

PROC SORT statement

BY statement
5-day default calendar

This example
3 creates a schedule calendar
3 uses one of the two default work patterns: 8-hour day, 5-day week

3 schedules activities around holidays

3 displays a 5-day week

Program

Create the activities data set. ALLACTY contains both personal and business activities
information for a bank president.
data allacty;
input date : date7. event $ 9-36 who $ 37-48 long;
datalines;
01JUL96 Dist. Mtg.
All
1
17JUL96 Bank Meeting
1st Natl
1
02JUL96 Mgrs. Meeting
District 6
2
11JUL96 Mgrs. Meeting
District 7
2
03JUL96 Interview
JW
1
08JUL96 Sales Drive
District 6
5
15JUL96 Sales Drive
District 7
5
08JUL96 Trade Show
Knox
3

The CALENDAR Procedure

22JUL96
11JUL96
18JUL96
25JUL96
12JUL96
19JUL96
18JUL96
05JUL96
19JUL96
16JUL96
24JUL96
25JUL96
;

Inventors Show
Planning Council
Planning Council
Planning Council
Seminar
Seminar
NewsLetter Deadline
VIP Banquet
Co. Picnic
Dentist
Birthday
Close Sale

Melvin
Group II
Group III
Group IV
White
White
All
JW
All
JW
Mary
WYGIX Co.

Program

115

3
1
1
1
1
1
1
1
1
1
1
2

Create the holidays data set.

data hol;
input date : date7. holiday $ 11-25 holilong @27;
datalines;
05jul96
Vacation
3
04jul96
Independence
1
;

Sort the activities data set by the variable that contains the starting date. You are not
required to sort the holidays data set.
proc sort data=allacty;
by date;
run;

Set LINESIZE= appropriately. If the line size is not long enough to print the variable values,
then PROC CALENDAR either truncates the values or produces no calendar output.
options nodate pageno=1 linesize=132 pagesize=60;

Create the schedule calendar. DATA= identies the activities data set; HOLIDATA= identies
the holidays data set. WEEKDAYS species that a week consists of ve eight-hour work days.
proc calendar data=allacty holidata=hol weekdays;

Specify an activity start date variable and an activity duration variable. The START
statement species the variable in the activities data set that contains the starting date of the
activities; DUR species the variable that contains the duration of each activity. Creating a
schedule calendar requires START and DUR.
start date;
dur long;

116

Program

Chapter 5

Retrieve holiday information. The HOLISTART, HOLIVAR, and HOLIDUR statements

specify the variables in the holidays data set that contain the start date, name, and duration of
each holiday, respectively. When you use a holidays data set, HOLISTART is required. Because
at least one holiday lasts more than one day, HOLIDUR is required.
holistart date;
holivar holiday;
holidur holilong;

Specify the titles.

title1 Summer Planning Calendar: Julia Cho;
title2 President, Community Bank;
run;

The CALENDAR Procedure

Output

117

Output

Output 5.4 Schedule Calendar: 5-Day Week with Holidays

Summer Planning Calendar:

Julia Cho

President, Community Bank

----------------------------------------------------------------------------------------------------------------------------------|
|
|

July

|
|
|

1996

|---------------------------------------------------------------------------------------------------------------------------------|
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
|-------------------------+-------------------------+-------------------------+-------------------------+-------------------------|
|
1
|
2
|
3
|
4
|
5
|
|
|
|
|******Independence*******|********Vacation*********|
|
|
|

|
|
|

|
|
|

|-------------------------+-------------------------+-------------------------+-------------------------+-------------------------|
|
8
|
9
|
10
|
11
|
12
|
|********Vacation*********|********Vacation*********|
|
|
|
|
|
|
|
|
|

|
|
|

|
|+Planning Council/Group +|+=====Seminar/White=====+|
|+==============================Trade Show/Knox==============================+|
|+==========================Sales Drive/District 6===========================>|

|
|
|+====VIP Banquet/JW=====+|+============Mgrs. Meeting/District 7=============+|
|-------------------------+-------------------------+-------------------------+-------------------------+-------------------------|
|
15
|
16
|
17
|
18
|
19
|
|
|
|

|
|
|

|
|

|
|
|

|
|
|+======Dentist/JW=======+|

|
|
|

|+====================================================Sales Drive/District 7=====================================================+|

|<=============Sales Drive/District 6==============+|+=Bank Meeting/1st Natl=+|+Planning Council/Group +|+=====Seminar/White=====+|
|-------------------------+-------------------------+-------------------------+-------------------------+-------------------------|
|
|
|
|
|
|

|
|
|
|
|
|

|
|
|+=====Birthday/Mary=====+|+==============Close Sale/WYGIX Co.===============+|
|+===========================Inventors Show/Melvin===========================+|+Planning Council/Group +|
|
|-------------------------+-------------------------+-------------------------+-------------------------+-------------------------|
|
29
|
30
|
31
|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

-----------------------------------------------------------------------------------------------------------------------------------

118

Example 2: Schedule Calendar Containing Multiple Calendars

Chapter 5

Example 2: Schedule Calendar Containing Multiple Calendars

Procedure features:

CALID statement:
_CAL_ variable
OUTPUT=COMBINE option
DUR statement
24-hour day, 7-day week

This example builds on Example 1 by identifying activities as belonging to one of two

calendars, business or personal. This example
3 produces a schedule calendar report
3 prints two calendars on the same output page
3 schedules activities around holidays
3 uses one of the two default work patterns: 24-hour day, 7-day week
3 identies activities and holidays by calendar name.

Program

Create the activities data set and identify separate calendars. ALLACTY2 contains both
personal and business activities for a bank president. The _CAL_ variable identies which
calendar an event belongs to.
data allacty2;
input date:date7. happen $ 10-34 who $ 35-47
datalines;
01JUL96 Dist. Mtg.
All
02JUL96 Mgrs. Meeting
District 6
03JUL96 Interview
JW
05JUL96 VIP Banquet
JW
06JUL96 Beach trip
family
08JUL96 Sales Drive
District 6
08JUL96 Trade Show
Knox
09JUL96 Orthodontist
Meagan
11JUL96 Mgrs. Meeting
District 7
11JUL96 Planning Council
Group II
12JUL96 Seminar
White
14JUL96 Co. Picnic
All
14JUL96 Business trip
Fred
15JUL96 Sales Drive
District 7
16JUL96 Dentist
JW
17JUL96 Bank Meeting
1st Natl
17JUL96 Real estate agent
Family
18JUL96 NewsLetter Deadline
All
18JUL96 Planning Council
Group III
19JUL96 Seminar
White
22JUL96 Inventors Show
Melvin
24JUL96 Birthday
Mary

_CAL_ $ long;
CAL1
CAL1
CAL1
CAL1
CAL2
CAL1
CAL1
CAL2
CAL1
CAL1
CAL1
CAL1
CAL2
CAL1
CAL1
CAL1
CAL2
CAL1
CAL1
CAL1
CAL1
CAL1

1
2
1
1
2
5
3
1
2
1
1
1
2
5
1
1
1
1
1
1
3
1

The CALENDAR Procedure

25JUL96
25JUL96
27JUL96
;

Planning Council
Close Sale
Ballgame

Group IV
WYGIX Co.
Family

CAL1
CAL1
CAL2

Program

119

1
2
1

Create the holidays data set and identify which calendar a holiday affects. The _CAL_
variable identies which calendar a holiday belongs to.
data vac;
input hdate:date7. holiday $ 11-25 _CAL_ $ ;
datalines;
29JUL96
vacation
CAL2
04JUL96
Independence
CAL1
;

Sort the activities data set by the variable that contains the starting date. When
creating a calendar with combined output, you sort only by the activity starting date, not by the
CALID variable. You are not required to sort the holidays data set.
proc sort data=allacty2;
by date;
run;

Set LINESIZE= appropriately. If the linesize is not long enough to print the variable values,
then PROC CALENDAR either truncates the values or produces no calendar output.
options nodate pageno=1 pagesize=60 linesize=132;

Create the schedule calendar. DATA= identies the activities data set; HOLIDATA=
identies the holidays data set. By default, the output calendar displays a 7-day week.
proc calendar data=allacty2 holidata=vac;

Combine all events and holidays on a single calendar. The CALID statement species the
variable that identies which calendar an event belongs to. OUTPUT=COMBINE places all
events and holidays on the same calendar.
calid _CAL_ / output=combine;

120

Program

Chapter 5

Retrieve holiday information. The HOLISTART and HOLIVAR statements specify the
variables in the holidays data set that contain the start date and name of each holiday,
respectively. HOLISTART is required when you use a holidays data set.
holistart hdate;
holivar holiday;

Specify the titles.

title1 Summer Planning Calendar: Julia Cho;
title2 President, Community Bank;
title3 Work and Home Schedule;
run;

The CALENDAR Procedure

Output

121

Output

Output 5.5 Schedule Calendar Containing Multiple Calendars

Summer Planning Calendar:

Julia Cho

President, Community Bank

Work and Home Schedule
-----------------------------------------------------------------------------------------------------------------------|
|
|
July 1996
|
|
|
|----------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
----------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
|
1
|
2
|
3
|
4
|
5
|
6
|
|.........|................|................|................|................|................|................|................|
| CAL2
|
|
|
|
|
|
|+Beach trip/fam>|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|

|
|
|

|
|
|

|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
7
|
8
|
9
|
10
|
11
|
12
|
13
|
|.........|................|................|................|................|................|................|................|
| CAL2
|<Beach trip/fam+|
|+Orthodontist/M+|
|
|
|
|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|

|
|
|

|
|
|

|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
14
|
15
|
16
|
17
|
18
|
19
|
20
|
|.........|................|................|................|................|................|................|................|
| CAL2
|+======Business trip/Fred=======+|
|+Real estate ag+|
|
|
|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|
|
|
|+Planning Counc+|
|
|
|
|

|
|

|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
21
|
22
|
23
|
24
|
25
|
26
|
27
|
|.........|................|................|................|................|................|................|................|
| CAL2
|
|
|
|
|
|
|+Ballgame/Famil+|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|
|
|+Birthday/Mary=+|+=====Close Sale/WYGIX Co.======+|
|
|
|
|+=============Inventors Show/Melvin==============+|+Planning Counc+|
|
|
|
|
|
|
|
|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
28
|
29
|
30
|
31
|
|
|
|
|.........|................|................|................|................|................|................|................|
| CAL2
|
|

|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
----------------------------------------------------------------------------------------------------------------------------------

122

Example 3: Multiple Schedule Calendars with Atypical Workshifts (Separated Output)

Chapter 5

Example 3: Multiple Schedule Calendars with Atypical Workshifts

(Separated Output)
Procedure features:

PROC CALENDAR statement options:

CALEDATA=
DATETIME
WORKDATA=
CALID statement:
_CAL_ variable
OUTPUT=SEPARATE option
DUR statement
OUTSTART statement
OUTFIN statement

This example
3 produces separate output pages for each calendar in a single PROC step

3 schedules activities around holidays

3 displays an 8-hour day, 5 1/2-day week
3 uses separate work patterns and holidays for each calendar.

Producing Different Output for Multiple Calendars

This example and Example 4 on page 127 use the same input data for multiple
calendars to produce different output. The only differences in these programs are how
the activities data set is sorted and how the OUTPUT= option is set.
To print

Sort the activities data set

And set
OUTPUT= to

See Example

Separate pages for each

calendar

calendar id and starting

date

SEPARATE

3, 8

All activities on the same

page and identify each
calendar

starting date

COMBINE

4, 2

All activities on the same

page and NOT identify each
calendar

starting date

MIX

Program

Specify a library so that you can permanently store the activities data set.
libname well SAS-data-library;

The CALENDAR Procedure

Program

123

Create the activities data set and identify separate calendars. WELL.ACT is a
permanent SAS data set that contains activities for a well construction project. The _CAL_
variable identies the calendar that an activity belongs to.
data well.act;
input task & $16. dur : 5. date : datetime16.
datalines;
Drill Well
3.50 01JUL96:12:00:00 CAL1
Lay Power Line
3.00 04JUL96:12:00:00 CAL1
Assemble Tank
4.00 05JUL96:08:00:00 CAL1
Build Pump House
3.00 08JUL96:12:00:00 CAL1
Pour Foundation
4.00 11JUL96:08:00:00 CAL1
Install Pump
4.00 15JUL96:14:00:00 CAL1
Install Pipe
2.00 19JUL96:08:00:00 CAL1
Erect Tower
6.00 20JUL96:08:00:00 CAL1
Deliver Material
2.00 01JUL96:12:00:00 CAL2
Excavate
4.75 03JUL96:08:00:00 CAL2
;

_cal_ $ cost;
1000
2000
1000
2000
1500
500
1000
2500
500
3500

Create the holidays data set. The _CAL_ variable identies the calendar that a holiday
belongs to.
data well.hol;
input date date. holiday $ 11-25 _cal_ $;
datalines;
09JUL96
Vacation
CAL2
04JUL96
Independence
CAL1
;

Create the calendar data set. Each observation denes the workshifts for an entire week.
The _CAL_ variable identies to which calendar the workshifts apply. CAL1 uses the default
8-hour workshifts for Monday through Friday. CAL2 uses a half day on Saturday and the
default 8-hour workshift for Monday through Friday.
data well.cal;
input _sun_ $
_fri_ $
datalines;
Holiday Holiday
Holiday Halfday
;

_sat_ $ _mon_ $ _tue_ $ _wed_ $ _thu_ $

_cal_ $;
Workday Workday Workday Workday Workday CAL1
Workday Workday Workday Workday Workday CAL2

Create the workdays data set. This data set denes the daily workshifts that are named in
the calendar data set. Each variable (not observation) contains one daily schedule of alternating
work and nonwork periods. The HALFDAY workshift lasts 4 hours.
data well.wor;
input halfday time5.;
datalines;

124

Program

Chapter 5

08:00
12:00
;

Sort the activities data set by the variables that contain the calendar identication
and the starting date, respectively. You are not required to sort the holidays data set.
proc sort data=well.act;
by _cal_ date;
run;

Set LINESIZE= appropriately. If the linesize is not long enough to print the variable values,
then PROC CALENDAR either truncates the values or produces no calendar output.
options nodate pageno=1 linesize=132 pagesize=60;

Create the schedule calendar. DATA= identies the activities data set; HOLIDATA=
identies the holidays data set; CALEDATA= identies the calendar data set; WORKDATA=
identies the workdays data set. DATETIME species that the variable specied with the
START statement contains values in SAS datetime format.
proc calendar data=well.act
holidata=well.hol
caledata=well.cal
workdata=well.wor
datetime;

Print each calendar on a separate page. The CALID statement species that the _CAL_
variable identies calendars. OUTPUT=SEPARATE prints information for each calendar on
separate pages.
calid _cal_ / output=separate;

Specify an activity start date variable and an activity duration variable. The START
statement species the variable in the activities data set that contains the activity starting
date; DUR species the variable that contains the activity duration. START and DUR are
required for a schedule calendar.
start date;
dur dur;

Retrieve holiday information. HOLISTART and HOLIVAR specify the variables in the
holidays data set that contain the start date and name of each holiday, respectively.
HOLISTART is required when you use a holidays data set.
holistart date;
holivar holiday;

The CALENDAR Procedure

Program

Customize the calendar appearance. OUTSTART and OUTFIN specify that the calendar
display a 6-day week, Monday through Saturday.
outstart Monday;
outfin Saturday;

Specify the title and format the Cost variable.

title1 Well Drilling Work Schedule: Separate Calendars;
format cost dollar9.2;
run;

125

126

Output

Chapter 5

Output

Output 5.6 Separate Output for Multiple Schedule Calendars

Well Drilling Work Schedule: Separate Calendars

............................................................ _cal_=CAL1 ............................................................

------------------------------------------------------------------------------------------------------------------------------|
|
|
July 1996
|
|
|
|-----------------------------------------------------------------------------------------------------------------------------|
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
1
|
2
|
3
|
4
|
5
|
6
|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|+====================Drill Well/$1,000.00====================>|

|
|
|

|
|
|

|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
8
|
9
|
10
|
11
|
12
|
13
|
|
|
|

|
|
|

|+===========================Build Pump House/$2,000.00============================+|

|
|<=============================Assemble Tank/$1,000.00=============================+|
|
|<=======Lay Power Line/$2,000.00========+|
|+=======Pour Foundation/$1,500.00=======>|

|
|
|

|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
15
|
16
|
17
|
18
|
19
|
20
|
|
|
|
|
|
|
|
|
|
|

|
|
|

|+=========================================Install Pump/$500.00=========================================+|
|<=================Pour Foundation/$1,500.00==================+|
|+Install Pipe/$1,00>|

|
|

|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
22
|
23
|
24
|
25
|
26
|
27
|
|
|
|
|
|
|
|
|
|
|

|
|
|

|+========================================Erect Tower/$2,500.00=========================================>|
|
|<========Install Pipe/$1,000.00=========+|
|
|
|
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
|
|
|
|

|
|
|
|
|

|
|

|
|
|

|
|
|
|
|
|
|
|<Erect Tower/$2,500+|
|
|
|
|
|
-------------------------------------------------------------------------------------------------------------------------------

The CALENDAR Procedure

Example 4: Multiple Schedule Calendars with Atypical Workshifts (Combined and Mixed Output)

127

Well Drilling Work Schedule: Separate Calendars

............................................................ _cal_=CAL2 ............................................................

------------------------------------------------------------------------------------------------------------------------------|
|
|
July 1996
|
|
|
|-----------------------------------------------------------------------------------------------------------------------------|
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
|
|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
|
|+===============================Excavate/$3,500.00================================>|

|+==================Deliver Material/$500.00==================+|
|
|
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
8
|
9
|
10
|
11
|
12
|
13
|
|
|
|

|
|
|

|
|

|<Excavate/$3,500.00>|
|<Excavate/$3,500.00+|
|
|
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
15
|
16
|
17
|
18
|
19
|
20
|
|
|
|

|
|
|

|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
22
|
23
|
24
|
25
|
26
|
27
|
|
|
|

|
|
|

|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
29
|
30
|
31
|
|
|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

|
|

-------------------------------------------------------------------------------------------------------------------------------

Example 4: Multiple Schedule Calendars with Atypical Workshifts

(Combined and Mixed Output)
Procedure features:

PROC CALENDAR statement options:

CALEDATA=
DATETIME
WORKDATA=
CALID statement:
_CAL_ variable

128

Two Programs and Two Pieces of Output

Chapter 5

OUTPUT=COMBINE option
OUTPUT=MIXED option
DUR statement
OUTSTART statement
OUTFIN statement
Data sets:

WELL.ACT on page 123, WELL.HOL on page 123, WELL.CAL on page 123,

WEL.WOR on page 123.

This example
3 produces a schedule calendar
3 schedules activities around holidays

3
3
3
3

uses separate work patterns and holidays for each calendar

uses an 8-hour day, 5 1/2-day work week
displays and identies multiple calendars on each calendar page (combined output)
displays but does not identify multiple calendars on each calendar page (mixed
output).

Two Programs and Two Pieces of Output

This example creates both combined and mixed output. Producing combined or
mixed calendar output requires only one change to a PROC CALENDAR step: the
setting of the OUTPUT= option in the CALID statement. Combined output is produced
rst, then mixed output.

Producing Different Output for Multiple Calendars

This example and Example 3 on page 122 use the same input data for multiple
calendars to produce different output. The only differences in these programs are how
the activities data set is sorted and how the OUTPUT= option is set.
To print

Sort the activities data set

And set
OUTPUT= to

See Example

Separate pages for each

calendar

calendar id and starting

date

SEPARATE

3, 8

All activities on the same

page and identify each
calendar

starting date

COMBINE

4, 2

All activities on the same

page and NOT identify each
calendar

starting date

MIX

The CALENDAR Procedure

Program for Combined Calendars

129

Program for Combined Calendars

Specify the SAS data library where the activities data set is stored.
libname well SAS-data-library;

Sort the activities data set by the variable that contains the starting date. Do not sort
by the CALID variable when producing combined calendar output.
proc sort data=well.act;
by date;
run;

Set PAGESIZE= and LINESIZE= appropriately. When you combine calendars, check the
value of PAGESIZE= to ensure that there is enough room to print the activities from multiple
calendars. If LINESIZE= is too small for the variable values to print, then PROC CALENDAR
either truncates the values or produces no calendar output.
options nodate pageno=1 linesize=132 pagesize=60;

Combine all events and holidays on a single calendar. The CALID statement species
that the _CAL_ variable identies the calendars. OUTPUT=COMBINE prints multiple
calendars on the same page and identies each calendar.
calid _cal_ / output=combine;

Specify an activity start date variable and an activity duration variable. The START
statement species the variable in the activities data set that contains the starting date of the
activities; DUR species the variable that contains the duration of each activity. START and
DUR are required for a schedule calendar.
start date;
dur dur;

130

Program for Combined Calendars

Chapter 5

Specify the title and format the Cost variable.

title1 Well Drilling Work Schedule: Combined Calendars;
format cost dollar9.2;
run;

The CALENDAR Procedure

Output for Combined Calendars

131

Output for Combined Calendars

Output 5.7 Multiple Schedule Calendars with Atypical Workshifts (Combined Output)
Well Drilling Work Schedule: Combined Calendars

-----------------------------------------------------------------------------------------------------------------------|
|
|
July 1996
|
|
|
|----------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
----------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
|
1
|
2
|
3
|
4
|
5
|
6
|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|
|
|
|**Independence**|+Assemble Tank/>|
|
|
|
|
|
|
|
|+Lay Power Line>|
|
|
|
|+==============Drill Well/$1,000.00==============>|
|<Drill Well/$1,+|
|
|.........|................|................|................|................|................|................|................|
| CAL2
|
|
|
|+=======================Excavate/$3,500.00========================>|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
7
|
8
|
9
|
10
|
11
|
12
|
13
|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|+===================Build Pump House/$2,000.00====================+|
|
|
|
|
|<=====================Assemble Tank/$1,000.00=====================+|
|
|
|
|
|<===Lay Power Line/$2,000.00====+|
|+===Pour Foundation/$1,500.00===>|
|
|.........|................|................|................|................|................|................|................|
| CAL2
|
|<Excavate/$3,50>|****Vacation****|<Excavate/$3,50+|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
14
|
15
|
16
|
17
|
18
|
19
|
20
|
|.........|................|................|................|................|................|................|................|
| CAL1
|
|+===============================Install Pump/$500.00===============================+|
|
|
|
|<===========Pour Foundation/$1,500.00============+|
|+Install Pipe/$>|
|
|
|
|

|
|
|

|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
28
|
29
|
30
|
31
|
|
|
|
|.........|................|................|................|................|................|................|................|
| CAL1
|

|
|

|<Erect Tower/$2+|
|
|

|
|

|
|
|

----------------------------------------------------------------------------------------------------------------------------------

132

Program for Mixed Calendars

Chapter 5

Program for Mixed Calendars

To produce mixed output instead of combined, use the same program and change the
setting of the OUTPUT= option to OUTPUT=MIX:
proc calendar data=well.act
holidata=well.hol
caledata=well.cal
workdata=well.wor
datetime;
calid _cal_ / output=mix;
start date;
dur dur;
holistart date;
holivar holiday;
outstart Monday;
outfin Saturday;
title1 Well Drilling Work Schedule: Mixed Calendars;
format cost dollar9.2;
run;

The CALENDAR Procedure

Output for Mixed Calendars

133

Output for Mixed Calendars

Output 5.8 Multiple Schedule Calendar with Atypical Workshifts (Mixed Output)
Well Drilling Work Schedule: Mixed Calendars

------------------------------------------------------------------------------------------------------------------------------|
|
|
July 1996
|
|
|
|-----------------------------------------------------------------------------------------------------------------------------|
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
|
|

|
|
|

|
|
|
|
|+Assemble Tank/$1,0>|
|
|
|
|+===============================Excavate/$3,500.00================================>|
|+==================Deliver Material/$500.00==================+|****Independence****|+Lay Power Line/$2,>|
|
|+====================Drill Well/$1,000.00====================>|****Independence****|<Drill Well/$1,000.+|
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
8
|
9
|
10
|
11
|
12
|
13
|
|
|

|
|

|+===========================Build Pump House/$2,000.00============================+|

|<=============================Assemble Tank/$1,000.00=============================+|
|<=======Lay Power Line/$2,000.00========+|
|
|

|
|
|

|<Excavate/$3,500.00>|******Vacation******|<Excavate/$3,500.00+|+=======Pour Foundation/$1,500.00=======>|
|
|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
15
|
16
|
17
|
18
|
19
|
20
|
|
|
|

|
|
|

|
|
|
|
|
|
|+=========================================Install Pump/$500.00=========================================+|
|<=================Pour Foundation/$1,500.00==================+|
|+Install Pipe/$1,00>|

|
|
|

|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
22
|
23
|
24
|
25
|
26
|
27
|
|
|
|

|
|
|

|
|
|
|
|
|
|+========================================Erect Tower/$2,500.00=========================================>|
|<========Install Pipe/$1,000.00=========+|
|
|
|

|
|
|

|--------------------+--------------------+--------------------+--------------------+--------------------+--------------------|
|
29
|
30
|
31
|
|
|
|
|
|
|
|
|
|
|
|
|

|
|

|
|
|

-------------------------------------------------------------------------------------------------------------------------------

134

Example 5: Schedule Calendar, Blank or with Holidays

Chapter 5

Example 5: Schedule Calendar, Blank or with Holidays

Procedure features:

PROC CALENDAR statement options:

FILL
HOLIDATA=
INTERVAL=WORKDAY
DUR statement
HOLIDUR statement
HOLISTART statement
HOLIVAR statement

This example produces a schedule calendar that displays only holidays. You can use
this same code to produce a set of blank calendars by removing the HOLIDATA= option
and the HOLISTART, HOLIVAR, and HOLIDUR statements from the PROC
CALENDAR step.

Program

Create the activities data set. Specify one activity in the rst month and one in the last, each
with a duration of 0. PROC CALENDAR does not print activities with zero durations in the
output.
data acts;
input sta : date7. act $ 11-30 dur;
datalines;
01JAN97
Start
0
31DEC97
Finish
0
;

Create the holidays data set.

data holidays;
input sta : date7. act $ 11-30 dur;
datalines;
01JAN97
New Years
1
28MAR97
Good Friday
1
30MAY97
Memorial Day
1
04JUL97
Independence Day
1
01SEP97
Labor Day
1
27NOV97
Thanksgiving
2
25DEC97
Christmas Break
5
;

The CALENDAR Procedure

Program

135

Set PAGESIZE= and LINESIZE= appropriately. To create larger boxes for each day in the
calendar output, increase the value of PAGESIZE=.
options nodate pageno=1 linesize=132 pagesize=30;

Create the calendar. DATA= identies the activities data set; HOLIDATA= identies the
holidays data set. FILL displays all months, even those with no activities. By default, only
months with activities appear in the report. INTERVAL=WORKDAY species that activities and
holidays are measured in 8-hour days and that PROC CALENDAR schedules activities only
Monday through Friday.
proc calendar data=acts holidata=holidays fill interval=workday;

Retrieve holiday information. The HOLISTART, HOLIVAR, and HOLIDUR statements

specify the variables in the holidays data set that contain the start date, name, and duration of
each holiday, respectively. When you use a holidays data set, HOLISTART is required. Because
at least one holiday lasts more than one day, HOLIDUR (or HOLIFIN) is required.
holistart sta;
holivar act;
holidur dur;

Specify the title.

title1 Calendar of Holidays Only;
run;

136

Output

Chapter 5

Output

Output 5.9 Schedule Calendars with Holidays Only (Partial Output).

Without INTERVAL=WORKDAY, the 5-day Christmas break would be scheduled through the weekend.

Calendar of Holidays Only

------------------------------------------------------------------------------------------------------------------------------|
|

January

1997

|
|

|
|
|-----------------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
|
|
|
1
|
2
|
3
|
4
|
|
|
|
|***New Years****|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
5
|
6
|
7
|
8
|
9
|
10
|
11
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
12
|
13
|
14
|
15
|
16
|
17
|
18
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
19
|
20
|
21
|
22
|
23
|
24
|
25
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
26
|
27
|
28
|
29
|
30
|
31
|
|
|
|
|
|
|
|
|
|
-------------------------------------------------------------------------------------------------------------------------------

Calendar of Holidays Only

------------------------------------------------------------------------------------------------------------------------------|
|
|
February 1997
|
|
|
|-----------------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
|
|
|
|
|
|
1
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
2
|
3
|
4
|
5
|
6
|
7
|
8
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
9
|
10
|
11
|
12
|
13
|
14
|
15
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
16
|
17
|
18
|
19
|
20
|
21
|
22
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
23
|
24
|
25
|
26
|
27
|
28
|
|
|
|
|
|
|
|
|
|
-------------------------------------------------------------------------------------------------------------------------------

The CALENDAR Procedure

Automating Your Scheduling Task with SAS/OR Software

137

Calendar of Holidays Only

------------------------------------------------------------------------------------------------------------------------------|
|
|

December

|
|
|

1997

|-----------------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
|

|
|

|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
7
|
8
|
9
|
10
|
11
|
12
|
13
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
14
|
15
|
16
|
17
|
18
|
19
|
20
|
|
|
|
|
|
|
|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
21
|
22
|
23
|
24
|
25
|
26
|
27
|
|
|
|
|
|*Christmas Break*|*Christmas Break*|
|
|-----------------+-----------------+-----------------+-----------------+-----------------+-----------------+-----------------|
|
28
|
29
|
30
|
31
|
|
|
|
|
|*Christmas Break*|*Christmas Break*|*Christmas Break*|
|
|
|
-------------------------------------------------------------------------------------------------------------------------------

Example 6: Calculating a Schedule Based on Completion of Predecessor

Tasks
Procedure features:

PROC CALENDAR statement

CALID statement
FIN statement
VAR statement
Other features:

PROC CPM step

PROC SORT step

Automating Your Scheduling Task with SAS/OR Software

When changes occur to a schedule, you have to adjust the activity starting dates
manually if you use PROC CALENDAR to produce a schedule calendar. Alternatively,
you can use PROC CPM in SAS/OR software to reschedule work when dates change.
Even more important, you can provide only an initial starting date for a project and let
PROC CPM calculate starting dates for activities, based on identied successor tasks,
that is, tasks that cannot begin until their predecessors end.
In order to use PROC CPM, you must
1 create an activities data set that contains activities with durations. (You can

indicate nonwork days, weekly work schedules, and workshifts with holidays,
calendar, and workshift data sets.)
2 indicate which activities are successors to others (precedence relationships).
3 dene resource limitations if you want them considered in the schedule.
4 provide an initial starting date.

138

Highlights of This Example

Chapter 5

PROC CPM can process your data to generate a data set that contains the start and
end dates for each activity. PROC CPM schedules the activities, based on the duration
information, weekly work patterns, workshifts, as well as holidays and nonwork days
that interrupt the schedule. You can generate several views of the schedule that is
computed by PROC CPM, from a simple listing of start and nish dates to a calendar, a
Gantt chart, or a network diagram.

Highlights of This Example

This example

3 calculates a project schedule containing multiple calendars (PROC CPM)

3 produces a listing of the PROC CPM output data set (PROC PRINT)
3 displays the schedule in calendar format (PROC CALENDAR).
This example features PROC CPMs ability to calculate a schedule that

3 is based on an initial starting date

3 applies different non-work periods to different calendars, such as personal
vacation days to each employees schedule

3 includes milestones (activities with a duration of 0).

See Also
This example introduces users of PROC CALENDAR to more advanced SAS
scheduling tools. For an introduction to project management tasks and tools and
several examples, see Project Management Using the SAS System. For more examples,
see SAS/OR Software: Project Management Examples. For complete reference
documentation, see SAS/OR Users Guide: Project Management.

Program

Set appropriate options. If the linesize is not long enough to print the variable values, then
PROC CALENDAR either truncates the values or produces no calendar output. A longer
linesize also makes it easier to view a listing of a PROC CPM output data set.
options nodate pageno=1 linesize=132 pagesize=60;

Create the activities data set and identify separate calendars. This data identies two
calendars: the professors (the value of _CAL_ is Prof.) and the students (the value of _CAL_ is
Student). The Succ1 variable identies which activity cannot begin until the current one ends.
For example Analyze Exp 1 cannot begin until Run Exp 1 is completed. The DAYS value of 0
for JOBNUM 3, 6, and 8 indicates that these are milestones.
data grant;
input jobnum Task $ 4-22 Days Succ1 $ 27-45 aldate : date7. altype $
_cal_ $;
format aldate date7.;
datalines;
1 Run Exp 1
11 Analyze Exp 1
.
.
Student
2 Analyze Exp 1
5 Send Report 1
.
.
Prof.
3 Send Report 1
0 Run Exp 2
.
.
Prof.

The CALENDAR Procedure

4
5
6
7
8
9
;

Run Exp 2
11
Analyze Exp 2
4
Send Report 2
0
Write Final Report 4
Send Final Report
0
Site Visit
1

Analyze Exp 2
Send Report 2
Write Final Report
Send Final Report

.
.
.
.
.
18jul96

.
.
.
.
.
ms

Program

139

Student
Prof.
Prof.
Prof.
Student
Prof.

Create the holidays data set and identify which calendar a nonwork day belongs to.
The two holidays are listed twice, once for the professors calendar and once for the students.
Because each person is associated with a separate calendar, PROC CPM can apply the personal
vacation days to the appropriate calendars.
data nowork;
format holista date7. holifin date7.;
input holista : date7. holifin : date7. name $ 17-32 _cal_ $;
datalines;
04jul96 04jul96 Independence Day Prof.
02sep96 02sep96 Labor Day
Prof.
04jul96 04jul96 Independence Day Student
02sep96 02sep96 Labor Day
Student
15jul96 16jul96 PROF Vacation
Prof.
15aug96 16aug96 STUDENT Vacation Student
;

Calculate the schedule with PROC CPM. PROC CPM uses information supplied in the
activities and holidays data sets to calculate start and nish dates for each activity. The DATE=
option supplies the starting date of the project. The CALID statement is not required, even
though this example includes two calendars, because the calendar identication variable has the
special name _CAL_.
proc cpm data=grant
date=01jul96d
interval=weekday
out=gcpm1
holidata=nowork;
activity task;
successor succ1;
duration days;
calid _cal_;
id task;
aligndate aldate;
aligntype altype;
holiday holista / holifin=holifin;
run;

140

Program

Chapter 5

Print the output data set that was created with PROC CPM. This step is not required.
PROC PRINT is a useful way to view the calculations produced by PROC CPM. See Output 5.10.
proc print data=gcpm1;
title Data Set GCPM1, Created with PROC CPM;
run;

Sort GCPM1 by the variable that contains the activity start dates before using it with
PROC CALENDAR.
proc sort data=gcpm1;
by e_start;
run;

Create the schedule calendar. GCPM1 is the activity data set. PROC CALENDAR uses the
S_START and S_FINISH dates, calculated by PROC CPM, to print the schedule. The VAR
statement selects only the variable TASK to display on the calendar output. See Output 5.11.
proc calendar data=gcpm1
holidata=nowork
interval=workday;
start e_start;
fin
e_finish;
calid _cal_ / output=combine;
holistart holista;
holifin
holifin;
holivar name;
var task;
title Schedule for Experiment X-15;
title2 Professor and Student Schedule;
run;

The CALENDAR Procedure

Output

141

Output

Output 5.10

The Data Set GCPM1

PROC PRINT displays the observations in GCPM1, showing the scheduling calculations created by PROC CPM.

Data Set GCPM1, Created with PROC CPM

Obs

Task

Succ1

Days

_cal_

E_START

E_FINISH

L_START

L_FINISH

T_FLOAT

F_FLOAT

Run Exp 1

Analyze Exp 1

Student

01JUL96

16JUL96

01JUL96

16JUL96

2
3
4

Analyze Exp 1
Send Report 1
Run Exp 2

Send Report 1
Run Exp 2
Analyze Exp 2

5
0
11

Prof.
Prof.
Student

17JUL96
24JUL96
24JUL96

23JUL96
24JUL96
07AUG96

17JUL96
24JUL96
24JUL96

23JUL96
24JUL96
07AUG96

0
0
0

5
6
7

Analyze Exp 2
Send Report 2
Write Final Report

Send Report 2
Write Final Report
Send Final Report

4
0
4

Prof.
Prof.
Prof.

08AUG96
14AUG96
14AUG96

13AUG96
14AUG96
19AUG96

08AUG96
14AUG96
14AUG96

13AUG96
14AUG96
19AUG96

0
0
0

8
9

Send Final Report

Site Visit

0
1

Student
Prof.

20AUG96
18JUL96

0
0

142

Output

Output 5.11

Chapter 5

Schedule Calendar Based on Output from PROC CPM

PROC CALENDAR created this schedule calendar by using the S_START and S_FINISH dates that were
calculated by PROC CPM. The activities on July 24th and August 14th, because they are milestones, do not
delay the start of a successor activity. Note that Site Visit occurs on July 18, the same day that Analyze Exp 1
occurs. To prevent this overallocation of resources, you can use resource constrained scheduling, available
in SAS/OR software.

Schedule for Experiment X-15

Professor and Student Schedule

-----------------------------------------------------------------------------------------------------------------------|
|

July

|
|

1996

|
|
|----------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
----------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
|
1
|
2
|
3
|
4
|
5
|
6
|
|.........|................|................|................|................|................|................|................|
| PROF.
|
|
|
|
|Independence Day|
|
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|+===================Run Exp 1====================>|Independence Day|<==Run Exp 1===>|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
7
|
8
|
9
|
10
|
11
|
12
|
13
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|
|
|
|

|<====================================Run Exp 1=====================================>|

|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
14
|
15
|
16
|
17
|
18
|
19
|
20
|
|.........|................|................|................|................|................|................|................|
| PROF.
|
|*PROF Vacation**|*PROF Vacation**|
|+==Site Visit==+|
|
|
|
|
|
|
|+=================Analyze Exp 1==================>|
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|<===========Run Exp 1===========+|
|
|
|
|
|
|
|
|
|
|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
21
|
22
|
23
|
24
|
25
|
26
|
27
|
|.........|................|................|................|................|................|................|................|
| PROF.
|
|<=========Analyze Exp 1=========+|+Send Report 1=+|
|
|
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|
|
|+===================Run Exp 2====================>|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
28
|
29
|
30
|
31
|
|
|
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|<===================Run Exp 2====================>|
|
|
|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
----------------------------------------------------------------------------------------------------------------------------------

The CALENDAR Procedure

Example 7: Summary Calendar with MEAN Values By Observation

143

Schedule for Experiment X-15

Professor and Student Schedule

-----------------------------------------------------------------------------------------------------------------------|
|
|
August 1996
|
|
|
|----------------------------------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
----------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
|
|
|
|
1
|
2
|
3
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|
|
|
|<===========Run Exp 2===========>|
|
|
|
|
|
|
|
|
|
|
|
|
|

|
|
|

|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
4
|
5
|
6
|
7
|
8
|
9
|
10
|
|.........|................|................|................|................|................|................|................|
| PROF.
|
|
|
|
|+=========Analyze Exp 2=========>|
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|<===================Run Exp 2====================+|
|
|
|
|
|

|
|

|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
11
|
12
|
13
|
14
|
15
|
16
|
17
|
|.........|................|................|................|................|................|................|................|
| PROF.
|
|
|
|+===============Write Final Report===============>|
|
|
|
|<=========Analyze Exp 2=========+|+Send Report 2=+|
|
|
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|
|
|
|STUDENT Vacation|STUDENT Vacation|
|
|
|
|
|
|
|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
18
|
19
|
20
|
21
|
22
|
23
|
24
|
|.........|................|................|................|................|................|................|................|
| PROF.
|
|<Write Final Re+|
|
|
|
|
|
|.........|................|................|................|................|................|................|................|
| STUDENT |
|
|+Send Final Rep+|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|---------+----------------+----------------+----------------+----------------+----------------+----------------+----------------|
|
|
|

|
|
|

|
|
|
|
|
|

|
|
|

|
|
|
|
|
|

|
|
|
|
|
|
|
|
|
----------------------------------------------------------------------------------------------------------------------------------

Example 7: Summary Calendar with MEAN Values By Observation

Procedure features:

CALID statement:
_CAL_ variable
OUTPUT=SEPARATE option
FORMAT statement
LABEL statement

144

MEAN Values by Number of Days

Chapter 5

MEAN statement
SUM statement
Other features:

PROC FORMAT:
PICTURE statement

This example
3 produces a summary calendar
3 displays holidays
3 produces sum and mean values by business day (observation) for three variables
3 prints a legend and uses variable labels
3 uses picture formats to display values.

MEAN Values by Number of Days

To produce MEAN values based on the number of days in the calendar month, use
MEANTYPE=NDAYS. By default, MEANTYPE=NOBS, which calculates the MEAN
values according to the number of days for which data exists.

Program

Create the activities data set. MEALS records how many meals were served for breakfast,
lunch, and dinner on the days that the cafeteria was open for business.
data meals;
input date
datalines;
02Dec96
03Dec96
04Dec96
05Dec96
06Dec96
09Dec96
10Dec96
11Dec96
12Dec96
13Dec96
16Dec96
17Dec96
18Dec96
19Dec96
20Dec96
23Dec96
;

: date7. Brkfst Lunch Dinner;

123
188
123
200
176
178
165
187
176
187
176
156
198
178
165
187

234
188
183
267
165
198
176
176
187
187
165
.
143
198
176
187

238
198
176
243
177
187
187
231
222
123
177
167
167
187
187
123

Create the holidays data set.

data closed;
input date date. holiday $ 11-25;

The CALENDAR Procedure

Program

145

datalines;
26DEC96
Repairs
27DEC96
Repairs
30DEC96
Repairs
31DEC96
Repairs
24DEC96
Christmas Eve
25DEC96
Christmas
;

Sort the activities data set by the activity starting date. You are not required to sort the
holidays data set.
proc sort data=meals;
by date;
run;

Create picture formats for the variables that indicate how many meals were served.
proc format;
picture bfmt other = 000 Brkfst;
picture lfmt other = 000 Lunch ;
picture dfmt other = 000 Dinner;
run;

Set PAGESIZE= and LINESIZE= appropriately. The legend box prints on the next page if
PAGESIZE= is not set large enough. LINESIZE= controls the width of the cells in the calendar.
options nodate pageno=1 linesize=132 pagesize=60;

Create the summary calendar. DATA= identies the activities data set; HOLIDATA=
identies the holidays data set. The START statement species the variable in the activities
data set that contains the activity starting date; START is required.
proc calendar data=meals holidata=closed;
start date;

Retrieve holiday information. The HOLISTART and HOLIVAR statements specify the
variables in the holidays data set that contain the start date and the name of each holiday,
respectively. HOLISTART is required when you use a holidays data set.
holistart date;
holiname holiday;

146

Program

Chapter 5

Calculate, label, and format the sum and mean values. The SUM and MEAN statements
calculate sum and mean values for three variables and print them with the specied format.
The LABEL statement prints a legend and uses labels instead of variable names. The FORMAT
statement associates picture formats with three variables.
sum brkfst lunch dinner / format=4.0;
mean brkfst lunch dinner / format=6.2;
label brkfst = Breakfasts Served
lunch =
Lunches Served
dinner =
Dinners Served;
format brkfst bfmt.
lunch lfmt.
dinner dfmt.;

Specify the titles.

title Meals Served in Company Cafeteria;
title2 Mean Number by Business Day;
run;

The CALENDAR Procedure

Example 8: Multiple Summary Calendars with Atypical Workshifts (Separated Output)

147

Output

Output 5.12

Summary Calendar with MEAN Values by Observation

Meals Served in Company Cafeteria
Mean Number by Business Day

-------------------------------------------------------------------------------------------|
|
|
December 1996
|
|
|
|------------------------------------------------------------------------------------------|
|
Sunday
|
Monday
| Tuesday
| Wednesday | Thursday |
Friday
| Saturday |
|------------+------------+------------+------------+------------+------------+------------|
|
1
|
2
|
3
|
4
|
5
|
6
|
7
|
|
|

|
|

|
| 234 Lunch | 188 Lunch | 183 Lunch | 267 Lunch | 165 Lunch |
|
|
| 238 Dinner | 198 Dinner | 176 Dinner | 243 Dinner | 177 Dinner |
|
|------------+------------+------------+------------+------------+------------+------------|
|
|
|

|
9
|
10
|
11
|
12
|
13
|
|
|
|
|
|
|
| 178 Brkfst | 165 Brkfst | 187 Brkfst | 176 Brkfst | 187 Brkfst |

|
|
|

|
| 198 Lunch | 176 Lunch | 176 Lunch | 187 Lunch | 187 Lunch |
|
|
| 187 Dinner | 187 Dinner | 231 Dinner | 222 Dinner | 123 Dinner |
|
|------------+------------+------------+------------+------------+------------+------------|
|
|
|
|
|

|
|
|
|
|

|------------+------------+------------+------------+------------+------------+------------|
|
22
|
23
|
24
|
25
|
26
|
27
|
28
|
|
|
|Christmas Ev|*Christmas**|**Repairs***|**Repairs***|
|
|
|
|

|
|
|

|------------+------------+------------+------------+------------+------------+------------|
|
29
|
30
|
31
|
|
|
|
|
|
|**Repairs***|**Repairs***|
|
|
|
|
|
|

|
|

|
|
|
|
|
|
|
|
-------------------------------------------------------------------------------------------------------------------------------|
| Sum | Mean |
|
|
|
|
| Breakfasts Served | 2763 | 172.69 |
|
Lunches Served | 2830 | 188.67 |
|
Dinners Served | 2990 | 186.88 |
-------------------------------------

Example 8: Multiple Summary Calendars with Atypical Workshifts

(Separated Output)
Procedure features:

PROC CALENDAR statement options:

DATETIME

148

Producing Different Output for Multiple Calendars

Chapter 5

LEGEND
CALID statement:
_CAL_ variable
OUTPUT=SEPARATE option
OUTSTART statement
OUTFIN statement
SUM statement
Data sets:

WELL.ACT on page 123 and WELL.HOL on page 123.

This example

3
3
3
3

produces a summary calendar for multiple calendars in a single PROC step

prints the calendars on separate pages
displays holidays
uses separate work patterns, work shifts, and holidays for each calendar

Producing Different Output for Multiple Calendars

This example produces separate output for multiple calendars. To produce combined
or mixed output for this data, you need to change only two things:

3 how the activities data set is sorted

3 how the OUTPUT= option is set.
To print

Sort the activities data set

And set
OUTPUT= to

See Example

Separate pages for each

calendar

calendar id and starting

date

SEPARATE

3, 8

All activities on the same

page and identify each
calendar

starting date

COMBINE

4, 2

All activities on the same

page and NOT identify each
calendar

starting date

MIX

Program

Specify the SAS data library where the activities data set is stored.
libname well SAS-data-library;
run;

The CALENDAR Procedure

Program

149

Sort the activities data set by the variables containing the calendar identication and
the starting date, respectively.
proc sort data=well.act;
by _cal_ date;
run;

Set PAGESIZE= and LINESIZE= appropriately. The legend box prints on the next page if
PAGESIZE= is not set large enough. LINESIZE= controls the width of the boxes.
options nodate pageno=1 linesize=132 pagesize=60;

Create the summary calendar. DATA= identies the activities data set; HOLIDATA=
identies the holidays data set; CALDATA= identies the calendar data set; WORKDATA=
identies the workdays data set. DATETIME species that the variable specied with the
START statement contains a SAS datetime value. LEGEND prints text that identies the
variables.
proc calendar data=well.act
holidata=well.hol
datetime legend;

Specify an activity start date variable and retrieve holiday information. The START
statement species the variable in the activities data set that contains the activity starting
date. The HOLISTART and HOLIVAR statements specify the variables in the holidays data set
that contain the start date and name of each holiday, respectively. These statements are
required when you use a holidays data set.
start date;
holistart date;
holivar holiday;

Calculate sum values. The SUM statement totals the COST variable for all observations in
each calendar.
sum cost / format=dollar10.2;

Display a 6-day week. OUTSTART and OUTFIN specify that the calendar display a 6-day
week, Monday through Saturday.
outstart Monday;
outfin Saturday;

150

Output

Chapter 5

Specify the titles and format the Cost variable.

title Well Drilling Cost Summary;
title2 Separate Calendars;
format cost dollar10.2;
run;

Output

Output 5.13

Separated Output for Multiple Summary Calendars

Well Drilling Cost Summary
Separate Calendars

............................................................ _cal_=CAL1 ............................................................

------------------------------------------------------------------------------------------------------------------|
|
|
July 1996
|
|
|
|-----------------------------------------------------------------------------------------------------------------|
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|------------------+------------------+------------------+------------------+------------------+------------------|
|
1
|
2
|
3
|
4
|
5
|
6
|
|
|
|
|***Independence***|
|
|
| Drill Well
|
|
3.5 |
|
$1,000.00 |

|
|
|

| Lay Power Line

| Assemble Tank
|
|
3 |
4 |
|
$2,000.00 |
$1,000.00 |

|
|
|

|------------------+------------------+------------------+------------------+------------------+------------------|
|
8
|
9
|
10
|
11
|
12
|
13
|
|
|
| Build Pump House |
|
3 |

|
|
|

|
|
| Pour Foundation |
|
4 |

|
|
|

|
$2,000.00 |
|
|
$1,500.00 |
|
|
|------------------+------------------+------------------+------------------+------------------+------------------|
|
15
|
16
|
17
|
18
|
19
|
20
|
|
| Install Pump
|

|
|
4 |

|
|
|

|
| Install Pipe
|

|
| Erect Tower
2 |

|
|
6 |

|
$500.00 |
|
|
|
$1,000.00 |
$2,500.00 |
|------------------+------------------+------------------+------------------+------------------+------------------|
|
22
|
23
|
24
|
25
|
26
|
27
|
|
|

|
|

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|------------------+------------------+------------------+------------------+------------------+------------------|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
|
|
|
|
|
------------------------------------------------------------------------------------------------------------------------------------------| Legend
|
| task

|
|
|

Sum

|
|
|

| dur
|
|
| cost
| $11,500.00 |
-------------------------

The CALENDAR Procedure

Output

Well Drilling Cost Summary

Separate Calendars

151

............................................................ _cal_=CAL2 ............................................................

------------------------------------------------------------------------------------------------------------------|
|
|

July

|
|
|

1996

|-----------------------------------------------------------------------------------------------------------------|
|
Monday
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
Saturday
|
|------------------+------------------+------------------+------------------+------------------+------------------|
|
1
|
2
|
3
|
4
|
5
|
6
|
|
|
|
|
|
|
|
| Deliver Material |
|
2 |
|
$500.00 |

| Excavate
|
|
4.75 |
|
$3,500.00 |

|
|
|

|------------------+------------------+------------------+------------------+------------------+------------------|
|
8
|
9
|
10
|
11
|
12
|
13
|
|
|*****Vacation*****|
|
|
|
|
|
|
|

|
|
|

|------------------+------------------+------------------+------------------+------------------+------------------|
|
15
|
16
|
17
|
18
|
19
|
20
|
|
|
|

|
|
|

|
|
|
|
|
|
|
|------------------+------------------+------------------+------------------+------------------+------------------|
|
22
|
23
|
24
|
25
|
26
|
27
|
|
|
|

|
|
|

|
|
|
|
|
|
|
|------------------+------------------+------------------+------------------+------------------+------------------|
|
|
|

|
|
|

|
|
|
|
|
|
|
|
|
|
|
|
|
|
------------------------------------------------------------------------------------------------------------------------------------------| Legend |
Sum
|
|
| task
| dur

|
|
|

| cost
| $4,000.00 |
-------------------------

152

153

CHAPTER

6
The CATALOG Procedure
Overview: CATALOG Procedure 153
Syntax: CATALOG Procedure 154
PROC CATALOG Statement 155
CHANGE Statement 157
CONTENTS Statement 157
COPY Statement 158
DELETE Statement 160
EXCHANGE Statement 161
EXCLUDE Statement 162
MODIFY Statement 163
SAVE Statement 164
SELECT Statement 165
Concepts: CATALOG Procedure 165
Interactive Processing with RUN Groups 165
Denition 165
How to End a PROC CATALOG Step 166
Error Handling and RUN Groups 166
Specifying an Entry Type 166
Four Ways to Supply an Entry Type 166
Why Use the ENTRYTYPE= Option? 167
Avoid a Common Error 167
The ENTRYTYPE= Option 167
Catalog Concatenation 168
Restrictions 168
Results: CATALOG Procedure 169
Examples: CATALOG Procedure 170
Example 1: Copying, Deleting, and Moving Catalog Entries from Multiple Catalogs 170
Example 2: Displaying Contents, Changing Names, and Changing a Description 174
Example 3: Using the FORCE Option with the KILL Option 176

Overview: CATALOG Procedure

The CATALOG procedure manages entries in SAS catalogs. PROC CATALOG is an
interactive, statement-driven procedure that enables you to
3 create a listing of the contents of a catalog
3 copy a catalog or selected entries within a catalog
3 rename, exchange, or delete entries within a catalog
3 change the name of a catalog entry
3 modify, by changing or deleting, the description of a catalog entry.

154

Syntax: CATALOG Procedure

Chapter 6

For more information on SAS data libraries and catalogs, refer to SAS Language
Reference: Concepts.
To learn how to use the SAS windowing environment to manage entries in a SAS
catalog, see the SAS online Help for the SAS Explorer window. You may prefer to use
the Explorer window instead of using PROC CATALOG. The window can do most of
what the procedure does.

Syntax: CATALOG Procedure

Tip:

Supports RUN-group processing.

Supports the Output Delivery System. See Output Delivery System on page 32
for details.

Tip:

ODS Table Name:

See: Results: CATALOG Procedure on page 169

You can perform similar functions with the SAS Explorer window and with
dictionary tables in the SQL procedure. For information on the Explorer window, see
the online Help. For information on PROC SQL, see Chapter 44, The SQL Procedure,
on page 1027.

Reminder:

See:

CATALOG Procedure in the documentation for your operating environment.

PROC CATALOG CATALOG=<libref.>catalog <ENTRYTYPE=etype> <FORCE>

<KILL>;
CONTENTS <OUT=SAS-data-set> <FILE=leref>;
COPY OUT=<libref.>catalog <options>;
SELECT entry(s) </ ENTRYTYPE=etype>;
EXCLUDE entry(s) </ ENTRYTYPE=etype>;
CHANGE old-name-1=new-name-1
<old-name-n=new-name-n>
</ ENTRYTYPE=etype>;
EXCHANGE name-1=other-name-1
<name-n=other-name-n>
</ ENTRYTYPE=etype>;
DELETE entry(s) </ ENTRYTYPE=etype>;
MODIFY entry (DESCRIPTION=< <>entry-description< >>)< / ENTRYTYPE=etype>;
SAVE entry(s) </ ENTRYTYPE=etype>;

To do this

Use this statement

Copy entries from one SAS catalog to another

Copy or move all entries

COPY (with MOVE option)

Copy entries to a new catalog (overwriting the catalog

if it already exists)

COPY (with NEW option)

Copy only selected entries

COPY, SELECT

Copy all except the entries specied

COPY, EXCLUDE

Delete entries from a SAS catalog

The CATALOG Procedure

To do this

PROC CATALOG Statement

155

Use this statement

Delete all entries

PROC CATALOG (with KILL option)

Delete all entries in catalog opened by another

resource environment

PROC CATALOG (with FORCE and

KILL options)

Delete all except the entries specied

SAVE

Alter names and descriptions

Change the names of catalog entries

CHANGE

Switch the names of two catalog entries

EXCHANGE

Change the description of a catalog entry

MODIFY

Print the contents of a catalog

CONTENTS

PROC CATALOG Statement

PROC CATALOG CATALOG=<libref.>catalog <ENTRYTYPE=etype> <FORCE>
<KILL>;

To do this

Use this option

Restrict processing to one entry type

ENTRYTYPE=

Delete all catalog entries

KILL

Force certain statements to execute on a catalog opened by

another resource environment

FORCE

Required Arguments
CATALOG=<libref.>catalog

species the SAS catalog to process.

Alias:

CAT=, C=

Default: If ENTRYTYPE= is not specied, PROC CATALOG processes all entries in

the catalog.

156

PROC CATALOG Statement

Chapter 6

Options
ENTRYTYPE=etype

restricts processing of the current PROC CATALOG step to one entry type.
Alias:

ET=

Default: If you omit ENTRYTYPE=, PROC CATALOG processes all entries in a

catalog.
Interaction: The specied entry type applies to any one-level entry names used in a

subordinate statement. You cannot override this specication in a subordinate

statement.
Interaction: ENTRYTYPE= does not restrict the effects of the KILL option.

In order to process multiple entry types in a single PROC CATALOG step, use
ENTRYTYPE= in a subordinate statement, not in the PROC CATALOG statement.

Tip:

Example 1 on page 170 and Example 2 on page 174

FORCE

forces statements to execute on a catalog that is opened by another resource

environment.
Some CATALOG statements require exclusive access to the catalog that they
operate on if the statement can radically change the contents of a catalog. If
exclusive access cannot be obtained, then the action fails. The statements and the
catalogs that are affected by FORCE are
KILL

affects the specied catalog

COPY

affects the OUT= catalog

COPY MOVE

affects the IN= and the OUT= catalogs

SAVE

affects the specied catalog.

Use FORCE to execute the statement, even if exclusive access cannot be

obtained.

Tip:

Featured in:

Example 3 on page 176

KILL

deletes all entries in a SAS catalog.

Interaction: The KILL option deletes all catalog entries even when ENTRYTYPE=

is specied.
Interaction: The SAVE statement has no effect because the KILL option deletes all

entries in a SAS catalog before any other statements are processed.

KILL deletes all entries but does not remove an empty catalog from the SAS
data library. You must use another method, such as PROC DATASETS or the DIR
window to delete an empty SAS catalog.

Tip:

Featured in:

Example 3 on page 176

CAUTION:

Do not attempt to limit the effects of the KILL option. This option deletes all entries in a
SAS catalog before any option or other statement takes effect. 4

The CATALOG Procedure

CONTENTS Statement

157

CHANGE Statement
Renames one or more catalog entries.
You can change multiple names in a single CHANGE statement or use multiple
CHANGE statements.

Tip:

Featured in: Example 2 on page 174

CHANGE old-name-1=new-name-1
<old-name-n=new-name-n>
</ ENTRYTYPE=etype>;

Required Arguments
old-name=new-name

species the current name of a catalog entry and the new name you want to assign to
it. Specify any valid SAS name.
Restriction: You must designate the type of the entry, either with the name

(ename.etype) or with the ENTRYTYPE= option.

Options
ENTRYTYPE=etype

restricts processing to one entry type.

See:

The ENTRYTYPE= Option on page 167

CONTENTS <OUT=SAS-data-set> <FILE=leref>;

Without Options
The output is sent to the procedure output.

158

COPY Statement

Chapter 6

Options
Note: The ENTRYTYPE= (ET=) option is not available for the CONTENTS
statement. 4
CATALOG=<libref.>catalog

species the SAS catalog to process.

Alias: CAT=, C=
Default: None
FILE=leref

sends the contents to an external le, identied with a SAS leref.

Interaction: If leref has not been previously assigned to a le, then the le is
created and named according to operating environment-dependent rules for
external les.
OUT=SAS-data-set

sends the contents to a SAS data set. When the statement executes, a message on
the SAS log reports that a data set has been created. The data set contains six
variables in this order:
LIBNAME

the libref

MEMNAME

the catalog name

NAME

the names of entries

TYPE

the types of entries

DESC

the descriptions of entries

DATE

the dates entries were last modied.

COPY Statement
Copies some or all of the entries in one catalog to another catalog.
Restriction: A COPY statements effect ends at a RUN statement or at the beginning of a
statement other than the SELECT or EXCLUDE statement.
Tip: Use SELECT or EXCLUDE statements, but not both, after the COPY statement to
limit which entries are copied.
Tip: You can copy entries from multiple catalogs in a single PROC step, not just the one
specied in the PROC CATALOG statement.
Tip: The ENTRYTYPE= option does not require a forward slash (/) in this statement.
Featured in: Example 1 on page 170

COPY OUT=<libref.>catalog <options>;

The CATALOG Procedure

COPY Statement

To do this

Use this option

Restrict processing to one type of entry

ENTRYTYPE=

Copy from a different catalog in the same step

IN=

Move (copy and then delete) a catalog entry

MOVE

Copy entries to a new catalog (overwriting the catalog if it

already exists)

NEW

Protect several types of SAS/AF entries from being edited with

PROC BUILD

NOEDIT

Not copy source lines from a PROGRAM, FRAME, or SCL entry

159

NOSOURCE

Required Arguments
OUT=<libref.>catalog

names the catalog to which entries are copied.

Options
ENTRYTYPE=etype

restricts processing to one entry type for the current COPY statement and any
subsequent SELECT or EXCLUDE statements.
The ENTRYTYPE= Option on page 167
See also: Specifying an Entry Type on page 166
See:

IN=<libref.>catalog

species the catalog to copy.

Interaction: The IN= option overrides a CATALOG= argument that was specied

in the PROC CATALOG statement.

Featured in:

Example 1 on page 170

MOVE

deletes the original catalog or entries after the new copy is made.
Interaction: When MOVE removes all entries from a catalog, the procedure deletes

the catalog from the library.

NEW

overwrites the destination (specied by OUT=) if it already exists. If you omit NEW,
PROC CATALOG updates the destination. For information about using the NEW
option with concatenated catalogs, see Catalog Concatenation on page 168.

160

DELETE Statement

Chapter 6

NOEDIT

prevents the copied version of the following SAS/AF entry types from being edited by
the BUILD procedure:
CBT

PROGRAM

FRAME

SCL

HELP

SYSTEM

Restriction: If you specify the NOEDIT option for an entry that is not one of these

types, it is ignored.
When creating SAS/AF applications for other users, use NOEDIT to protect the
application by preventing certain catalog entries from being altered.

Tip:

Featured in:

Example 1 on page 170

NOSOURCE

omits copying the source lines when you copy a SAS/AF PROGRAM, FRAME, or SCL
entry.
Alias:

NOSRC

Restriction: If you specify this option for an entry other than a PROGRAM,

FRAME, or SCL entry, it is ignored.

DELETE Statement
Deletes entries from a SAS catalog.
Tip: Use DELETE to delete only a few entries; use SAVE when it is more convenient to
specify which entries not to delete.
Tip:

You can specify multiple entries. You can also use multiple DELETE statements.

Example 1 on page 170

DELETE entry(s) </ ENTRYTYPE=etype>;

Required Arguments
entry(s)

species the name of one or more SAS catalog entries.

Restriction: You must designate the type of the entry, either with the name

(ename.etype) or with the ENTRYTYPE= option.

The CATALOG Procedure

EXCHANGE Statement

161

Options
ENTRYTYPE=etype

restricts processing to one entry type.

See: The ENTRYTYPE= Option on page 167
See also: Specifying an Entry Type on page 166

EXCHANGE Statement
Switches the name of two catalog entries.
Restriction:

The catalog entries must be of the same type.

EXCHANGE name-1=other-name-1
<name-n=other-name-n>
</ ENTRYTYPE=etype>;

Required Arguments
name=other-name

species two catalog entry names that the procedure will switch.
Interaction: You can specify only the entry name without the entry type if you use
the ENTRYTYPE= option on either the PROC CATALOG statement or the
EXCHANGE statement.
See also: Specifying an Entry Type on page 166

Options
ENTRYTYPE=etype

restricts processing to one entry type.

See: The ENTRYTYPE= Option on page 167
See also: Specifying an Entry Type on page 166

162

EXCLUDE Statement

Chapter 6

EXCLUDE Statement
Species entries that the COPY statement does not copy.
Requires the COPY statement.
Restriction: Do not use the EXCLUDE statement with the SELECT statement.
Restriction:
Tip:

You can specify multiple entries in a single EXCLUDE statement.

You can use multiple EXCLUDE statements with a single COPY statement within
a RUN group.
See also: COPY Statement on page 158 and SELECT Statement on page 165
Tip:

Featured in:

Example 1 on page 170

EXCLUDE entry(s) </ ENTRYTYPE=etype>;

Required Arguments
entry(s)

species the name of one or more SAS catalog entries.

Restriction: You must designate the type of the entry, either when you specify the

name (ename.etype) or with the ENTRYTYPE= option.

See also: Specifying an Entry Type on page 166

Options
ENTRYTYPE=etype

restricts processing to one entry type.

See: The ENTRYTYPE= Option on page 167
See also: Specifying an Entry Type on page 166

The CATALOG Procedure

MODIFY Statement

163

MODIFY Statement
Changes the description of a catalog entry.
Featured in: Example 2 on page 174

MODIFY entry (DESCRIPTION=< <>entry-description< >>) </ ENTRYTYPE=etype>;

Required Arguments
entry

species the name of one SAS catalog entry. Optionally, you can specify the entry
type with the name.
Restriction: You must designate the type of the entry, either when you specify the

name (ename.etype) or with the ENTRYTYPE= option.

See also: Specifying an Entry Type on page 166
DESCRIPTION=<<>entry-description<>>

changes the description of a catalog entry by replacing it with a new description, up

to 256 characters long, or by removing it altogether. Optionally, you can enclose the
description in single or double quotes.
Alias:
Tip:

DESC
Use DESCRIPTION= with no text to remove the current description.

164

SAVE Statement

Chapter 6

Options
ENTRYTYPE=etype

restricts processing to one entry type.

See: The ENTRYTYPE= Option on page 167
See also: Specifying an Entry Type on page 166

SAVE Statement
Specify entries not to delete from a SAS catalog.
Restriction:

Cannot limit the effects of the KILL option.

Tip: Use SAVE to delete all but a few entries in a catalog. Use DELETE when it is
more convenient to specify which entries to delete.

You can specify multiple entries and use multiple SAVE statements.
See also: DELETE Statement on page 160
Tip:

SAVE entry(s) < / ENTRYTYPE=etype>;

Required Arguments
entry(s)

species the name of one or more SAS catalog entries.

Restriction: You must designate the type of the entry, either with the name
(ename.etype) or with the ENTRYTYPE= option.

Options
ENTRYTYPE=etype

restricts processing to one entry type.

See: The ENTRYTYPE= Option on page 167
See also: Specifying an Entry Type on page 166

The CATALOG Procedure

Interactive Processing with RUN Groups

165

SELECT Statement
Species entries that the COPY statement will copy.
Restriction:

Requires the COPY statement.

Cannot be used with an EXCLUDE statement.

Tip: You can specify multiple entries in a single SELECT statement.
Restriction:

You can use multiple SELECT statements with a single COPY statement within a
RUN group.
See also: COPY Statement on page 158 and EXCLUDE Statement on page 162
Tip:

Featured in: Example 1 on page 170

SELECT entry(s) </ ENTRYTYPE=etype>;

Required Arguments
entry(s)

species the name of one or more SAS catalog entries.

Restriction: You must designate the type of the entry, either when you specify the
name (ename.etype) or with the ENTRYTYPE= option.

Options
ENTRYTYPE=etype

restricts processing to one entry type.

See: The ENTRYTYPE= Option on page 167.
See also: Specifying an Entry Type on page 166.

Concepts: CATALOG Procedure

Interactive Processing with RUN Groups

Denition
The CATALOG procedure is interactive. Once you submit a PROC CATALOG
statement, you can continue to submit and execute statements or groups of statements
without repeating the PROC CATALOG statement.
A set of procedure statements ending with a RUN statement is called a RUN group.
The changes specied in a given group of statements take effect when a RUN statement
is encountered.

166

Specifying an Entry Type

Chapter 6

How to End a PROC CATALOG Step

In the DATA step and most SAS procedures, a RUN statement is a step boundary
and ends the step. A simple RUN statement does not, however, end an interactive
procedure. To terminate a PROC CATALOG step, you can

3 submit a QUIT statement

3 submit a RUN statement with the CANCEL option
3 submit another DATA or PROC statement
3 end your SAS session.

Note: When you enter a QUIT, DATA, or PROC statement, any statements following
the last RUN group execute before the CATALOG procedure terminates. If you enter a
RUN statement with the CANCEL option, however, the remaining statements do not
execute before the procedure ends. 4
See Example 2 on page 174.

Error Handling and RUN Groups

Error handling is based in part on the division of statements into RUN groups. If a
syntax error is encountered, none of the statements in the current RUN group execute,
and execution proceeds to the next RUN group.
For example, the following statements contain a misspelled DELETE statement:
proc catalog catalog=misc entrytype=help;
copy out=drink;
select coffee tea;
del juices;
/* INCORRECT!!! */
exchange glass=plastic;
run;
change calstats=nutri;
run;

Because the DELETE statement is incorrectly specied as DEL, no statements in

that RUN group execute, except the PROC CATALOG statement itself. The CHANGE
statement does execute, however, because it is in a different RUN group.
CAUTION:

Be careful when setting up batch jobs in which one RUN groups statements depend on the
effects of a previous RUN group, especially when deleting and renaming entries. 4

Specifying an Entry Type

Four Ways to Supply an Entry Type

There is no default entry type, so if you do not supply one, PROC CATALOG
generates an error. You can supply an entry type in one of four ways. See Table 6.1 on
page 167.

The CATALOG Procedure

Specifying an Entry Type

167

Table 6.1 Supplying an Entry Type

You can supply an entry
type with

Example

the entry name

delete
test1.program
test1.log test2.log;

ET= in parentheses

delete
test1 (et=program);

ET= after a slash1

delete test1 (et=program)

test1 test2 / et=log;

ENTRYTYPE= without a
slash2

proc catalog catalog=mycat et=log;

delete test1 test2;

1 in a subordinate statement
2 in the PROC CATALOG or the COPY statement

Note: All statements, except the CONTENTS statement, accept the ENTRYTYPE=
(alias ET=) option. 4

Why Use the ENTRYTYPE= Option?

ENTRYTYPE= can save keystrokes when you are processing multiple entries of the
same type.
To create a default for entry type for all statements in the current step, use
ENTRYTYPE= in the PROC CATALOG statement. To set the default for only the
current statement, use ENTRYTYPE= in a subordinate statement.
If many entries are of one type, but a few are of other types, you can use
ENTRYTYPE= to specify a default and then override that for individual entries with
(ENTRYTYPE=) in parentheses after those entries.

Avoid a Common Error

You cannot specify the ENTRYTYPE= option in both the PROC CATALOG statement
and a subordinate statement. For example, these statements generate an error and do
not delete any entries because the ENTRYTYPE= specications contradict each other:
/* THIS IS INCORRECT CODE. */
proc catalog cat=sample et=help;
delete a b c / et=program;
run;

The ENTRYTYPE= Option

The ENTRYTYPE= option is available in every statement in the CATALOG
procedure except CONTENTS.
ENTRYTYPE=etype
not in parentheses, sets a default entry type for the entire PROC step when used
in the PROC CATALOG statement. In all other statements, this option sets a
default entry type for the current statement.
Alias: ET=
Default: If you omit ENTRYTYPE=, PROC CATALOG processes all entries in the
catalog.

168

Catalog Concatenation

Chapter 6

Interaction: If you specify ENTRYTYPE= in the PROC CATALOG statement, do

not specify either ENTRYTYPE= or (ENTRYTYPE=) in a subordinate statement.

Interaction: (ENTRYTYPE=etype) in parentheses immediately following an entry
name overrides ENTRYTYPE= in that same statement.
Tip: On all statements except the PROC CATALOG and COPY statements, this
option follows a slash.
Tip: To process multiple entry types in a single PROC CATALOG step, use
ENTRYTYPE= in a subordinate statement, not in the PROC CATALOG
statement.
See also: Specifying an Entry Type on page 166.
Featured in: Example 1 on page 170
(ENTRYTYPE=etype)
in parentheses, identies the type of the entry just preceding it.
Alias: (ET=)
Restriction: (ENTRYTYPE=etype) immediately following an entry name in a
subordinate statement cannot override an ENTRYTYPE= option in the PROC
CATALOG statement. It generates a syntax error.
Interaction: (ENTRYTYPE=etype) immediately following an entry name
overrides ENTRYTYPE= in that same statement.
Tip: This form is useful mainly for specifying exceptions to an ENTRYTYPE=
option used in a subordinate statement. The following statement deletes
A.HELP, B.FORMAT, and C.HELP:
delete a b (et=format) c / et=help;

For the CHANGE and EXCHANGE statements, specify (ENTRYTYPE=) in

parentheses only once for each pair of names following the second name in the
pair. For example,

Tip:

change old1=new1 (et=log)

old1=new2 (et=help);

The CATALOG Procedure

Results: CATALOG Procedure

169

For example, the following code demonstrates these two rules, and the copy fails:
libname first path-name1;
libname second path-name2;
/* create concat.x */
libname concat (first second);
/* fails rule #1 */
proc catalog c=concat.x;
copy out=first.x new;
run;
quit;
/* fails rule #2 */
proc catalog c=first.x;
copy out=concat.x new;
run;
quit;

In summary, the following table shows when copies are allowed. In the table, A and
B are libraries, and each contains catalog X. Catalog C is an implicit concatenation of A
and B, and catalog D is an implicit concatenation of B and A.
Input catalog

Output catalog

Copy allowed?

C.X

B.X

C.X

D.X

C.X

A.X

B.X

Yes

B.X

A.X

Yes

C.X

A.X

B.X

C.X

Yes

A.X

C.X

Results: CATALOG Procedure

The CATALOG procedure produces output when the CONTENTS statement is
executed without options. The procedure output is assigned a name. You can use this
name to reference the table when using the Output Delivery System (ODS) to select
tables and create output data sets. For more information, see SAS Output Delivery
System: Users Guide.

170

Examples: CATALOG Procedure

Chapter 6

Table 6.2 ODS Tables Produced by the CATALOG Procedure

Table Name

The CATALOG procedure generates this table:

Catalog_Random

when the catalog is in a random-access data library.

Catalog_Sequential

when the catalog is in a sequential data library.

Examples: CATALOG Procedure

Example 1: Copying, Deleting, and Moving Catalog Entries from Multiple

Catalogs
Procedure features:

PROC CATALOG statement:

CATALOG= argument
COPY statement options:
IN=
MOVE
NOEDIT
DELETE statement options:
ENTRYTYPE= or ET=
EXCLUDE statement options:
ENTRYTYPE= or ET=
(ENTRYTYPE=) or (ET=)
QUIT statement
RUN statement
SELECT statement options:
ENTRYTYPE= or ET=

This example

3 copies entries by excluding a few entries

3 copies entries by specifying a few entries
3 protects entries from being edited
3 moves entries
3 deletes entries
3 processes entries from multiple catalogs
3 processes entries in multiple run groups.

The CATALOG Procedure

The CATALOG Procedure

Log

1
libname perm SAS-data-library;
NOTE: Directory for library PERM contains files of mixed engine types.
NOTE: Libref PERM was successfully assigned as follows:
Engine:
V9
Physical Name: SAS-data-library
2
options nodate pageno=1 linesize=80 pagesize=60 source;
3
proc catalog cat=perm.sample;
4
delete credit.program credit.log;
5
run;
NOTE: Deleting entry CREDIT.PROGRAM in catalog PERM.SAMPLE.
NOTE: Deleting entry CREDIT.LOG in catalog PERM.SAMPLE.
6
copy out=tcatall;
7
run;
NOTE: Copying entry DEFAULT.FORM from catalog PERM.SAMPLE to catalog
WORK.TCATALL.
NOTE: Copying entry FSLETTER.FORM from catalog PERM.SAMPLE to catalog
WORK.TCATALL.
NOTE: Copying entry LOAN.FRAME from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry LOAN.HELP from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry BUILD.KEYS from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry LOAN.KEYS from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry TEST1.LOG from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry TEST2.LOG from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry TEST3.LOG from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry LOAN.PMENU from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry TEST1.PROGRAM from catalog PERM.SAMPLE to catalog
WORK.TCATALL.
NOTE: Copying entry TEST2.PROGRAM from catalog PERM.SAMPLE to catalog
WORK.TCATALL.
NOTE: Copying entry TEST3.PROGRAM from catalog PERM.SAMPLE to catalog
WORK.TCATALL.
NOTE: Copying entry LOAN.SCL from catalog PERM.SAMPLE to catalog WORK.TCATALL.
NOTE: Copying entry PASSIST.SLIST from catalog PERM.SAMPLE to catalog
WORK.TCATALL.
NOTE: Copying entry PRTINFO.XPRINTER from catalog PERM.SAMPLE to catalog
WORK.TCATALL.

173

174

Example 2: Displaying Contents, Changing Names, and Changing a Description

Chapter 6

8
copy out=testcat;
9
exclude test1 test2 test3 passist (et=slist) / et=log;
10
run;
NOTE: Copying entry DEFAULT.FORM from catalog PERM.SAMPLE to catalog
WORK.TESTCAT.
NOTE: Copying entry FSLETTER.FORM from catalog PERM.SAMPLE to catalog
WORK.TESTCAT.
NOTE: Copying entry LOAN.FRAME from catalog PERM.SAMPLE to catalog WORK.TESTCAT.
NOTE: Copying entry LOAN.HELP from catalog PERM.SAMPLE to catalog WORK.TESTCAT.
NOTE: Copying entry BUILD.KEYS from catalog PERM.SAMPLE to catalog WORK.TESTCAT.
NOTE: Copying entry LOAN.KEYS from catalog PERM.SAMPLE to catalog WORK.TESTCAT.
NOTE: Copying entry LOAN.PMENU from catalog PERM.SAMPLE to catalog WORK.TESTCAT.
NOTE: Copying entry TEST1.PROGRAM from catalog PERM.SAMPLE to catalog
WORK.TESTCAT.
NOTE: Copying entry TEST2.PROGRAM from catalog PERM.SAMPLE to catalog
WORK.TESTCAT.
NOTE: Copying entry TEST3.PROGRAM from catalog PERM.SAMPLE to catalog
WORK.TESTCAT.
NOTE: Copying entry LOAN.SCL from catalog PERM.SAMPLE to catalog WORK.TESTCAT.
NOTE: Copying entry PRTINFO.XPRINTER from catalog PERM.SAMPLE to catalog
WORK.TESTCAT.
11
copy out=logcat move;
12
select test1 test2 test3 / et=log;
13
run;
NOTE: Moving entry TEST1.LOG from catalog PERM.SAMPLE to catalog WORK.LOGCAT.
NOTE: Moving entry TEST2.LOG from catalog PERM.SAMPLE to catalog WORK.LOGCAT.
NOTE: Moving entry TEST3.LOG from catalog PERM.SAMPLE to catalog WORK.LOGCAT.
14
copy out=perm.finance noedit;
15
select loan.frame loan.help loan.keys loan.pmenu;
16
run;
NOTE: Copying entry LOAN.FRAME from catalog PERM.SAMPLE to catalog PERM.FINANCE.
NOTE: Copying entry LOAN.HELP from catalog PERM.SAMPLE to catalog PERM.FINANCE.
NOTE: Copying entry LOAN.KEYS from catalog PERM.SAMPLE to catalog PERM.FINANCE.
NOTE: Copying entry LOAN.PMENU from catalog PERM.SAMPLE to catalog PERM.FINANCE.
17
copy in=perm.formats out=perm.finance;
18
select revenue.format dept.formatc;
19
quit;
NOTE: Copying entry REVENUE.FORMAT from catalog PERM.FORMATS to catalog
PERM.FINANCE.
NOTE: Copying entry DEPT.FORMATC from catalog PERM.FORMATS to catalog
PERM.FINANCE.

Example 2: Displaying Contents, Changing Names, and Changing a

Description
Procedure features:

PROC CATALOG statement

CHANGE statement options:
(ENTRYTYPE=) or (ET=)
CONTENTS statement options:
FILE=
MODIFY statement
RUN statement
QUIT statement

The CATALOG Procedure

Program

175

This example
3 lists the entries in a catalog and routes the output to a le
3 changes entry names
3 changes entry descriptions

3 processes entries in multiple run groups.

Program

Set the SAS system options. The system option SOURCE writes the source code to the log.
options nodate pageno=1 linesize=80 pagesize=60 source;

Assign a library reference. The LIBNAME statement assigns a libref to the SAS data library
that contains a permanent SAS catalog.
libname perm SAS-data-library;

List the entries in a catalog and route the output to a le. The CONTENTS statement
creates a listing of the contents of the SAS catalog PERM.FINANCE and routes the output to a
le.
proc catalog catalog=perm.finance;
contents;
title1 Contents of PERM.FINANCE before changes are made;
run;

Change entry names. The CHANGE statement changes the name of an entry that contains a
user-written character format. (ET=) species the entry type.
change dept=deptcode (et=formatc);
run;

Process entries in multiple run groups. The MODIFY statement changes the description of
an entry. The CONTENTS statement creates a listing of the contents of PERM.FINANCE after
all the changes have been applied. QUIT ends the procedure.
modify loan.frame (description=Loan analysis app. - ver1);
contents;
title1 Contents of PERM.FINANCE after changes are made;
run;
quit;

176

Output

Chapter 6

Output

Contents of PERM.FINANCE before changes are made

Contents of Catalog PERM.FINANCE

# Name
Type
Create Date
Modified Date Description
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
1 REVENUE FORMAT
16OCT1996:13:48:11 16OCT1996:13:48:11 FORMAT:MAXLEN=16,16,12
2 DEPT
FORMATC 30OCT1996:13:40:42 30OCT1996:13:40:42 FORMAT:MAXLEN=1,1,14
3 LOAN
FRAME
30OCT1996:13:40:43 30OCT1996:13:40:43 Loan analysis
application
4 LOAN
HELP
16OCT1996:13:48:10 16OCT1996:13:48:10 Information about
the application
5 LOAN
KEYS
16OCT1996:13:48:10 16OCT1996:13:48:10 Custom key definitions
for application
6 LOAN
PMENU
16OCT1996:13:48:10 16OCT1996:13:48:10 Custom menu
definitions for
application
7 LOAN
SCL
16OCT1996:13:48:10 16OCT1996:13:48:10 SCL code for loan
analysis application

Contents of PERM.FINANCE after changes are made

Contents of Catalog PERM.FINANCE

# Name
Type
Create Date
Modified Date Description
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
1 REVENUE FORMAT
16OCT1996:13:48:11 16OCT1996:13:48:11 FORMAT:MAXLEN=
16,16,12
2 DEPTCODE FORMATC 30OCT1996:13:40:42 30OCT1996:13:40:42 FORMAT:MAXLEN=1,1,14
3 LOAN
FRAME
30OCT1996:13:40:43 11FEB2002:13:20:50 Loan analysis
app. - ver1
4 LOAN
HELP
16OCT1996:13:48:10 16OCT1996:13:48:10 Information about
the application
5 LOAN
KEYS
16OCT1996:13:48:10 16OCT1996:13:48:10 Custom key
definitions for
application
6 LOAN
PMENU
16OCT1996:13:48:10 16OCT1996:13:48:10 Custom menu
definitions for
application
7 LOAN
SCL
16OCT1996:13:48:10 16OCT1996:13:48:10 SCL code for loan
analysis application

Example 3: Using the FORCE Option with the KILL Option

Procedure features:

PROC CATALOG statement:

CATALOG= argument
KILL option
FORCE option
QUIT statement
RUN statement

The CATALOG Procedure

Add the FORCE Option to the PROC CATALOG Statement

177

This example
3 creates a resource environment

3 tries to delete all catalog entries by using the KILL option but receives an error
3 species the FORCE option to successfully delete all catalog entries by using the
KILL option.

Program

Start a process (resource environment) by opening the catalog entry MATT in the
WORK.SASMACR catalog.
%macro matt;
%put &syscc;
%mend matt;

Specify the KILL option to delete all catalog entries in WORK.SASMACR. Since there is
a resource environment (process using the catalog), KILL will not work and an error is sent to
the log.
proc catalog c=work.sasmacr kill;
run;
quit;

Log
ERROR: You cannot open WORK.SASMACR.CATALOG for update access because
WORK.SASMACR.CATALOG is in use by you in resource environment
Line Mode Process.
WARNING: Command CATALOG not processed because of errors noted above.
NOTE: The SAS System stopped processing this step because of errors.
NOTE: PROCEDURE CATALOG used (Total process time):
real time
0.04 seconds
cpu time
0.03 seconds

Add the FORCE Option to the PROC CATALOG Statement

Add the FORCE option to the KILL option to delete the catalog entries.
proc catalog c=work.sasmacr kill force;
run;
quit;

178

Log

Chapter 6

Log
NOTE: Deleting entry MATT.MACRO in catalog WORK.SASMACR.

179

CHAPTER

7
The CHART Procedure
Overview: CHART Procedure 179
What Does the CHART Procedure Do? 179
What Types of Charts Can PROC CHART Create? 180
Syntax: CHART Procedure 184
PROC CHART Statement 185
BLOCK Statement 187
BY Statement 188
HBAR Statement 189
PIE Statement 189
STAR Statement 190
VBAR Statement 191
Customizing All Types of Charts 191
Concepts: CHART Procedure 197
Results: CHART Procedure 197
Missing Values 197
ODS Table Names 198
Portability of ODS Output with PROC CHART 198
Examples: CHART Procedure 198
Example 1: Producing a Simple Frequency Count 198
Example 2: Producing a Percentage Bar Chart 201
Example 3: Subdividing the Bars into Categories 203
Example 4: Producing Side-by-Side Bar Charts 206
Example 5: Producing a Horizontal Bar Chart for a Subset of the Data
Example 6: Producing Block Charts for BY Groups 210
References 213

209

Overview: CHART Procedure

What Does the CHART Procedure Do?
The CHART procedure produces vertical and horizontal bar charts, block charts, pie
charts, and star charts. These types of charts graphically display values of a variable or
a statistic associated with those values. The charted variable can be numeric or
character.
PROC CHART is a useful tool that lets you visualize data quickly, but if you need to
produce presentation-quality graphics that include color and various fonts, then use
SAS/GRAPH software. The GCHART procedure in SAS/GRAPH software produces the

180

What Types of Charts Can PROC CHART Create?

Chapter 7

same types of charts as PROC CHART does. In addition, PROC GCHART can produce
donut charts.

What Types of Charts Can PROC CHART Create?

Bar Charts
Horizontal and vertical bar charts display the magnitude of data with bars, each of
which represents a category of data. The length or height of the bars represents the
value of the chart statistic for each category.
Output 7.1 shows a vertical bar chart that displays the number of responses for the
ve categories from the survey data. The following statements produce the output:
options nodate pageno=1 linesize=80
pagesize=30;
proc chart data=survey;
vbar response / sumvar=count
midpoints=Always Usually
Sometimes Rarely Never;
run;

Output 7.1

Vertical Bar Chart

The SAS System

Count Sum
200 +
*****
|
*****
|
*****
|
*****
|
*****
150 +
*****
|
*****
|
*****
|
*****
*****
|
*****
*****
*****
100 +
*****
*****
*****
*****
|
*****
*****
*****
*****
|
*****
*****
*****
*****
|
*****
*****
*****
*****
|
*****
*****
*****
*****
50 +
*****
*****
*****
*****
|
*****
*****
*****
*****
*****
|
*****
*****
*****
*****
*****
|
*****
*****
*****
*****
*****
|
*****
*****
*****
*****
*****
-------------------------------------------------------------------Always
Usually
Sometimes
Rarely
Never
Response

Output 7.2 shows the same data presented in a horizontal bar chart. The two types
of bar charts have essentially the same characteristics, except that horizontal bar
charts by default display a table of statistic values to the right of the bars. The
following statements produce the output:

The CHART Procedure

What Types of Charts Can PROC CHART Create?

181

options nodate pageno=1 linesize=80

pagesize=60;
proc chart data=survey;
hbar response / sumvar=count
midpoints=Always Usually
Sometimes Rarely Never;
run;

Output 7.2 Horizontal Bar Chart

The SAS System
Response

Always
Usually
Sometimes
Rarely
Never

1
Count
Sum

|
|*********************
|
|****************************************
|
|************************
|
|*******************
|
|*********
|
----+---+---+---+---+---+---+---+---+---+
20 40 60 80 100 120 140 160 180 200

106.0000
202.0000
119.0000
97.0000
44.0000

Count Sum

Block Charts
Block charts display the relative magnitude of data by using blocks of varying height,
each set in a square that represents a category of data. Output 7.3 shows the number
of each survey response in the form of a block chart.
options nodate pageno=1 linesize=80
pagesize=30;
proc chart data=survey;
block response / sumvar=count
midpoints=Always Usually
Sometimes Rarely Never;
run;

182

What Types of Charts Can PROC CHART Create?

Output 7.3

Chapter 7

Block Chart
The SAS System

Sum of Count by Response

___
/_ /|
|**| |
|**| |
|**| |
___
___
|**| |
/_ /|
___
/_ /|
|**| |
|**| |
/_ /|
-|**| |--------|**| |--------|**| |--------|**| |--------------------/ |**| |
/ |**| |
/ |**| |
/ |**| |
/
___
/
/ |**| |
/ |**| |
/ |**| |
/ |**| |
/
/_ /|
/
/
|**| |
/
|**| |
/
|**| |
/
|**| |
/
|**| |
/
/
|**|/
/
|**|/
/
|**|/
/
|**|/
/
|**|/
/
/
/
/
/
/
/
/
106
/
202
/
119
/
97
/
44
/
/-------------/-------------/-------------/-------------/-------------/
Always

Usually

Sometimes

Rarely

Never

Response

Pie Charts
Pie charts represent the relative contribution of parts to the whole by displaying data
as wedge-shaped slices of a circle. Each slice represents a category of the data. Output
7.4 shows the survey results divided by response into ve pie slices. The following
statements produce the output:
options nodate pageno=1 linesize=80
pagesize=35;
proc chart data=survey;
pie response / sumvar=count;
run;

The CHART Procedure

What Types of Charts Can PROC CHART Create?

183

Output 7.4 Pie Chart

The SAS System

Sum of Count by Response

Never
***********
Rarely
****
.
****
**
.
.
**
**
. 44
.
**
*
.7.75%.
*
Always
**
97
.
..
**
**
17.08%
.
.
**
* ..
.
.
106
*
*
..
. .
18.66%
*
*
.. .
..
*
*
. .
*
*
+ . . .. . .. . .*
*
119
*
*
20.95%
..
*
Sometimes *
.
*
*
.
*
**
.
202
**
*
..
35.56%
*
*
.
*
** .
**
**
**
****
****
***********
Usually

Star Charts
With PROC CHART, you can produce star charts that show group frequencies, totals,
or mean values. A star chart is similar to a vertical bar chart, but the bars on a star
chart radiate from a center point, like spokes in a wheel. Star charts are commonly used
for cyclical data, such as measures taken every month or day or hour, or for data like
these in which the categories have an inherent order (always meaning more frequent
than usually which means more frequent than sometimes). Output 7.5 shows the
survey data displayed in a star chart. The following statements produce the output:
options nodate pageno=1 linesize=80
pagesize=60;
proc chart data=survey;
star response / sumvar=count;
run;

184

Syntax: CHART Procedure

Output 7.5

Chapter 7

Star Chart
The SAS System

Center = 0

Sum of Count by Response

Outside = 202

Never
*************
44
*****
*****
***
***
***
***
**
*
Rarely **
97
*
**
*
*
**
*
**
*
*
*
*
*
*
*
**
*
**
*
*
*
*
Sometimes **
119
*
**

**
*
**
*
**
*

*......
. ..
......*.
..
..
.. ...
.
..
.
....
.
.. ..
...
.
.+..............*
.
.. ..
.
.
...
.
.
.
.
.
.
.. ...
.
..
. ..
.
.
*.
..
.
..
.
..
..
.
.
*
..
.
..
*
...
..
.
*
..
.
.
**
...
. ..
*
.
. .
**
***
... . .
***
***
.... ***
*****
*.***
*************
Usually
202

*
**
*
**
*
*
*
*
*
*
*
**
*
**
*

Always
106

Syntax: CHART Procedure

You must use at least one of the chart-producing statements.
Tip: Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Names: See: ODS Table Names on page 198
Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See
Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.
Requirement:

The CHART Procedure

PROC CHART Statement

185

PROC CHART <option(s)>;

BLOCK variable(s) </ option(s)>;
BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
HBAR variable(s) </ option(s)>;
PIE variable(s) < / option(s)>;
STAR variable(s) </ option(s)>;
VBAR variable(s) </ option(s)>;

PROC CHART Statement

PROC CHART <option(s)>;

Options
DATA=SAS-data-set

identies the input SAS data set.

Main discussion: Input Data Sets on page 19
Restriction: You cannot use PROC CHART with an engine that supports
concurrent access if another user is updating the data set at the same time.
FORMCHAR <(position(s))>=formatting-character(s)

denes the characters to use for constructing the horizontal and vertical axes,
reference lines, and other structural parts of a chart. It also denes the symbols to
use to create the bars, blocks, or sections in the output.
position(s)
identies the position of one or more characters in the SAS formatting-character
string. A space or a comma separates the positions.
Default: Omitting (position(s)), is the same as specifying all 20 possible SAS
formatting characters, in order.
Range: PROC CHART uses 6 of the 20 formatting characters that SAS provides.
Table 7.1 on page 186 shows the formatting characters that PROC CHART uses.
Figure 7.1 on page 186 illustrates the use of formatting characters commonly
used in PROC CHART.
formatting-character(s)
lists the characters to use for the specied positions. PROC CHART assigns
characters in formatting-character(s) to position(s), in the order that they are
listed. For instance, the following option assigns the asterisk (*) to the second
formatting character, the pound sign (#) to the seventh character, and does not
alter the remaining characters:
formchar(2,7)=*#

Interaction: The SAS system option FORMCHAR= species the default formatting

characters. The system option denes the entire string of formatting characters.
The FORMCHAR= option in a procedure can redene selected characters.

186

PROC CHART Statement

Chapter 7

You can use any character in formatting-characters, including hexadecimal

characters. If you use hexadecimal characters, then you must put an x after the
closing quotation mark. For instance the following option assigns the hexadecimal
character 2D to the second formatting character, the hexadecimal character 7C to
the seventh character, and does not alter the remaining characters:

Tip:

formchar(2,7)=2D7Cx

See also: For information on which hexadecimal codes to use for which characters,

consult the documentation for your hardware.

Table 7.1 Formatting Characters Used by PROC CHART

Position

Default

Used to draw

Vertical axes in bar charts, the sides of the blocks in block charts, and
reference lines in horizontal bar charts. In side-by-side bar charts, the rst
and second formatting characters appear around each value of the group
variable (below the chart) to indicate the width of each group.

Horizontal axes in bar charts, the horizontal lines that separate the blocks
in a block chart, and reference lines in vertical bar charts. In side-by-side
bar charts, the rst and second formatting characters appear around each
value of the group variable (below the chart) to indicate the width of each
group.

Tick marks in bar charts and the centers in pie and star charts.

Intersection of axes in bar charts.

Ends of blocks and the diagonal lines that separate blocks in a block chart.

Circles in pie and star charts.

Figure 7.1 Formatting Characters Commonly Used in PROC CHART Output

Mean Yearly Pie Sales Grouped by Flavor
within Bakery Location

Pies_Sold Mean

400 +
| ***
***
300 +--***-------***---------***-------***-----------------------------------| ***
***
***
***
***
200 +--***--***--***---------***--***--***---------***-------***-------------| *** *** ***
*** *** ***
***
***
100 +--***--***--***---------***--***--***---------***--***--***-------------| *** *** *** ***
*** *** *** ***
*** *** *** ***
-------------------------------------------------------------------------a
b
c
r
a
b
c
r
a
b
c
r
Flavor
9
p
l
h
h
p
l
h
h
p
l
h
h
p
u
e
u
p
u
e
u
p
u
e
u
l
e
r
b
l
e
r
b
l
e
r
b
e
b
r
a
e
b
r
a
e
b
r
a
e
y
r
e
y
r
e
y
r
r
b
r
b
r
b
r
r
r
|----- Clyde ----|

LPI=value

|------ Oak -----|

|---- Samford ---|

Bakery

The CHART Procedure

BLOCK Statement

187

species the proportions of PIE and STAR charts. The value is determined by
(lines per inch = columns per inch)

For example, if you have a printer with 8 lines per inch and 12 columns per inch,
then specify LPI=6.6667.
Default: 6

BLOCK Statement
Produces a block chart.
Featured in: Example 6 on page 210

BLOCK variable(s) </ option(s)>;

Required Arguments
variable(s)

species the variables for which PROC CHART produces a block chart, one chart for
each variable.

Options
The options available on the BLOCK, HBAR, PIE, STAR, and VBAR statements are
documented in Customizing All Types of Charts on page 191.

Statement Results
Because each block chart must t on one output page, you may have to adjust the
SAS system options LINESIZE= and PAGESIZE= if you have a large number of charted
values for the BLOCK variable and for the variable specied in the GROUP= option.
Table 7.2 on page 187 shows the maximum number of charted values of BLOCK
variables for selected LINESIZE= (LS=) specications that can t on a 66-line page.
Table 7.2 Maximum Number of Bars of BLOCK Variables
GROUP= Value

LS= 132

LS= 120

LS= 105

LS= 90

LS= 76

LS= 64

0,1

5,6

188

BY Statement

Chapter 7

If the value of any GROUP= level is longer than three characters, then the maximum
number of charted values for the BLOCK variable that can t might be reduced by one.
BLOCK level values truncate to 12 characters. If you exceed these limits, then PROC
CHART produces a horizontal bar chart instead.

BY Statement
Produces a separate chart for each BY group.
Main discussion: BY on page 58
Featured in:

Example 6 on page 210

BY < DESCENDING> variable-1

<< DESCENDING> variable-n>
<NOTSORTED>;

Required Arguments
variable

Options
DESCENDING

species that the observations are sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

The CHART Procedure

PIE Statement

189

HBAR Statement
Produces a horizontal bar chart.
Tip: HBAR charts can print either the name or the label of the chart variable.
Featured in: Example 5 on page 209

HBAR variable(s) </ option(s)>;

Required Argument
variable(s)

species the variables for which PROC CHART produces a horizontal bar chart, one
chart for each variable.

Options
The options available on the BLOCK, HBAR, PIE, STAR, and VBAR statements are
documented in Customizing All Types of Charts on page 191.

Statement Results
Each chart occupies one or more output pages, depending on the number of bars;
each bar occupies one line, by default.
By default, for horizontal bar charts of TYPE=FREQ, CFREQ, PCT, or CPCT, PROC
CHART prints the following statistics: frequency, cumulative frequency, percentage,
and cumulative percentage. If you use one or more of the statistics options, then PROC
CHART prints only the statistics that you request, plus the frequency.

PIE Statement
Produces a pie chart.

PIE variable(s) </ option(s)>;

Required Argument
variable(s)

species the variables for which PROC CHART produces a pie chart, one chart for
each variable.

Options
The options available on the BLOCK, HBAR, PIE, STAR, and VBAR statements are
documented in Customizing All Types of Charts on page 191.

190

STAR Statement

Chapter 7

Statement Results
PROC CHART determines the number of slices for the pie in the same way that it
determines the number of bars for vertical bar charts. Any slices of the pie accounting
for less than three print positions are grouped together into an "OTHER" category.
The pies size is determined only by the SAS system options LINESIZE= and
PAGESIZE=. By default, the pie looks elliptical if your printer does not print 6 lines per
inch and 10 columns per inch. To make a circular pie chart on a printer that does not
print 6 lines and 10 columns per inch, use the LPI= option on the PROC CHART
statement. See the description of LPI= on page 186 for the formula that gives you the
proper LPI= value for your printer.
If you try to create a PIE chart for a variable with more than 50 levels, then PROC
CHART produces a horizontal bar chart instead.

STAR Statement
Produces a star chart.

STAR variable(s) </ option(s)>;

Required Argument
variable(s)

species the variables for which PROC CHART produces a star chart, one chart for
each variable.

Options
The options available on the BLOCK, HBAR, PIE, STAR, and VBAR statements are
documented in Customizing All Types of Charts on page 191.

Statement Results
The number of points in the star is determined in the same way as the number of
bars for vertical bar charts.
If all the data values are positive, then the center of the star represents zero and the
outside circle represents the maximum value. If any data values are negative, then the
center represents the minimum. See the description of the AXIS= option on page 193
for more information about how to specify maximum and minimum values. For
information about how to specify the proportion of the chart, see the description of the
LPI= option on page 186.
If you try to create a star chart for a variable with more than 24 levels, then PROC
CHART produces a horizontal bar chart instead.

The CHART Procedure

Customizing All Types of Charts

191

VBAR Statement
Produces a vertical bar chart.
Featured in: Example 1 on page 198, Example 2 on page 201, Example 3 on page 203,
Example 4 on page 206

VBAR variable(s) </ option(s)>;

Required Argument
variable(s)

species the variables for which PROC CHART produces a vertical bar chart, one
chart for each variable.

Options
The options available on the BLOCK, HBAR, PIE, STAR, and VBAR statements are
documented in Customizing All Types of Charts on page 191.

Statement Results
PROC CHART prints one page per chart. Along the vertical axis, PROC CHART
describes the chart frequency, the cumulative frequency, the chart percentage, the
cumulative percentage, the sum, or the mean. At the bottom of each bar, PROC CHART
prints a value according to the value of the TYPE= option, if specied. For character
variables or discrete numeric variables, this value is the actual value represented by
the bar. For continuous numeric variables, the value gives the midpoint of the interval
represented by the bar.
PROC CHART can automatically scale the vertical axis, determine the bar width,
and choose spacing between the bars. However, by using options, you can choose bar
intervals and the number of bars, include missing values in the chart, produce
side-by-side charts, and subdivide the bars. If the number of characters per line
(LINESIZE=) is not sufcient to display all vertical bars, then PROC CHART produces
a horizontal bar chart instead.

Customizing All Types of Charts

Many options in PROC CHART are valid in more than one statement. This section
describes the options that you can use on the chart-producing statements.
To do this

Use this option

Specify that numeric variables are discrete

DISCRETE

Specify a frequency variable

FREQ=

Specify that missing values are valid levels

MISSING

192

Customizing All Types of Charts

Chapter 7

To do this

Use this option

Specify the variable for which values or means are displayed

SUMVAR=

Specify the statistic represented in the chart

TYPE=

Specify groupings
Group the bars in side-by-side charts

GROUP=

Specify that group percentages sum to 100

G100

Group the bars in side-by-side charts

GROUP=

Specify the number of bars for continuous variables

LEVELS=

Dene ranges for continuous variables

MIDPOINTS=

Divide the bars into categories

SUBGROUP=

Compute statistics
Compute the cumulative frequency for each bar

CFREQ

Compute the cumulative percentage for each bar

CPERCENT

Compute the frequency for each bar

FREQ

Compute the mean of the observations for each bar

MEAN

Compute the percentage of total observations for each bar

PERCENT

Compute the total number of observations for each bar

SUM

Control output format

Print the bars in ascending order of size

ASCENDING

Specify the values for the response axis

AXIS=

Print the bars in descending order of size

DESCENDING

Specify extra space between groups of bars

GSPACE=

Suppress the default header line

NOHEADER

Allow no space between vertical bars

NOSPACE

Suppress the statistics

NOSTATS

Suppress the subgroup legend or symbol table

NOSYMBOL

Suppress the bars with zero frequency

NOZEROS

Draw reference lines

REF=

Specify the spaces between bars

SPACE=

Specify the symbols within bars or blocks

SYMBOL=

Specify the width of bars

WIDTH=

Options
ASCENDING

prints the bars and any associated statistics in ascending order of size within groups.
Alias: ASC
Restriction: Available only on the HBAR and VBAR statements

The CHART Procedure

Customizing All Types of Charts

193

AXIS=value-expression

species the values for the response axis, where value-expression is a list of
individual values, each separated by a space, or a range with a uniform interval for
the values. For example, the following range species tick marks on a bar chart from
0 to 100 at intervals of 10:
hbar x / axis=0 to 100 by 10;

Restriction: Not available on the PIE statement

Restriction: Values must be uniformly spaced, even if you specify them individually.
Restriction: For frequency charts, values must be integers.
Interaction: For BLOCK charts, AXIS= sets the scale of the tallest block. To set

the scale, PROC CHART uses the maximum value from the AXIS= list. If no value
is greater than 0, then PROC CHART ignores the AXIS= option.
Interaction: For HBAR and VBAR charts, AXIS= determines tick marks on the
response axis. If the AXIS= specication contains only one value, then the value
determines the minimum tick mark if the value is less than 0, or determines the
maximum tick mark if the value is greater than 0.
Interaction: For STAR charts, a single AXIS= value sets the minimum (the center
of the chart) if the value is less than zero, or sets the maximum (the outside circle)
if the value is greater than zero. If the AXIS= specication contains more than one
value, then PROC CHART uses the minimum and maximum values from the list.
Interaction: If you use AXIS= and the BY statement, then PROC CHART produces
uniform axes over BY groups.
CAUTION:

Values in value-expression override the range of the data. For example, if the data
range is 1 to 10 and you specify a range of 3 to 5, then only the data in the range 3
to 5 appears on the chart. Values out of range produce a warning message in the
SAS log. 4
CFREQ

prints the cumulative frequency.

Restriction: Available only on the HBAR statement
CPERCENT

prints the cumulative percentages.

Restriction: Available only on the HBAR statement
DESCENDING

prints the bars and any associated statistics in descending order of size within groups.
Alias: DESC
Restriction: Available only on the HBAR and VBAR statements
DISCRETE

species that a numeric chart variable is discrete rather than continuous. Without
DISCRETE, PROC CHART assumes that all numeric variables are continuous and
automatically chooses intervals for them unless you use MIDPOINTS= or LEVELS=.
FREQ

prints the frequency of each bar to the side of the chart.

Restriction: Available only on the HBAR statement
FREQ=variable

species a data set variable that represents a frequency count for each observation.
Normally, each observation contributes a value of one to the frequency counts. With
FREQ=, each observation contributes its value of the FREQ= value.

194

Customizing All Types of Charts

Chapter 7

Restriction: If the FREQ= values are not integers, then PROC CHART truncates

them.
Interaction: If you use SUMVAR=, then PROC CHART multiplies the sums by the

FREQ= value.
GROUP=variable

produces side-by-side charts, with each chart representing the observations that have
a common value for the GROUP= variable. The GROUP= variable can be character
or numeric and is assumed to be discrete. For example, the following statement
produces a frequency bar chart for men and women in each department:
vbar gender / group=dept;

Missing values for a GROUP= variable are treated as valid levels.

Restriction: Available only on the BLOCK, HBAR, and VBAR statements
Featured in: Example 4 on page 206, Example 5 on page 209, Example 6 on page
210
GSPACE=n

species the amount of extra space between groups of bars. Use GSPACE=0 to leave
no extra space between adjacent groups of bars.
Restriction: Available only on the HBAR and VBAR statements
Interaction: PROC CHART ignores GSPACE= if you omit GROUP=
G100

species that the sum of percentages for each group equals 100. By default, PROC
CHART uses 100 percent as the total sum. For example, if you produce a bar chart
that separates males and females into three age categories, then the six bars, by
default, add to 100 percent; however, with G100, the three bars for females add to
100 percent, and the three bars for males add to 100 percent.
Restriction: Available only on the BLOCK, HBAR, and VBAR statements
Interaction: PROC CHART ignores G100 if you omit GROUP=.
LEVELS=number-of-midpoints

species the number of bars that represent each chart variable when the variables
are continuous.
MEAN

prints the mean of the observations represented by each bar.

Restriction: Available only on the HBAR statement and only when you use
SUMVAR= and TYPE=
Restriction: Not available when TYPE=CFREQ, CPERCENT, FREQ, or PERCENT
MIDPOINTS=midpoint-specication | OLD

denes the range of values that each bar, block, or section represents by specifying
the range midpoints.
The value for MIDPOINTS= is one of the following:
midpoint-specication
species midpoints, either individually, or across a range at a uniform interval.
For example, the following statement produces a chart with ve bars; the rst bar
represents the range of values of X with a midpoint of 10, the second bar
represents the range with a midpoint of 20, and so on:
vbar x / midpoints=10 20 30 40 50;

Here is an example of a midpoint specication for a character variable:

vbar x / midpoints=JAN FEB MAR;

The CHART Procedure

Customizing All Types of Charts

195

Here is an example of specifying midpoints across a range at a uniform interval:

vbar x / midpoints=10 to 100 by 5;

OLD
species an algorithm that PROC CHART used in previous versions of SAS to
choose midpoints for continuous variables. The old algorithm was based on the
work of Nelder (1976). The current algorithm that PROC CHART uses if you omit
OLD is based on the work of Terrell and Scott (1985).
Default: Without MIDPOINTS=, PROC CHART displays the values in the SAS

Systems normal sorted order.

Restriction: When the VBAR variables are numeric, the midpoints must be given

in ascending order.
MISSING

species that missing values are valid levels for the chart variable.
NOHEADER

suppresses the default header line printed at the top of a chart.

NOHEADING

Alias:

Restriction: Available only on the BLOCK, PIE, and STAR statements

Featured in:

Example 6 on page 210

NOSTATS

suppresses the statistics on a horizontal bar chart.

NOSTAT

Alias:

Restriction: Available only on the HBAR statement

NOSYMBOL

suppresses printing of the subgroup symbol or legend table.

NOLEGEND

Alias:

Restriction: Available only on the BLOCK, HBAR, and VBAR statements

Interaction: PROC CHART ignores NOSYMBOL if you omit SUBGROUP=.
NOZEROS

suppresses any bar with zero frequency.

Restriction: Available only on the HBAR and VBAR statements
PERCENT

prints the percentages of observations having a given value for the chart variable.
Restriction: Available only on the HBAR statement
REF=value(s)

draws reference lines on the response axis at the specied positions.

Restriction: Available only on the HBAR and VBAR statements
Tip:

The REF= values should correspond to values of the TYPE= statistic.

Featured in:

Example 4 on page 206

SPACE=n

species the amount of space between individual bars.

Restriction: Available only on the HBAR and VBAR statements
Tip:

Use SPACE=0 to leave no space between adjacent bars.

Use the GSPACE= option to specify the amount of space between the bars
within each group.

Tip:

196

Customizing All Types of Charts

Chapter 7

SUBGROUP=variable

subdivides each bar or block into characters that show the contribution of the values
of variable to that bar or block. PROC CHART uses the rst character of each value
to ll in the portion of the bar or block that corresponds to that value, unless more
than one value begins with the same rst character. In that case, PROC CHART
uses the letters A, B, C, and so on to ll in the bars or blocks. If the variable is
formatted, then PROC CHART uses the rst character of the formatted value.
The characters used in the chart and the values that they represent are given in a
legend at the bottom of the chart. The subgroup symbols are ordered A through Z
and 0 through 9 with the characters in ascending order.
PROC CHART calculates the height of a bar or block for each subgroup
individually and then rounds the percentage of the total bar up or down. So the total
height of the bar may be higher or lower than the same bar without the
SUBGROUP= option.
Restriction: Available only on the BLOCK, HBAR, and VBAR statements
Interaction: If you use both TYPE=MEAN and SUBGROUP=, then PROC CHART
rst calculates the mean for each variable that is listed in the SUMVAR= option,
then subdivides the bar into the percentages that each subgroup contributes.
Featured in: Example 3 on page 203
SUM

prints the total number of observations that each bar represents.

Restriction: Available only on the HBAR statement and only when you use both
SUMVAR= and TYPE=
Restriction: Not available when TYPE=CFREQ, CPERCENT, FREQ, or PERCENT
SUMVAR=variable

species the variable for which either values or means (depending on the value of
TYPE=) PROC CHART displays in the chart.
Interaction: If you use SUMVAR= and you use TYPE= with a value other than
MEAN or SUM, then TYPE=SUM overrides the specied TYPE= value.
Tip: Both HBAR and VBAR charts can print labels for SUMVAR= variables if you
use a LABEL statement.
Featured in: Example 3 on page 203, Example 4 on page 206, Example 5 on page
209, Example 6 on page 210
SYMBOL=character(s)

species the character or characters that PROC CHART uses in the bars or blocks of
the chart when you do not use the SUBGROUP= option.
Default: asterisk (*)
Restriction: Available only on the BLOCK, HBAR, and VBAR statements
Interaction: If the SAS system option OVP is in effect and if your printing device
supports overprinting, then you can specify up to three characters to produce
overprinted charts.
Featured in: Example 6 on page 210
TYPE=statistic

species what the bars or sections in the chart represent. The statistic is one of the
following:
CFREQ
species that each bar, block, or section represent the cumulative frequency.
CPERCENT
species that each bar, block, or section represent the cumulative percentage.

The CHART Procedure

Missing Values

197

Alias: CPCT

FREQ
species that each bar, block, or section represent the frequency with which a
value or range occurs for the chart variable in the data.
MEAN
species that each bar, block, or section represent the mean of the SUMVAR=
variable across all observations that belong to that bar, block, or section.
Interaction: With TYPE=MEAN, you can only compute MEAN and FREQ
statistics.
Featured in: Example 4 on page 206
PERCENT
species that each bar, block, or section represent the percentage of observations
that have a given value or that fall into a given range of the chart variable.
Alias: PCT
Featured in: Example 2 on page 201
SUM
species that each bar, block, or section represent the sum of the SUMVAR=
variable for the observations that correspond to each bar, block, or section.
Default: FREQ (unless you use SUMVAR=, which causes a default of SUM)
Interaction: With TYPE=SUM, you can only compute SUM and FREQ statistics.
WIDTH=n

species the width of the bars on bar charts.

Restriction: Available only on the HBAR and VBAR statements

Concepts: CHART Procedure

The following are variable characteristics for the CHART procedure:
3 Character variables and formats cannot exceed a length of 16.
3 For continuous numeric variables, PROC CHART automatically selects display
intervals, although you can explicitly dene interval midpoints.
3 For character variables and discrete numeric variables, which contain several
distinct values rather than a continuous range, the data values themselves dene
the intervals.

Results: CHART Procedure

Missing Values
PROC CHART follows these rules when handling missing values:
3 Missing values are not considered as valid levels for the chart variable when you
use the MISSING option.
3 Missing values for a GROUP= or SUBGROUP= variable are treated as valid levels.
3 PROC CHART ignores missing values for the FREQ= option and the SUMVAR=
option.

198

ODS Table Names

Chapter 7

3 If the value of the FREQ= variable is missing, zero, or negative, then the
observation is excluded from the calculation of the chart statistic.

3 If the value of the SUMVAR= variable is missing, then the observation is excluded
from the calculation of the chart statistic.

ODS Table Names

The CHART procedure assigns a name to each table that it creates. You can use
these names to reference the table when using the Output Delivery System (ODS) to
select tables and create output data sets. For more information, see SAS Output
Delivery System: Users Guide.
Table 7.3 ODS Tables Produced by the CHART Procedure
Name

Description

Statement Used

BLOCK

A block chart

BLOCK

HBAR

A horizontal bar chart

HBAR

PIE

A pie chart

PIE

STAR

A star chart

STAR

VBAR

A vertical bar chart

VBAR

Portability of ODS Output with PROC CHART

Under certain circumstances, using PROC CHART with the Output Delivery System
produces les that are not portable. If the SAS system option FORMCHAR= in your
SAS session uses nonstandard line-drawing characters, then the output might include
strange characters instead of lines in operating environments in which the SAS
Monospace font is not installed. To avoid this problem, specify the following OPTIONS
statement before executing PROC CHART:
options formchar="|----|+|---+=|-/\<>*";

Examples: CHART Procedure

Example 1: Producing a Simple Frequency Count

Procedure features:

VBAR statement

This example produces a vertical bar chart that shows a frequency count for the
values of the chart variable.

The CHART Procedure

Program

199

Program

Create the SHIRTS data set. SHIRTS contains the sizes of a particular shirt that is sold
during a week at a clothing store, with one observation for each shirt that is sold.
data shirts;
input Size $ @@;
datalines;
medium
large
large
large
large
medium
medium
small
small
medium
medium
large
small
medium
large
large
large
small
medium
medium
medium
medium
medium
large
small
small
;

Create a vertical bar chart with frequency counts. The VBAR statement produces a
vertical bar chart for the frequency counts of the Size values.
proc chart data=shirts;
vbar size;

Specify the title.

title Number of Each Shirt Size Sold;
run;

200

Output

Chapter 7

Output

The frequency chart shows the stores sales of the shirt for the week: 9
large shirts, 11 medium shirts, and 6 small shirts.

Number of Each Shirt Size Sold

Frequency
11 +
*****
|
*****
|
*****
|
*****
10 +
*****
|
*****
|
*****
|
*****
9 +
*****
*****
|
*****
*****
|
*****
*****
|
*****
*****
8 +
*****
*****
|
*****
*****
|
*****
*****
|
*****
*****
7 +
*****
*****
|
*****
*****
|
*****
*****
|
*****
*****
6 +
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
5 +
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
4 +
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
3 +
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
2 +
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
1 +
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
-------------------------------------------large
medium
small
Size

The CHART Procedure

Program

201

Example 2: Producing a Percentage Bar Chart

Procedure features:

VBAR statement option:

TYPE=
Data set: SHIRTS on page 199

This example produces a vertical bar chart. The chart statistic is the percentage for
each category of the total number of shirts sold.

Program

Create a vertical bar chart with percentages. The VBAR statement produces a vertical bar
chart. TYPE= species percentage as the chart statistic for the variable Size.
proc chart data=shirts;
vbar size / type=percent;

Specify the title.

title Percentage of Total Sales for Each Shirt Size;
run;

202

Output

Chapter 7

Output

The chart shows the percentage of total sales for each shirt size. Of all
the shirts sold, about 42.3 percent were medium, 34.6 were large, and
23.1 were small.

Percentage of Total Sales for Each Shirt Size

Percentage

|
*****
|
*****
+
*****
|
*****
|
*****
|
*****
|
*****
+
*****
*****
|
*****
*****
|
*****
*****
|
*****
*****
|
*****
*****
+
*****
*****
|
*****
*****
|
*****
*****
|
*****
*****
|
*****
*****
+
*****
*****
|
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
+
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
+
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
+
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
+
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
|
*****
*****
*****
-------------------------------------------large
medium
small
Size

The CHART Procedure

Program

203

Example 3: Subdividing the Bars into Categories

Procedure features:

VBAR statement options:

SUBGROUP=
SUMVAR=

This example

3 produces a vertical bar chart for categories of one variable with bar lengths that
represent the values of another variable.

3 subdivides each bar into categories based on the values of a third variable.

Program

Create the PIESALES data set. PIESALES contains the number of each avor of pie that is
sold for two years at three bakeries that are owned by the same company. One bakery is on
Samford Avenue, one on Oak Street, and one on Clyde Drive.
data piesales;
input Bakery $ Flavor $ Year Pies_Sold;
datalines;
Samford apple
1995 234
Samford apple
1996 288
Samford blueberry 1995 103
Samford blueberry 1996 143
Samford cherry
1995 173
Samford cherry
1996 195
Samford rhubarb
1995
26
Samford rhubarb
1996
28
Oak
apple
1995 319
Oak
apple
1996 371
Oak
blueberry 1995 174
Oak
blueberry 1996 206
Oak
cherry
1995 246
Oak
cherry
1996 311
Oak
rhubarb
1995
51
Oak
rhubarb
1996
56
Clyde
apple
1995 313
Clyde
apple
1996 415
Clyde
blueberry 1995 177
Clyde
blueberry 1996 201

204

Program

Chapter 7

Clyde
Clyde
Clyde
Clyde
;

cherry
cherry
rhubarb
rhubarb

1995
1996
1995
1996

250
328
60
59

Create a vertical bar chart with the bars that are subdivided into categories. The
VBAR statement produces a vertical bar chart with one bar for each pie avor. SUBGROUP=
divides each bar into sales for each bakery.
proc chart data=piesales;
vbar flavor / subgroup=bakery

Specify the bar length variable. SUMVAR= species Pies_Sold as the variable whose values
are represented by the lengths of the bars.
sumvar=pies_sold;

Specify the title.

title Pie Sales by Flavor Subdivided by Bakery Location;
run;

The CHART Procedure

Output

205

Output

The bar that represents the sales of apple pies, for example, shows 1,940 total pies across both
years and all three bakeries. The symbol for the Samford Avenue bakery represents the 522
pies at the top, the symbol for the Oak Street bakery represents the 690 pies in the middle, and
the symbol for the Clyde Drive bakery represents the 728 pies at the bottom of the bar for apple
pies. By default, the labels along the horizontal axis are truncated to eight characters.

Pie Sales by Flavor Subdivided by Bakery Location

Pies_Sold Sum

1800

1600

1400

1200

1000

800

600

400

200

Symbol Bakery
C

Clyde

Symbol Bakery
O

Oak

Symbol Bakery
S

Samford

206

Example 4: Producing Side-by-Side Bar Charts

Chapter 7

Example 4: Producing Side-by-Side Bar Charts

Procedure features:

VBAR statement options:

GROUP=
REF=
SUMVAR=
TYPE=
Data set: PIESALESProgram on page 203

This example
3 charts the mean values of a variable for the categories of another variable

3 creates side-by-side bar charts for the categories of a third variable

3 draws reference lines across the charts.

Program

Create a side-by-side vertical bar chart. The VBAR statement produces a side-by-side
vertical bar chart to compare the sales across values of Bakery, specied by GROUP=. Each
Bakery group contains a bar for each Flavor value.
proc chart data=piesales;
vbar flavor / group=bakery

The CHART Procedure

Program

Create reference lines. REF= draws reference lines to mark pie sales at 100, 200, and 300.
ref=100 200 300

Specify the bar length variable. SUMVAR= species Pies_Sold as the variable that is
represented by the lengths of the bars.
sumvar=pies_sold

Specify the statistical variable. TYPE= averages the sales for 1995 and 1996 for each
combination of bakery and avor.
type=mean;

Specify the titles.

title Mean Yearly Pie Sales Grouped by Flavor;
title2 within Bakery Location;
run;

207

208

Output

Chapter 7

Output

The side-by-side bar charts compare the sales of apple pies, for example, across bakeries. The
mean for the Clyde Drive bakery is 364, the mean for the Oak Street bakery is 345, and the
mean for the Samford Avenue bakery is 261.

Mean Yearly Pie Sales Grouped by Flavor

within Bakery Location

Pies_Sold Mean
| ***
350 + ***
***
| ***
***
| ***
***
| ***
***
| ***
***
300 +--***-------------------***---------------------------------------------| ***
***
***
| ***
***
***
***
| ***
***
***
***
| ***
***
***
***
***
250 + ***
***
***
***
***
| ***
***
***
***
***
| ***
***
***
***
***
| ***
***
***
***
***
| ***
***
***
***
***
200 +--***-------***---------***-------***---------***-----------------------| *** *** ***
*** *** ***
***
| *** *** ***
*** *** ***
***
***
| *** *** ***
*** *** ***
***
***
| *** *** ***
*** *** ***
***
***
150 + *** *** ***
*** *** ***
***
***
| *** *** ***
*** *** ***
***
***
| *** *** ***
*** *** ***
***
***
| *** *** ***
*** *** ***
*** *** ***
| *** *** ***
*** *** ***
*** *** ***
100 +--***--***--***---------***--***--***---------***--***--***-------------| *** *** ***
*** *** ***
*** *** ***
| *** *** ***
*** *** ***
*** *** ***
| *** *** ***
*** *** ***
*** *** ***
| *** *** *** ***
*** *** ***
*** *** ***
50 + *** *** *** ***
*** *** *** ***
*** *** ***
| *** *** *** ***
*** *** *** ***
*** *** ***
| *** *** *** ***
*** *** *** ***
*** *** *** ***
| *** *** *** ***
*** *** *** ***
*** *** *** ***
| *** *** *** ***
*** *** *** ***
*** *** *** ***
-------------------------------------------------------------------------a
b
c
r
a
b
c
r
a
b
c
r
Flavor
p
l
h
h
p
l
h
h
p
l
h
h
p
u
e
u
p
u
e
u
p
u
e
u
l
e
r
b
l
e
r
b
l
e
r
b
e
b
r
a
e
b
r
a
e
b
r
a
e
y
r
e
y
r
e
y
r
r
b
r
b
r
b
r
r
r
|----- Clyde ----|

|------ Oak -----|

|---- Samford ---|

Bakery

The CHART Procedure

Program

209

Example 5: Producing a Horizontal Bar Chart for a Subset of the Data

Procedure features:

HBAR statement options:

GROUP=
SUMVAR=
Other features:

WHERE= data set option

Data set: PIESALESProgram on page 203

This example
3 produces horizontal bar charts only for observations with a common value
3 charts the values of a variable for the categories of another variable
3 creates side-by-side bar charts for the categories of a third variable.

Program

Specify the variable value limitation for the horizontal bar chart. WHERE= limits the
chart to only the 1995 sales totals.
proc chart data=piesales(where=(year=1995));

Create a side-by-side horizontal bar chart. The HBAR statement produces a side-by-side
horizontal bar chart to compare sales across values of Flavor, specied by GROUP=. Each
Flavor group contains a bar for each Bakery value.
hbar bakery / group=flavor

Specify the bar length variable. SUMVAR= species Pies_Sold as the variable whose values
are represented by the lengths of the bars.
sumvar=pies_sold;

Specify the title.

title 1995 Pie Sales for Each Bakery According to Flavor;
run;

210

Output

Chapter 7

Output
1995 Pie Sales for Each Bakery According to Flavor
Flavor

Bakery

apple

Clyde
Oak
Samford

blueberr

Clyde
Oak
Samford

cherry

Clyde
Oak
Samford

rhubarb

Clyde
Oak
Samford

Pies_Sold
Sum
|
|******************************************
|*******************************************
|*******************************
|
|************************
|***********************
|**************
|
|*********************************
|*********************************
|***********************
|
|********
|*******
|***
|
----+---+---+---+---+---+---+---+---+---+--30 60 90 120 150 180 210 240 270 300
Pies_Sold Sum

Example 6: Producing Block Charts for BY Groups

Procedure features:

BLOCK statement options:

GROUP=
NOHEADER=
SUMVAR=
SYMBOL=
BY statement
Other features:

PROC SORT
SAS system options:
NOBYLINE
OVP
TITLE statement:
#BYVAL specication
Data set:

PIESALESProgram on page 203

This example
3 sorts the data set

313.0000
319.0000
234.0000
177.0000
174.0000
103.0000
250.0000
246.0000
173.0000
60.0000
51.0000
26.0000

The CHART Procedure

Program

211

3 produces a block chart for each BY group

3 organizes the blocks into a three-dimensional chart
3 prints BY group-specic titles.

Program

Sort the input data set PIESALES. PROC SORT sorts PIESALES by year. Sorting is
required to produce a separate chart for each year.
proc sort data=piesales out=sorted_piesales;
by year;
run;

Suppress BY lines and allow overprinted characters in the block charts. NOBYLINE
suppresses the usual BY lines in the output. OVP allows overprinted characters in the charts.
options nobyline ovp;

Specify the BY group for multiple block charts. The BY statement produces one chart for
1995 sales and one for 1996 sales.
proc chart data=sorted_piesales;
by year;

Create a block chart. The BLOCK statement produces a block chart for each year. Each chart
contains a grid (Bakery values along the bottom, Flavor values along the side) of cells that
contain the blocks.
block bakery / group=flavor

Specify the bar length variable. SUMVAR= species Pies_Sold as the variable whose values
are represented by the lengths of the blocks.
sumvar=pies_sold

212

Output

Chapter 7

Suppress the default header line. NOHEADER suppresses the default header line.
noheader

Specify the block symbols. SYMBOL= species the symbols in the blocks.
symbol=OX;

Specify the titles. The #BYVAL specication inserts the year into the second line of the title.
title Pie Sales for Each Bakery and Flavor;
title2 #byval(year);
run;

Reset the printing of the default BY line. The SAS system option BYLINE resets the
printing of the default BY line.
options byline;

Output

Flavor

Bakery

The CHART Procedure

References

Flavor

Bakery

References
Nelder, J.A. (1976), A Simple Algorithm for Scaling Graphs, Applied Statistics,
Volume 25, Number 1, London: The Royal Statistical Society.
Terrell, G.R. and Scott, D.W. (1985), Oversmoothed Nonparametric Density
Estimates, Journal of the American Statistical Association, 80, 389, 209214.

213

214

215

CHAPTER

8
The CIMPORT Procedure
Overview: CIMPORT Procedure 215
What Does the CIMPORT Procedure Do? 215
General File Transport Process 215
Syntax: CIMPORT Procedure 216
PROC CIMPORT Statement 216
EXCLUDE Statement 219
SELECT Statement 220
Results: CIMPORT Procedure 221
Examples: CIMPORT Procedure 222
Example 1: Importing an Entire Data Library 222
Example 2: Importing Individual Catalog Entries 223
Example 3: Importing a Single Indexed SAS Data Set 224

Overview: CIMPORT Procedure

What Does the CIMPORT Procedure Do?
The CIMPORT procedure imports a transport le that was created (exported) by the
CPORT procedure. PROC CIMPORT restores the transport le to its original form as a
SAS catalog, SAS data set, or SAS data library. Transport les are sequential les that
each contain a SAS data library, a SAS catalog, or a SAS data set in transport format.
The transport format that PROC CPORT writes is the same for all environments and
for many releases of SAS.
PROC CIMPORT can read only transport les that PROC CPORT creates. For
information on the transport les that the transport engine creates, see the section on
SAS les in SAS Language Reference: Concepts.
PROC CIMPORT also converts SAS les, which means that it changes the format of
a SAS le from the format appropriate for one version of SAS to the format appropriate
for another version. For example, you can use PROC CPORT and PROC CIMPORT to
move les from earlier releases of SAS to more recent releases. In such cases, PROC
CIMPORT automatically converts the contents of the transport le as it imports it.
PROC CIMPORT produces no output, but it does write notes to the SAS log.

General File Transport Process

To export and import les, follow these steps:
1 Use PROC CPORT to export the SAS les that you want to transport.

216

Syntax: CIMPORT Procedure

Chapter 8

2 If you are changing operating environments, move the transport le to the new

machine by using either communications software or a magnetic medium.

Note: If you use communications software to move the transport le, be sure that
it treats the transport le as a binary le and that it modies neither the
attributes nor the contents of the le.

3 Use PROC CIMPORT to translate the transport le into the format appropriate

for the new operating environment or release.

Syntax: CIMPORT Procedure

See:

CIMPORT Procedure in the documentation for your operating environment.

PROC CIMPORT destination=libref | <libref.>member-name <option(s)>;

EXCLUDE SAS le(s) | catalog entry(s)</ MEMTYPE=mtype></
ENTRYTYPE=entry-type>;
SELECT SAS le(s) | catalog entry(s)< / MEMTYPE=mtype></
ENTRYTYPE=entry-type>;

PROC CIMPORT Statement

PROC CIMPORT destination=libref | <libref.> member-name<option(s)>;

To do this

Use this option

Identify the input transport le

Specify a previously dened leref or the lename of the
transport le to read

INFILE=

Read the input transport le from a tape

TAPE

Select les to import

Exclude specied entry types from the import process

EET=

Specify entry types to import

ET=

Control the contents of the transport le

Import a SAS le without changing the created and modied
date and time

DATECOPY

Specify whether to extend by 1 byte the length of short

numerics (less than 8 bytes) when you import them

EXTENDSN=

Specify that only data sets, only catalogs, or both, be moved

when a library is imported

MEMTYPE=

The CIMPORT Procedure

PROC CIMPORT Statement

To do this

Use this option

Enable access to a locked catalog

FORCE

Create a new catalog for the imported transport le, and delete
any existing catalog with the same name

NEW

Import SAS/AF PROGRAM and SCL entries without edit

capability

NOEDIT

Suppress the importing of source code for SAS/AF entries that

contain compiled SCL code

217

NOSRC

Required Arguments
destination=libref | < libref. >member-name

identies the type of le to import and species the specic catalog, SAS data set, or
SAS data library to import.
destination
identies the le or les in the transport le as a single catalog, as a single SAS
data set, or as the members of a SAS data library. The destination argument can
be one of the following:
CATALOG | CAT | C
DATA | DS | D
LIBRARY | LIB | L
libref | <libref. > member-name
species the specic catalog, SAS data set, or SAS data library as the destination
of the transport le. If the destination argument is CATALOG or DATA, you can
specify both a libref and a member name. If the libref is omitted, PROC CIMPORT
uses the default library as the libref, which is usually the WORK library. If the
destination argument is LIBRARY, specify only a libref.

Options
DATECOPY

copies the SAS internal date and time when the SAS le was created and the date
and time when it was last modied to the resulting destination le. Note that the
operating environment date and time are not preserved.
Restriction: DATECOPY can be used only when the destination le uses the V8 or

V9 engine.
You can alter the le creation date and time with the DTC= option on the
MODIFY statementMODIFY Statement on page 348 in a PROC DATASETS step.

Tip:

EET=(etype(s))

excludes specied entry types from the import process. If the etype is a single entry
type, then you can omit the parentheses. Separate multiple values with spaces.
Interaction: You cannot specify both the EET= option and the ET= option in the

same PROC CIMPORT step.

218

PROC CIMPORT Statement

Chapter 8

ET=(etype(s))

species the entry types to import. If the etype is a single entry type, then you can
omit the parentheses. Separate multiple values with spaces.
Interaction: You cannot specify both the EET= option and the ET= option in the

same PROC CIMPORT step.

EXTENDSN=YES | NO

species whether to extend by 1 byte the length of short numerics (fewer than 8
bytes) when you import them. You can avoid a loss of precision when you transport a
short numeric in IBM format to IEEE format if you extend its length. You cannot
extend the length of an 8-byte short numeric.
Default: YES
Restriction: This option applies only to data sets.
Tip:

Do not store fractions as short numerics.

FORCE

enables access to a locked catalog. By default, PROC CIMPORT locks the catalog
that it is updating to prevent other users from accessing the catalog while it is being
updated. The FORCE option overrides this lock, which allows other users to access
the catalog while it is being imported, or allows you to import a catalog that is
currently being accessed by other users.
CAUTION:

The FORCE option can lead to unpredictable results. The FORCE option allows
multiple users to access the same catalog entry simultaneously. 4
INFILE=leref | lename

species a previously dened leref or the lename of the transport le to read. If

you omit the INFILE= option, then PROC CIMPORT attempts to read from a
transport le with the leref SASCAT. If a leref SASCAT does not exist, then PROC
CIMPORT attempts to read from a le named SASCAT.DAT.
Alias:

FILE=

Featured in:

Example 1 on page 222.

MEMTYPE=mtype

species that only data sets, only catalogs, or both, be moved when a SAS library is
imported. Values for mtype can be
ALL
both catalogs and data sets
CATALOG | CAT
catalogs
DATA | DS
SAS data sets
NEW

creates a new catalog to contain the contents of the imported transport le when the
destination you specify has the same name as an existing catalog. NEW deletes any
existing catalog with the same name as the one you specify as a destination for the
import. If you do not specify NEW, and the destination you specify has the same
name as an existing catalog, PROC CIMPORT appends the imported transport le to
the existing catalog.
NOEDIT

imports SAS/AF PROGRAM and SCL entries without edit capability.

The CIMPORT Procedure

EXCLUDE Statement

219

You obtain the same results if you create a new catalog to contain SCL code by
using the MERGE statement with the NOEDIT option in the BUILD procedure of
SAS/AF software.
Note: The NOEDIT option affects only SAS/AF PROGRAM and SCL entries. It
does not affect FSEDIT SCREEN and FSVIEW FORMULA entries. 4
Alias:

NEDIT

NOSRC

suppresses the importing of source code for SAS/AF entries that contain compiled
SCL code.
You obtain the same results if you create a new catalog to contain SCL code by
using the MERGE statement with the NOSOURCE option in the BUILD procedure
of SAS/AF software.
Alias: NSRC
Interaction: PROC CIMPORT ignores the NOSRC option if you use it with an

entry type other than FRAME, PROGRAM, or SCL.

TAPE

reads the input transport le from a tape.

Default: PROC CIMPORT reads from disk.

EXCLUDE Statement
Excludes specied les or entries from the import process.
There is no limit to the number of EXCLUDE statements you can use in one
invocation of PROC CIMPORT.

Tip:

Interaction: You can use either EXCLUDE statements or SELECT statements in a PROC
CIMPORT step, but not both.

EXCLUDE SAS le(s) | catalog entry(s)</ MEMTYPE=mtype></ ENTRYTYPE=

entry-type>;

Required Arguments
SAS le(s) | catalog entry(s)

species either the name(s) of one or more SAS les or the name(s) of one or more
catalog entries to be excluded from the import process. Specify SAS lenames if you
import a data library; specify catalog entry names if you import an individual SAS
catalog. Separate multiple lenames or entry names with a space. You can use
shortcuts to list many like-named les in the EXCLUDE statement. For more
information, see Shortcuts for Specifying Lists of Variable Names on page 24.

220

SELECT Statement

Chapter 8

Options
ENTRYTYPE=entry-type

species a single entry type for the catalog entry(s) listed in the EXCLUDE
statement. See SAS Language Reference: Concepts for a complete list of catalog entry
types.
Restriction: ENTRYTYPE= is valid only when you import an individual SAS

catalog.
Alias:

ETYPE=, ET=

MEMTYPE=mtype

species a single member type for the SAS le(s) listed in the EXCLUDE statement.
Values for mtype can be
ALL
both catalogs and data sets
CATALOG
catalogs
DATA
SAS data sets.
You can also specify the MEMTYPE= option, enclosed in parentheses, immediately
after the name of a le. In parentheses, MEMTYPE= identies the type of the
lename that just precedes it. When you use this form of the option, it overrides the
MEMTYPE= option that follows the slash in the EXCLUDE statement, but it must
match the MEMTYPE= option in the PROC CIMPORT statement.
Restriction: MEMTYPE= is valid only when you import a SAS data library.
Alias:

MTYPE=, MT=

Default: ALL

SELECT Statement
Species individual les or entries to import.
Tip: There is no limit to the number of SELECT statements you can use in one
invocation of PROC CIMPORT.
Interaction: You can use either EXCLUDE statements or SELECT statements in a PROC
CIMPORT step, but not both.
Featured in:

Example 2 on page 223

SELECT SAS le(s) | catalog entry(s)< / MEMTYPE=mtype>< /

ENTRYTYPE=entry-type>;

The CIMPORT Procedure

Results: CIMPORT Procedure

221

Required Arguments
SAS le(s) | catalog entry(s)

species either the name(s) of one or more SAS les or the name(s) of one or more
catalog entries to import. Specify SAS lenames if you import a data library; specify
catalog entry names if you import an individual SAS catalog. Separate multiple
lenames or entry names with a space. You can use shortcuts to list many
like-named les in the SELECT statement. For more information, see Shortcuts for
Specifying Lists of Variable Names on page 24.

Options
ENTRYTYPE=entry-type

species a single entry type for the catalog entry(s) listed in the SELECT statement.
See SAS Language Reference: Concepts for a complete list of catalog entry types.
Restriction: ENTRYTYPE= is valid only when you import an individual SAS

catalog.
Alias:

ETYPE=, ET=

MEMTYPE=mtype

species a single member type for the SAS le(s) listed in the SELECT statement.
Valid values are CATALOG or CAT, DATA, or ALL.
You can also specify the MEMTYPE= option, enclosed in parentheses, immediately
after the name of a le. In parentheses, MEMTYPE= identies the type of the
lename that just precedes it. When you use this form of the option, it overrides the
MEMTYPE= option that follows the slash in the SELECT statement, but it must
match the MEMTYPE= option in the PROC CIMPORT statement.
Restriction: MEMTYPE= is valid only when you import a SAS data library.
Alias:

MTYPE=, MT=

Default: ALL

Results: CIMPORT Procedure

A common problem when you create or import a transport le under the z/OS
environment is a failure to specify the correct Data Control Block (DCB) characteristics.
When you reference a transport le you must specify the following DCB characteristics:
LRECL: 80
BLKSIZE: 8000
RECFM: FB
Note: A BLKSIZE value of less than 8000 may be more efcient for your storage
device in some cases. The BLKSIZE value must be an exact multiple of the LRECL
value. 4
Another common problem can occur if you use communications software to move les
from another environment to z/OS. In some cases, the transport le does not have the
proper DCB characteristics when it arrives on z/OS. If the communications software
does not allow you to specify le characteristics, try the following approach for z/OS:
1 Create a le under z/OS with the correct DCB characteristics and initialize the le.

222

Examples: CIMPORT Procedure

Chapter 8

2 Move the transport le from the other environment to the newly created le under

z/OS using binary transfer.

Examples: CIMPORT Procedure

Example 1: Importing an Entire Data Library

Procedure features:

PROC CIMPORT statement option:

INFILE=

This example shows how to use PROC CIMPORT to read from disk a transport le,
named TRANFILE, that PROC CPORT created from a SAS data library in another
operating environment. The transport le was moved to the new operating environment
by means of communications software or magnetic medium. PROC CIMPORT imports
the transport le to a SAS data library, called NEWLIB, in the new operating
environment.

Program

Specify the library name and lename. The LIBNAME statement species a libname for
the new SAS data library. The FILENAME statement species the lename of the transport le
that PROC CPORT created and enables you to specify any operating environment options for
le characteristics.
libname newlib SAS-data-library;
filename tranfile transport-file
host-option(s)-for-file-characteristics;

Import the SAS data library in the NEWLIB library. PROC CIMPORT imports the SAS
data library into the library named NEWLIB.
proc cimport library=newlib infile=tranfile;
run;

The CIMPORT Procedure

Program

223

SAS Log
NOTE: Proc
NOTE: Entry
NOTE: Entry
NOTE: Entry
NOTE: Entry
NOTE: Entry
NOTE: Total
NOTE:
NOTE:
NOTE:
NOTE:

CIMPORT begins to create/update catalog NEWLIB.FINANCE

LOAN.FRAME has been imported.
LOAN.HELP has been imported.
LOAN.KEYS has been imported.
LOAN.PMENU has been imported.
LOAN.SCL has been imported.
number of entries processed in catalog NEWLIB.FINANCE: 5

Proc CIMPORT begins to create/update catalog NEWLIB.FORMATS

Entry REVENUE.FORMAT has been imported.
Entry DEPT.FORMATC has been imported.
Total number of entries processed in catalog NEWLIB.FORMATS: 2

Example 2: Importing Individual Catalog Entries

Procedure features:

PROC CIMPORT statement options:

INFILE=
SELECT statement

This example shows how to use PROC CIMPORT to import the individual catalog
entries LOAN.PMENU and LOAN.SCL from the transport le TRANS2, which was
created from a single SAS catalog.

Program

Specify the library name, lename, and operating environment options. The LIBNAME
statement species a libname for the new SAS data library. The FILENAME statement species
the lename of the transport le that PROC CPORT created and enables you to specify any
operating environment options for le characteristics.
libname newlib SAS-data-library;
filename trans2 transport-file
host-option(s)-for-file-characteristics;

Import the specied catalog entries to the new SAS catalog. PROC CIMPORT imports
the individual catalog entries from the TRANS2 transport le and stores them in a new SAS
catalog called NEWLIB.FINANCE. The SELECT statement selects only the two specied
entries from the transport le to be imported into the new catalog.
proc cimport catalog=newlib.finance infile=trans2;
select loan.pmenu loan.scl;
run;

224

SAS Log

Chapter 8

SAS Log
NOTE:
NOTE:
NOTE:
NOTE:

Proc CIMPORT begins to create/update catalog NEWLIB.FINANCE

Entry LOAN.PMENU has been imported.
Entry LOAN.SCL has been imported.
Total number of entries processed in catalog NEWLIB.FINANCE: 2

Example 3: Importing a Single Indexed SAS Data Set

Procedure features:

PROC CIMPORT statement option:

INFILE=

This example shows how to use PROC CIMPORT to import an indexed SAS data set
from a transport le that was created by PROC CPORT from a single SAS data set.

Program

Specify the library name, lename, and operating environment options. The LIBNAME
statement species a libname for the new SAS data library. The FILENAME statement species
the lename of the transport le that PROC CPORT created and enables you to specify any
operating environment options for le characteristics.
libname newdata SAS-data-library;
filename trans3 transport-file
host-option(s)-for-file-characteristics;

Import the SAS data set. PROC CIMPORT imports the single SAS data set that you identify
with the DATA= specication in the PROC CIMPORT statement. PROC CPORT exported the
data set NEWDATA.TIMES in the transport le TRANS3.
proc cimport data=newdata.times infile=trans3;
run;

SAS Log
NOTE: Proc CIMPORT begins to create/update data set NEWDATA.TIMES
NOTE: The data set index x is defined.
NOTE: Data set contains 2 variables and 2 observations.
Logical record length is 16

225

CHAPTER

9
The COMPARE Procedure
Overview: COMPARE Procedure 226
What Does the COMPARE Procedure Do? 226
What Information Does PROC COMPARE Provide? 226
How Can PROC COMPARE Output Be Customized? 227
Syntax: COMPARE Procedure 229
PROC COMPARE Statement 229
BY Statement 236
ID Statement 237
VAR Statement 239
WITH Statement 239
Concepts: COMPARE Procedure 240
Comparisons Using PROC COMPARE 240
A Comparison by Position of Observations 240
A Comparison with an ID Variable 241
The Equality Criterion 242
Using the CRITERION= Option 242
Denition of Difference and Percent Difference 244
How PROC COMPARE Handles Variable Formats 244
Results: COMPARE Procedure 244
Results Reporting 244
SAS Log 244
Macro Return Codes (SYSINFO) 244
Procedure Output 246
Procedure Output Overview 246
Data Set Summary 246
Variables Summary 246
Observation Summary 247
Values Comparison Summary 248
Value Comparison Results 249
Table of Summary Statistics 250
Comparison Results for Observations (Using the TRANSPOSE Option) 252
ODS Table Names 253
Output Data Set (OUT=) 254
Output Statistics Data Set (OUTSTATS=) 255
Examples: COMPARE Procedure 256
Example 1: Producing a Complete Report of the Differences 256
Example 2: Comparing Variables in Different Data Sets 261
Example 3: Comparing a Variable Multiple Times 263
Example 4: Comparing Variables That Are in the Same Data Set 264
Example 5: Comparing Observations with an ID Variable 266
Example 6: Comparing Values of Observations Using an Output Data Set (OUT=) 270

226

Overview: COMPARE Procedure

Chapter 9

Example 7: Creating an Output Data Set of Statistics (OUTSTATS=)

273

Overview: COMPARE Procedure

What Does the COMPARE Procedure Do?
The COMPARE procedure compares the contents of two SAS data sets, selected
variables in different data sets, or variables within the same data set.
PROC COMPARE compares two data sets: the base data set and the comparison
data set. The procedure determines matching variables and matching observations.
Matching variables are variables with the same name or variables that you explicitly
pair by using the VAR and WITH statements. Matching variables must be of the same
type. Matching observations are observations that have the same values for all ID
variables that you specify or, if you do not use the ID statement, that occur in the same
position in the data sets. If you match observations by ID variables, then both data sets
must be sorted by all ID variables.

What Information Does PROC COMPARE Provide?

PROC COMPARE generates the following information about the two data sets that
are being compared:

3
3
3
3
3
3

whether matching variables have different values

whether one data set has more observations than the other
what variables the two data sets have in common
how many variables are in one data set but not in the other
whether matching variables have different formats, labels, or types.
a comparison of the values of matching observations.

Further, PROC COMPARE creates two kinds of output data sets that give detailed
information about the differences between observations of variables it is comparing.
The following example compares the data sets PROCLIB.ONE and PROCLIB.TWO,
which contain similar data about students:
data proclib.one(label=First Data Set);
input student year $ state $ gr1 gr2;
label year=Year of Birth;
format gr1 4.1;
datalines;
1000 1970 NC 85 87
1042 1971 MD 92 92
1095 1969 PA 78 72
1187 1970 MA 87 94
;
data proclib.two(label=Second Data Set);
input student $ year $ state $ gr1
gr2 major $;
label state=Home State;
format gr1 5.2;
datalines;

The COMPARE Procedure

1000
1042
1095
1187
1204
;

1970
1971
1969
1970
1971

NC
MA
PA
MD
NC

84
92
79
87
82

87
92
73
74
96

How Can PROC COMPARE Output Be Customized?

227

Math
History
Physics
Dance
French

How Can PROC COMPARE Output Be Customized?

PROC COMPARE produces lengthy output. You can use one or more options to
determine the kinds of comparisons to make and the degree of detail in the report. For
example, in the following PROC COMPARE step, the NOVALUES option suppresses
the part of the output that shows the differences in the values of matching variables:
proc compare base=proclib.one
compare=proclib.two novalues;
run;

Output 9.1 Comparison of Two Data Sets

The SAS System

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Data Set Summary
Dataset

Created

PROCLIB.ONE
PROCLIB.TWO

Modified

NVar

NObs

13MAY98:15:01:42
13MAY98:15:01:44

5
6

4
5

Label
First Data Set
Second Data Set

Variables Summary
Number
Number
Number
Number

of
of
of
of

Variables
Variables
Variables
Variables

in Common: 5.
in PROCLIB.TWO but not in PROCLIB.ONE: 1.
with Conflicting Types: 1.
with Differing Attributes: 3.

Listing of Common Variables with Conflicting Types

Variable

Dataset

Type

Length

student

PROCLIB.ONE
PROCLIB.TWO

Num
Char

8
8

Listing of Common Variables with Differing Attributes

Variable

Dataset

Type

Length

year

PROCLIB.ONE
PROCLIB.TWO
PROCLIB.ONE
PROCLIB.TWO

Char
Char
Char
Char

8
8
8
8

state

Format

Label
Year of Birth

Home State

228

How Can PROC COMPARE Output Be Customized?

Chapter 9

The SAS System

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Listing of Common Variables with Differing Attributes
Variable

Dataset

Type

gr1

PROCLIB.ONE
PROCLIB.TWO

Num
Num

Length
8
8

Format

Label

4.1
5.2

Observation Summary
Observation
First
First
Last
Last
Last

Obs
Unequal
Unequal
Match
Obs

Base

Compare

1
1
4
4
.

1
1
4
4
5

Number of Observations in Common: 4.

Number of Observations in PROCLIB.TWO but not in PROCLIB.ONE: 1.
Total Number of Observations Read from PROCLIB.ONE: 4.
Total Number of Observations Read from PROCLIB.TWO: 5.
Number of Observations with Some Compared Variables Unequal: 4.
Number of Observations with All Compared Variables Equal: 0.

The SAS System

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Values Comparison Summary
Number of Variables Compared with All Observations Equal: 1.
Number of Variables Compared with Some Observations Unequal: 3.
Total Number of Values which Compare Unequal: 6.
Maximum Difference: 20.

Variables with Unequal Values

Variable

Type

Len

state
gr1
gr2

CHAR
NUM
NUM

8
8
8

Compare Label
Home State

Ndif

MaxDif

2
2
2

1.000
20.000

Procedure Output on page 246 shows the default output for these two data sets.
Example 1 on page 256 shows the complete output for these two data sets.

The COMPARE Procedure

PROC COMPARE Statement

229

Syntax: COMPARE Procedure

You must use the VAR statement when you use the WITH statement.
Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Names: See: ODS Table Names on page 253
Reminder: You can use the LABEL, ATTRIB, FORMAT, and WHERE statements. See
Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.
Restriction:
Tip:

PROC COMPARE <option(s)>;

BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
ID <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
VAR variable(s);
WITH variable(s);

To do this

Use this statement

Produce a separate comparison for each BY group

Identify variables to use to match observations

Restrict the comparison to values of specic variables

VAR

Compare variables of different names

WITH and VAR

Compare two variables in the same data set

WITH and VAR

PROC COMPARE Statement

If you omit COMPARE=, then you must use the WITH and VAR statements.
Restriction: PROC COMPARE reports errors differently if one or both of the compared
data sets are not RADIX addressable. Version 6 compressed les are not RADIX
addressable, while, beginning with Version 7, compressed les are RADIX addressable.
(The integrity of the data is not compromised; the procedure simply numbers the
observations differently.)
Reminder: You can use data set options with the BASE= and COMPARE= options.
Restriction:

PROC COMPARE <option(s)>;

230

PROC COMPARE Statement

Chapter 9

To do this

Use this option

Specify the data sets to compare

Specify the base data set

BASE=

Specify the comparison data set

COMPARE=

Control the output data set

Create an output data set

OUT=

Write an observation for each observation in the BASE=

and COMPARE= data sets

OUTALL

Write an observation for each observation in the BASE=

data set

OUTBASE

Write an observation for each observation in the

COMPARE= data set

OUTCOMP

Write an observation that contains the differences for

each pair of matching observations

OUTDIF

Suppress the writing of observations when all values

are equal

OUTNOEQUAL

Write an observation that contains the percent

differences for each pair of matching observations

OUTPERCENT

Create an output data set that contains summary statistics

OUTSTATS=

Specify how the values are compared

Specify the criterion for judging the equality of numeric
values

CRITERION=

Specify the method for judging the equality of numeric

values

METHOD=

Judge missing values equal to any value

NOMISSBASE and NOMISSCOMP

Control the details in the default report

Include the values for all matching observations

ALLOBS

Print a table of summary statistics for all pairs of

matching variables

ALLSTATS and STATS

Include in the report the values and differences for all

matching variables

ALLVARS

Print only a short comparison summary

BRIEFSUMMARY

Change the report for numbers between 0 and 1

FUZZ=

Restrict the number of differences to print

MAXPRINT=

Suppress the print of creation and last-modied dates

NODATE

Suppress all printed output

NOPRINT

Suppress the summary reports

NOSUMMARY

Suppress the value comparison results.

NOVALUES

Produce a complete listing of values and differences

PRINTALL

The COMPARE Procedure

To do this
Print the value differences by observation, not by
variable

PROC COMPARE Statement

231

Use this option

TRANSPOSE

Control the listing of variables and observations

List all variables and observations found in only one
data set

LISTALL

List all variables and observations found only in the

base data set

LISTBASE

List all observations found only in the base data set

LISTBASEOBS

List all variables found only in the base data set

LISTBASEVAR

List all variables and observations found only in the

comparison data set

LISTCOMP

List all observations found only in the comparison data

set

LISTCOMPOBS

List all variables found only in the comparison data set

LISTCOMPVAR

List variables whose values are judged equal

LISTEQUALVAR

List all observations found in only one data set

LISTOBS

List all variables found in only one data set

LISTVAR

Options
ALLOBS

includes in the report of value comparison results the values and, for numeric
variables, the differences for all matching observations, even if they are judged equal.
Default: If you omit ALLOBS, then PROC COMPARE prints values only for

observations that are judged unequal.

Interaction: When used with the TRANSPOSE option, ALLOBS invokes the

ALLVARS option and displays the values for all matching observations and
variables.
ALLSTATS

prints a table of summary statistics for all pairs of matching variables.

See also: Table of Summary Statistics on page 250 for information on the

statistics produced
ALLVARS

includes in the report of value comparison results the values and, for numeric
variables, the differences for all pairs of matching variables, even if they are judged
equal.
Default: If you omit ALLVARS, then PROC COMPARE prints values only for

variables that are judged unequal.

Interaction: When used with the TRANSPOSE option, ALLVARS displays unequal

values in context with the values for other matching variables. If you omit the
TRANSPOSE option, then ALLVARS invokes the ALLOBS option and displays the
values for all matching observations and variables.

232

PROC COMPARE Statement

Chapter 9

BASE=SAS-data-set

species the data set to use as the base data set.

DATA=

Alias:

Default: the most recently created SAS data set

You can use the WHERE= data set option with the BASE= option to limit the
observations that are available for comparison.

Tip:

BRIEFSUMMARY

produces a short comparison summary and suppresses the four default summary
reports (data set summary report, variables summary report, observation summary
report, and values comparison summary report).
BRIEF

Alias:

By default, a listing of value differences accompanies the summary reports. To

suppress this listing, use the NOVALUES option.

Tip:

Featured in:

Example 4 on page 264

COMPARE=SAS-data-set

species the data set to use as the comparison data set.

Aliases:

COMP=, C=

Default: If you omit COMPARE=, then the comparison data set is the same as the

base data set, and PROC COMPARE compares variables within the data set.
Restriction: If you omit COMPARE=, then you must use the WITH statement.

You can use the WHERE= data set option with COMPARE= to limit the
observations that are available for comparison.

Tip:

CRITERION=

species the criterion for judging the equality of numeric values. Normally, the value
of (gamma) is positive, in which case the number itself becomes the equality
criterion. If you use a negative value for , then PROC COMPARE uses an equality
criterion proportional to the precision of the computer on which SAS is running.
Default: 0.00001
See also: The Equality Criterion on page 242 for more information
ERROR

displays an error message in the SAS log when differences are found.
Interaction: This option overrides the WARNING option.
FUZZ=number

alters the values comparison results for numbers less than number. PROC
COMPARE prints

3 0 for any variable value that is less than number

3 a blank for difference or percent difference if it is less than number
3 0 for any summary statistic that is less than number.
Default 0
Range:
Tip:

0-1

A report that contains many trivial differences is easier to read in this form.

LISTALL

lists all variables and observations that are found in only one data set.
Alias

LIST

Interaction: using LISTALL is equivalent to using the following four options:

LISTBASEOBS, LISTCOMPOBS, LISTBASEVAR, and LISTCOMPVAR.

The COMPARE Procedure

PROC COMPARE Statement

233

LISTBASE

lists all observations and variables that are found in the base data set but not in the
comparison data set.
Interaction: Using LISTBASE is equivalent to using the LISTBASEOBS and
LISTBASEVAR options.
LISTBASEOBS

lists all observations that are found in the base data set but not in the comparison
data set.
LISTBASEVAR

lists all variables that are found in the base data set but not in the comparison data
set.
LISTCOMP

lists all observations and variables that are found in the comparison data set but not
in the base data set.
Interaction: Using LISTCOMP is equivalent to using the LISTCOMPOBS and
LISTCOMPVAR options.
LISTCOMPOBS

lists all observations that are found in the comparison data set but not in the base
data set.
LISTCOMPVAR

lists all variables that are found in the comparison data set but not in the base data
set.
LISTEQUALVAR

prints a list of variables whose values are judged equal at all observations in addition
to the default list of variables whose values are judged unequal.
LISTOBS

lists all observations that are found in only one data set.
Interaction: Using LISTOBS is equivalent to using the LISTBASEOBS and
LISTCOMPOBS options.
LISTVAR

lists all variables that are found in only one data set.
Interaction: Using LISTVAR is equivalent to using both the LISTBASEVAR and
LISTCOMPVAR options.
MAXPRINT=total | (per-variable, total)

species the maximum number of differences to print, where

total
is the maximum total number of differences to print. The default value is 500
unless you use the ALLOBS option (or both the ALLVAR and TRANSPOSE
options), in which case the default is 32000.
per-variable
is the maximum number of differences to print for each variable within a BY
group. The default value is 50 unless you use the ALLOBS option (or both the
ALLVAR and TRANSPOSE options), in which case the default is 1000.
The MAXPRINT= option prevents the output from becoming extremely large when
data sets differ greatly.
METHOD=ABSOLUTE | EXACT | PERCENT | RELATIVE<()>

species the method for judging the equality of numeric values. The constant
(delta) is a number between 0 and 1 that species a value to add to the denominator
when calculating the equality measure. By default, is 0.

234

PROC COMPARE Statement

Chapter 9

Unless you use the CRITERION= option, the default method is EXACT. If you use
the CRITERION= option, then the default method is RELATIVE(), where (phi) is
a small number that depends on the numerical precision of the computer on which
SAS is running and on the value of CRITERION=.
See also: The Equality Criterion on page 242
NODATE

suppresses the display in the data set summary report of the creation dates and the
last modied dates of the base and comparison data sets.
NOMISSBASE

judges a missing value in the base data set equal to any value. (By default, a missing
value is equal only to a missing value of the same kind, that is .=., .^=.A, .A=.A,
.A^=.B, and so on.)
You can use this option to determine the changes that would be made to the
observations in the comparison data set if it were used as the master data set and
the base data set were used as the transaction data set in a DATA step UPDATE
statement. For information on the UPDATE statement, see the chapter on SAS
language statements in SAS Language Reference: Dictionary.
NOMISSCOMP

judges a missing value in the comparison data set equal to any value. (By default, a
missing value is equal only to a missing value of the same kind, that is .=., .^=.A,
.A=.A, .A^=.B, and so on.)
You can use this option to determine the changes that would be made to the
observations in the base data set if it were used as the master data set and the
comparison data set were used as the transaction data set in a DATA step UPDATE
statement. For information on the UPDATE statement, see the chapter on SAS
language statements in SAS Language Reference: Dictionary.
NOMISSING

judges missing values in both the base and comparison data sets equal to any value.
By default, a missing value is only equal to a missing value of the same kind, that is
.=., .^=.A, .A=.A, .A^=.B, and so on.
Alias:

NOMISS

Interaction: Using NOMISSING is equivalent to using both NOMISSBASE and

NOMISSCOMP.
NOPRINT

suppresses all printed output.

Tip: You may want to use this option when you are creating one or more output
data sets.
Featured in:

Example 6 on page 270

NOSUMMARY

suppresses the data set, variable, observation, and values comparison summary
reports.
NOSUMMARY produces no output if there are no differences in the matching
values.
Featured in: Example 2 on page 261
Tips:

NOTE

displays notes in the SAS log that describe the results of the comparison, whether or
not differences were found.
NOVALUES

suppresses the report of the value comparison results.

The COMPARE Procedure

Featured in:

PROC COMPARE Statement

235

Overview: COMPARE Procedure on page 226

OUT=SAS-data-set

names the output data set. If SAS-data-set does not exist, then PROC COMPARE
creates it. SAS-data-set contains the differences between matching variables.
See also: Output Data Set (OUT=) on page 254
Featured in: Example 6 on page 270
OUTALL

writes an observation to the output data set for each observation in the base data set
and for each observation in the comparison data set. The option also writes
observations to the output data set that contains the differences and percent
differences between the values in matching observations.
Tip: Using OUTALL is equivalent to using the following four options: OUTBASE,
OUTCOMP, OUTDIF, and OUTPERCENT.
See also: Output Data Set (OUT=) on page 254
OUTBASE

writes an observation to the output data set for each observation in the base data set,
creating observations in which _TYPE_=BASE.
See also: Output Data Set (OUT=) on page 254
Featured in: Example 6 on page 270
OUTCOMP

writes an observation to the output data set for each observation in the comparison
data set, creating observations in which _TYPE_=COMP.
See also: Output Data Set (OUT=) on page 254
Featured in: Example 6 on page 270
OUTDIF

writes an observation to the output data set for each pair of matching observations.
The values in the observation include values for the differences between the values
in the pair of observations. The value of _TYPE_ in each observation is DIF.
Default: The OUTDIF option is the default unless you specify the OUTBASE,
OUTCOMP, or OUTPERCENT option. If you use any of these options, then you
must explicitly specify the OUTDIF option to create _TYPE_=DIF observations in
the output data set.
See also: Output Data Set (OUT=) on page 254
Featured in: Example 6 on page 270
OUTNOEQUAL

suppresses the writing of an observation to the output data set when all values in
the observation are judged equal. In addition, in observations containing values for
some variables judged equal and others judged unequal, the OUTNOEQUAL option
uses the special missing value ".E" to represent differences and percent differences
for variables judged equal.
See also: Output Data Set (OUT=) on page 254
Featured in: Example 6 on page 270
OUTPERCENT

writes an observation to the output data set for each pair of matching observations.
The values in the observation include values for the percent differences between the
values in the pair of observations. The value of _TYPE_ in each observation is
PERCENT.
See also: Output Data Set (OUT=) on page 254

236

BY Statement

Chapter 9

OUTSTATS=SAS-data-set

writes summary statistics for all pairs of matching variables to the specied
SAS-data-set.
If you want to print a table of statistics in the procedure output, then use the
STATS, ALLSTATS, or PRINTALL option.
See also: Output Statistics Data Set (OUTSTATS=) on page 255 and Table of
Summary Statistics on page 250.
Tip:

Featured in:

Example 7 on page 273

PRINTALL

invokes the following options: ALLVARS, ALLOBS, ALLSTATS, LISTALL, and

WARNING.
Featured in:

Example 1 on page 256

STATS

prints a table of summary statistics for all pairs of matching numeric variables that
are judged unequal.
See also: Table of Summary Statistics on page 250 for information on the

statistics produced.
TRANSPOSE

prints the reports of value differences by observation instead of by variable.

Interaction: If you also use the NOVALUES option, then the TRANSPOSE option
lists only the names of the variables whose values are judged unequal for each
observation, not the values and differences.
See also: Comparison Results for Observations (Using the TRANSPOSE Option)
on page 252.
WARNING

displays a warning message in the SAS log when differences are found.
Interaction: The ERROR option overrides the WARNING option.

BY Statement
Produces a separate comparison for each BY group.
Main discussion: BY on page 58

BY < DESCENDING> variable-1

<< DESCENDING> variable-n>
<NOTSORTED>;

Required Arguments
variable

species the variable that the procedure uses to form BY groups. You can specify
more than one variable. If you do not use the NOTSORTED option in the BY
statement, then the observations in the data set must be sorted by all the variables
that you specify. Variables in a BY statement are called BY variables.

The COMPARE Procedure

ID Statement

237

Options
DESCENDING

species that the observations are sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The observations are grouped in another way, for example, chronological order.
The requirement for ordering observations according to the values of BY variables is
suspended for BY-group processing when you use the NOTSORTED option. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations with the same values for the BY variables
are not contiguous, then the procedure treats each contiguous set as a separate BY
group.

BY Processing with PROC COMPARE

To use a BY statement with PROC COMPARE, you must sort both the base and
comparison data sets by the BY variables. The nature of the comparison depends on
whether all BY variables are in the comparison data set and, if they are, whether their
attributes match those of the BY variables in the base data set. The following table
shows how PROC COMPARE behaves under different circumstances:
Condition

Behavior of PROC COMPARE

All BY variables are in the comparison

data set and all attributes match exactly

Compares corresponding BY groups

None of the BY variables are in the

comparison data set

Compares each BY group in the base data set with

the entire comparison data set

Some BY variables are not in the

comparison data set

Writes an error message to the SAS log and

terminates

Some BY variables have different types in

the two data sets

Writes an error message to the SAS log and

terminates

ID Statement
Lists variables to use to match observations.
See also: A Comparison with an ID Variable on page 241
Featured in: Example 5 on page 266

ID <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;

238

ID Statement

Chapter 9

Required Arguments
variable

species the variable that the procedure uses to match observations. You can specify
more than one variable, but the data set must be sorted by the variable or variables
you specify. These variables are ID variables. ID variables also identify observations
on the printed reports and in the output data set.

Options
DESCENDING

species that the data set is sorted in descending order by the variable that
immediately follows the word DESCENDING in the ID statement.
If you use the DESCENDING option, then you must sort the data sets. SAS does
not use an index to process an ID statement with the DESCENDING option.
Further, the use of DESCENDING for ID variables must correspond to the use of the
DESCENDING option in the BY statement in the PROC SORT step that was used to
sort the data sets.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data are grouped in another way, for example, chronological order.
See also: Comparing Unsorted Data on page 238

Requirements for ID Variables

3 ID variables must be in the BASE= data set or PROC COMPARE stops processing.
3 If an ID variable is not in the COMPARE= data set, then PROC COMPARE writes
a warning message to the SAS log and does not use that variable to match
observations in the comparison data set (but does write it to the OUT= data set).
3 ID variables must be of the same type in both data sets.
3 You should sort both data sets by the common ID variables (within the BY
variables, if any) unless you specify the NOTSORTED option.

Comparing Unsorted Data

If you do not want to sort the data set by the ID variables, then you can use the
NOTSORTED option. When you specify the NOTSORTED option, or if the ID
statement is omitted, PROC COMPARE matches the observations one-to-one. That is,
PROC COMPARE matches the rst observation in the base data set with the rst
observation in the comparison data set, the second with the second, and so on. If you
use NOTSORTED, and the ID values of corresponding observations are not the same,
then PROC COMPARE prints an error message and stops processing.
If the data sets are not sorted by the common ID variables and if you do not specify
the NOTSORTED option, then PROC COMPARE writes a warning message to the SAS
log and continues to process the data sets as if you had specied NOTSORTED.

Avoiding Duplicate ID Values

The observations in each data set should be uniquely labeled by the values of the ID
variables. If PROC COMPARE nds two successive observations with the same ID
values in a data set, then it

The COMPARE Procedure

WITH Statement

239

3 prints the warning Duplicate Observations for the rst occurrence for that data
set

3 prints the total number of duplicate observations found in the data set in the
observation summary report

3 uses the rst observation with the duplicate value for the comparison.
When the data sets are not sorted, PROC COMPARE detects only those duplicate
observations that occur in succession.

VAR Statement
Restricts the comparison of the values of variables to those named in the VAR statement.
Example 2 on page 261, Example 3 on page 263, and Example 4 on page 264

Featured in:

VAR variable(s);

Required Arguments
variable(s)

one or more variables that appear in the BASE= and COMPARE= data sets or only
in the BASE= data set.

Details
3 If you do not use the VAR statement, then PROC COMPARE compares the values
of all matching variables except those appearing in BY and ID statements.

3 If a variable in the VAR statement does not exist in the COMPARE= data set, then
PROC COMPARE writes a warning message to the SAS log and ignores the
variable.

3 If a variable in the VAR statement does not exist in the BASE= data set, then
PROC COMPARE stops processing and writes an error message to the SAS log.

3 The VAR statement restricts only the comparison of values of matching variables.
PROC COMPARE still reports on the total number of matching variables and
compares their attributes. However, it produces neither error nor warning
messages about these variables.

WITH Statement
Compares variables in the base data set with variables that have different names in the
comparison data set, and compares different variables that are in the same data set.
Restriction:

You must use the VAR statement when you use the WITH statement.

Featured in:

Example 2 on page 261, Example 3 on page 263, and Example 4 on page 264

240

Concepts: COMPARE Procedure

Chapter 9

WITH variable(s);

Required Arguments
variable(s)

one or more variables to compare with variables in the VAR statement.

Comparing Selected Variables

If you want to compare variables in the base data set with variables that have
different names in the comparison data set, then specify the names of the variables in
the base data set in the VAR statement and specify the names of the matching
variables in the WITH statement. The rst variable that you list in the WITH
statement corresponds to the rst variable that you list in the VAR statement, the
second with the second, and so on. If the WITH statement list is shorter than the VAR
statement list, then PROC COMPARE assumes that the extra variables in the VAR
statement have the same names in the comparison data set as they do in the base data
set. If the WITH statement list is longer than the VAR statement list, then PROC
COMPARE ignores the extra variables.
A variable name can appear any number of times in the VAR statement or the WITH
statement. By selecting VAR and WITH statement lists, you can compare the variables
in any permutation.
If you omit the COMPARE= option in the PROC COMPARE statement, then you
must use the WITH statement. In this case, PROC COMPARE compares the values of
variables with different names in the BASE= data set.

Concepts: COMPARE Procedure

Comparisons Using PROC COMPARE

PROC COMPARE rst compares the following:
3 data set attributes (set by the data set options TYPE= and LABEL=).
3 variables. PROC COMPARE checks each variable in one data set to determine
whether it matches a variable in the other data set.
3 attributes (type, length, labels, formats, and informats) of matching variables.
3 observations. PROC COMPARE checks each observation in one data set to
determine whether it matches an observation in the other data set. PROC
COMPARE either matches observations by their position in the data sets or by the
values of the ID variable.
After making these comparisons, PROC COMPARE compares the values in the parts
of the data sets that match. PROC COMPARE either compares the data by the position
of observations or by the values of an ID variable.

A Comparison by Position of Observations

Figure 9.1 on page 241 shows two data sets. The data inside the shaded boxes shows
the part of the data sets that the procedure compares. Assume that variables with the
same names have the same type.

The COMPARE Procedure

A Comparison with an ID Variable

241

Figure 9.1 Comparison by the Positions of Observations

Data Set ONE
IDNUM

NAME

GENDER

GPA

2998

Bagwell

3.722

9866

Metcalf

3.342

2118

Gray

3.177

3847

Baglione

4.000

2342

Hall

3.574

Data Set TWO

IDNUM

NAME

GENDER

GPA

YEAR

2998

Bagwell

3.722

9866

Metcalf

3.342

2118

Gray

3.177

3847

Baglione

4.000

2342

Hall

3.574

7565

Gold

3.609

1755

Syme

3.883

When you use PROC COMPARE to compare data set TWO with data set ONE, the
procedure compares the rst observation in data set ONE with the rst observation in
data set TWO, and it compares the second observation in the rst data set with the
second observation in the second data set, and so on. In each observation that it
compares, the procedure compares the values of the IDNUM, NAME, GENDER, and
GPA.
The procedure does not report on the values of the last two observations or the
variable YEAR in data set TWO because there is nothing to compare them with in data
set ONE.

A Comparison with an ID Variable

In a simple comparison, PROC COMPARE uses the observation number to determine
which observations to compare. When you use an ID variable, PROC COMPARE uses
the values of the ID variable to determine which observations to compare. ID variables
should have unique values and must have the same type.
For the two data sets shown in Figure 9.2 on page 242, assume that IDNUM is an ID
variable and that IDNUM has the same type in both data sets. The procedure compares
the observations that have the same value for IDNUM. The data inside the shaded
boxes shows the part of the data sets that the procedure compares.

242

The Equality Criterion

Figure 9.2

Chapter 9

Comparison by the Value of the ID Variable

Data Set ONE

IDNUM

NAME

GENDER

GPA

2998

Bagwell

3.722

9866

Metcalf

3.342

2118

Gray

3.177

3847

Baglione

4.000

2342

Hall

3.574

Data Set TWO

IDNUM

NAME

GENDER

GPA

YEAR

2998

Bagwell

3.722

9866

Metcalf

3.342

2118

Gray

3.177

3847

Baglione

4.000

2342

Hall

3.574

7565

Gold

3.609

1755

Syme

3.883

The data sets contain three matching variables: NAME, GENDER, and GPA. They
also contain ve matching observations: the observations with values of 2998, 9866,
2118, 3847, and 2342 for IDNUM.
Data Set TWO contains two observations (IDNUM=7565 and IDNUM=1755) for
which data set ONE contains no matching observations. Similarly, no variable in data
set ONE matches the variable YEAR in data set TWO.
See Example 5 on page 266 for an example that uses an ID variable.

The Equality Criterion

Using the CRITERION= Option

The COMPARE procedure judges numeric values unequal if the magnitude of their
difference, as measured according to the METHOD= option, is greater than the value of
the CRITERION= option. PROC COMPARE provides four methods for applying
CRITERION=:

3 The EXACT method tests for exact equality.

3 The ABSOLUTE method compares the absolute difference to the value specied by
CRITERION=.

3 The RELATIVE method compares the absolute relative difference to the value
specied by CRITERION=.

3 The PERCENT method compares the absolute percent difference to the value
specied by CRITERION=.
For a numeric variable compared, let x be its value in the base data set and let y be
its value in the comparison data set. If both x and y are nonmissing, then the values

The COMPARE Procedure

The Equality Criterion

243

are judged unequal according to the value of METHOD= and the value of CRITERION=
( ) as follows:
3 If METHOD=EXACT, then the values are unequal if y does not equal x.
3 If METHOD=ABSOLUTE, then the values are unequal if

ABS (y 0 x) >
3 If METHOD=RELATIVE, then the values are unequal if

ABS (y 0 x) = ((ABS(x) + ABS(y)) =2 + ) >

The values are equal if x=y=0.

3 If METHOD=PERCENT, then the values are unequal if

100 (ABS (y 0 x) =ABS (x)) > for x 6= 0

y=0
6

for x = 0

If x or y is missing, then the comparison depends on the NOMISSING option. If the

NOMISSING option is in effect, then a missing value will always be judged equal to
anything. Otherwise, a missing value is judged equal only to a missing value of the
same type (that is, .=., .^=.A, .A=.A, .A^=.B, and so on).
If the value that is specied for CRITERION= is negative, then the actual criterion
that is used is made equal to the absolute value of times a very small number
(epsilon) that depends on the numerical precision of the computer. This number is
dened as the smallest positive oating-point value such that, using machine
arithmetic, 1<1<1+. Round-off or truncation error in oating-point computations is
typically a few orders of magnitude larger than . This means that CRITERION=1000
often provides a reasonable test of the equality of computed results at the machine level
of precision.
The value added to the denominator in the RELATIVE method is specied in
parentheses after the method name: METHOD=RELATIVE(). If not specied in
METHOD=, then defaults to 0. The value of can be used to control the behavior of
the error measure when both x and y are very close to 0. If is not given and x and y
are very close to 0, then any error produces a large relative error (in the limit, 2).
Specifying a value for avoids this extreme sensitivity of the RELATIVE method for
small values. If you specify METHOD=RELATIVE() CRITERION= when both x and y
are much smaller than in absolute value, then the comparison is as if you had
specied METHOD=ABSOLUTE CRITERION= . However, when either x or y is much
larger than in absolute value, the comparison is like METHOD=RELATIVE
CRITERION= . For moderate values of x and y, METHOD=RELATIVE()
CRITERION= is, in effect, a compromise between METHOD=ABSOLUTE
CRITERION= and METHOD=RELATIVE CRITERION= .
For character variables, if one value is longer than the other, then the shorter value
is padded with blanks for the comparison. Nonblank character values are judged equal
only if they agree at each character. If the NOMISSING option is in effect, then blank
character values are judged equal to anything.

244

How PROC COMPARE Handles Variable Formats

Chapter 9

Denition of Difference and Percent Difference

In the reports of value comparisons and in the OUT= data set, PROC COMPARE
displays difference and percent difference values for the numbers compared. These
quantities are dened using the value from the base data set as the reference value.
For a numeric variable compared, let x be its value in the base data set and let y be its
value in the comparison data set. If x and y are both nonmissing, then the difference
and percent difference are dened as follows:
Difference = y

Percent Difference = (y

0 x) =x 3 100

for x 6= 0

Percent Difference = missing for x = 0 :

How PROC COMPARE Handles Variable Formats

PROC COMPARE compares unformatted values. If you have two matching variables
that are formatted differently, then PROC COMPARE lists the formats of the variables.

Results: COMPARE Procedure

Results Reporting
PROC COMPARE reports the results of its comparisons in the following ways:

3
3
3
3

the SAS log

return codes stored in the automatic macro SYSINFO
procedure output
output data sets.

SAS Log
When you use the WARNING, PRINTALL, or ERROR option, PROC COMPARE
writes a description of the differences to the SAS log.

Macro Return Codes (SYSINFO)

PROC COMPARE stores a return code in the automatic macro variable SYSINFO.
The value of the return code provides information about the result of the comparison.
By checking the value of SYSINFO after PROC COMPARE has run and before any
other step begins, SAS macros can use the results of a PROC COMPARE step to
determine what action to take or what parts of a SAS program to execute.
Table 9.1 on page 245 is a key for interpreting the SYSINFO return code from PROC
COMPARE. For each of the conditions listed, the associated value is added to the
return code if the condition is true. Thus, the SYSINFO return code is the sum of the
codes listed in Table 9.1 on page 245 for the applicable conditions:

The COMPARE Procedure

Macro Return Codes (SYSINFO)

Table 9.1 Macro Return Codes

Bit

Condition

Code

Hex

Description

DSLABEL

0001X

Data set labels differ

DSTYPE

0002X

Data set types differ

INFORMAT

0004X

Variable has different informat

FORMAT

0008X

Variable has different format

LENGTH

0010X

Variable has different length

LABEL

base

VALUE

4096

1000X

A value comparison was unequal

TYPE

8192

2000X

Conicting variable types

BYVAR

16384

4000X

BY variables do not match

ERROR

32768

8000X

Fatal error: comparison not done

These codes are ordered and scaled to enable a simple check of the degree to which
the data sets differ. For example, if you want to check that two data sets contain the
same variables, observations, and values, but you do not care about differences in
labels, formats, and so forth, then use the following statements:
proc compare base=SAS-data-set
compare=SAS-data-set;
run;
%if &sysinfo >= 64 %then
%do;
handle error;
%end;

You can examine individual bits in the SYSINFO value by using DATA step
bit-testing features to check for specic conditions. For example, to check for the
presence of observations in the base data set that are not in the comparison data set,
use the following statements:
proc compare base=SAS-data-set
compare=SAS-data-set;
run;

245

246

Procedure Output

Chapter 9

%let rc=&sysinfo;
data _null_;
if &rc=1......b then
put Observations in Base but not
in Comparison Data Set;
run;

PROC COMPARE must run before you check SYSINFO and you must obtain the
SYSINFO value before another SAS step starts because every SAS step resets
SYSINFO.

Procedure Output
Procedure Output Overview
The following sections show and describe the default output of the two data sets
shown in Overview: COMPARE Procedure on page 226. Because PROC COMPARE
produces lengthy output, the output is presented in seven pieces.

Data Set Summary

This report lists the attributes of the data sets that are being compared. These
attributes include the following:
3 the data set names
3 the data set types, if any
3 the data set labels, if any
3 the dates created and last modied
3 the number of variables in each data set
3 the number of observations in each data set.
Output 9.2 shows the Data Set Summary.

Output 9.2

Partial Output
COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Data Set Summary

Dataset

PROCLIB.ONE
PROCLIB.TWO

Created

Modified

NVar

NObs

11SEP97:15:11:07
11SEP97:15:11:10

11SEP97:15:11:09
11SEP97:15:11:10

5
6

4
5

Label

First Data Set

Second Data Set

Variables Summary
This report compares the variables in the two data sets. The rst part of the report
lists the following:

The COMPARE Procedure

Procedure Output

247

3 the number of variables the data sets have in common

3 the number of variables in the base data set that are not in the comparison data
set and vice versa
3 the number of variables in both data sets that have different types
3 the number of variables that differ on other attributes (length, label, format, or
informat)
3 the number of BY, ID, VAR, and WITH variables specied for the comparison.
The second part of the report lists matching variables with different attributes and
shows how the attributes differ. (The COMPARE procedure omits variable labels if the
line size is too small for them.)
Output 9.3 shows the Variables Summary.

Output 9.3 Partial Output

Variables Summary
Number
Number
Number
Number

of
of
of
of

Variables
Variables
Variables
Variables

in Common: 5.
in PROCLIB.TWO but not in PROCLIB.ONE: 1.
with Conflicting Types: 1.
with Differing Attributes: 3.

Listing of Common Variables with Conflicting Types

Variable

Dataset

Type

Length

student

PROCLIB.ONE
PROCLIB.TWO

Num
Char

8
8

Listing of Common Variables with Differing Attributes

Variable

Dataset

Type

Length

year

PROCLIB.ONE
PROCLIB.TWO
PROCLIB.ONE
PROCLIB.TWO
PROCLIB.ONE
PROCLIB.TWO

Char
Char
Char
Char
Num
Num

8
8
8
8
8
8

state
gr1

Format

Label
Year of Birth

Home State
4.1
5.2

Observation Summary
This report provides information about observations in the base and comparison data
sets. First of all, the report identies the rst and last observation in each data set, the
rst and last matching observations, and the rst and last differing observations. Then,
the report lists the following:
3 the number of observations that the data sets have in common
3 the number of observations in the base data set that are not in the comparison
data set and vice versa
3 the total number of observations in each data set
3 the number of matching observations for which PROC COMPARE judged some
variables unequal
3 the number of matching observations for which PROC COMPARE judged all
variables equal.

248

Procedure Output

Chapter 9

Output 9.4 shows the Observation Summary.

Output 9.4

Partial Output
Observation Summary
Observation
First
First
Last
Last
Last

Obs
Unequal
Unequal
Match
Obs

Base

Compare

1
1
4
4
.

1
1
4
4
5

Number of Observations in Common: 4.

Values Comparison Summary

This report rst lists the following:

3
3
3
3
3

the number of variables compared with all observations equal

the number of variables compared with some observations unequal
the number of variables with differences involving missing values, if any
the total number of values judged unequal
the maximum difference measure between unequal values for all pairs of matching
variables (for differences not involving missing values).

In addition, for the variables for which some matching observations have unequal
values, the report lists
3 the name of the variable

3 other variable attributes

3 the number of times PROC COMPARE judged the variable unequal
3 the maximum difference measure found between values (for differences not
involving missing values)
3 the number of differences caused by comparison with missing values, if any.
Output 9.5 shows the Values Comparison Summary.

The COMPARE Procedure

Procedure Output

249

Output 9.5 Partial Output

Values Comparison Summary
Number of Variables Compared with All Observations Equal: 1.
Number of Variables Compared with Some Observations Unequal: 3.
Total Number of Values which Compare Unequal: 6.
Maximum Difference: 20.
Variables with Unequal Values
Variable

Type

Len

state
gr1
gr2

CHAR
NUM
NUM

8
8
8

Compare Label
Home State

Ndif

MaxDif

2
2
2

1.000
20.000

Value Comparison Results

This report consists of a table for each pair of matching variables judged unequal at
one or more observations. When comparing character values, PROC COMPARE
displays only the rst 20 characters. When you use the TRANSPOSE option, it displays
only the rst 12 characters. Each table shows
3 the number of the observation or, if you use the ID statement, the values of the ID
variables

3
3
3
3

the
the
the
the

value of the variable in the base data set

value of the variable in the comparison data set
difference between these two values (numeric variables only)
percent difference between these two values (numeric variables only).

Output 9.6 shows the Value Comparison Results for Variables.

250

Procedure Output

Output 9.6

Chapter 9

Partial Output
Value Comparison Results for Variables
__________________________________________________________
|| Home State
|| Base Value
Compare Value
Obs || state
state
________ || ________
________
||
2 || MD
MA
4 || MA
MD
__________________________________________________________

__________________________________________________________
||
Base
Compare
Obs ||
gr2
gr2
Diff.
% Diff
________ || _________ _________ _________ _________
||
3 ||
72.0000
73.0000
1.0000
1.3889
4 ||
94.0000
74.0000
-20.0000
-21.2766
__________________________________________________________

You can suppress the value comparison results with the NOVALUES option. If you
use both the NOVALUES and TRANSPOSE options, then PROC COMPARE lists for
each observation the names of the variables with values judged unequal but does not
display the values and differences.

Table of Summary Statistics

If you use the STATS, ALLSTATS, or PRINTALL option, then the Value Comparison
Results for Variables section contains summary statistics for the numeric variables that
are being compared. The STATS option generates these statistics for only the numeric
variables whose values are judged unequal. The ALLSTATS and PRINTALL options
generate these statistics for all numeric variables, even if all values are judged equal.
Note: In all cases PROC COMPARE calculates the summary statistics based on all
matching observations that do not contain missing values, not just on those containing
unequal values. 4
Output 9.7 shows the following summary statistics for base data set values,
comparison data set values, differences, and percent differences:
N
the number of nonmissing values
MEAN
the mean, or average, of the values
STD
the standard deviation

The COMPARE Procedure

Procedure Output

251

MAX
the maximum value
MIN
the minimum value
STDERR
the standard error of the mean
T
the T ratio (MEAN/STDERR)
PROB> | T |
the probability of a greater absolute T value if the true population mean is 0.
NDIF
the number of matching observations judged unequal, and the percent of the
matching observations that were judged unequal.
DIFMEANS
the difference between the mean of the base values and the mean of the
comparison values. This line contains three numbers. The rst is the mean
expressed as a percentage of the base values mean. The second is the mean
expressed as a percentage of the comparison values mean. The third is the
difference in the two means (the comparison mean minus the base mean).
R
the correlation of the base and comparison values for matching observations that
are nonmissing in both data sets.
RSQ
the square of the correlation of the base and comparison values for matching
observations that are nonmissing in both data sets.
Output 9.7 is from the ALLSTATS option using the two data sets shown in
Overview:

252

Procedure Output

Output 9.7

Chapter 9

Partial Output
Value Comparison Results for Variables
__________________________________________________________
||
Base
Compare
Obs ||
gr1
gr1
Diff.
% Diff
________ || _________ _________ _________ _________
||
1 ||
85.0
84.00
-1.0000
-1.1765
3 ||
78.0
79.00
1.0000
1.2821
________ || _________ _________ _________ _________
||
N
||
4
4
4
4
Mean
||
85.5000
85.5000
0
0.0264
Std
||
5.8023
5.4467
0.8165
1.0042
Max
||
92.0000
92.0000
1.0000
1.2821
Min
||
78.0000
79.0000
-1.0000
-1.1765
StdErr
||
2.9011
2.7234
0.4082
0.5021
t
||
29.4711
31.3951
0.0000
0.0526
Prob>|t| ||
<.0001
<.0001
1.0000
0.9614
||
Ndif
||
2
50.000%
DifMeans ||
0.000%
0.000%
0
r, rsq
||
0.991
0.983
__________________________________________________________
__________________________________________________________
||
Base
Compare
Obs ||
gr2
gr2
Diff.
% Diff
________ || _________ _________ _________ _________
||
3 ||
72.0000
73.0000
1.0000
1.3889
4 ||
94.0000
74.0000
-20.0000
-21.2766
________ || _________ _________ _________ _________
||
N
||
4
4
4
4
Mean
||
86.2500
81.5000
-4.7500
-4.9719
Std
||
9.9457
9.4692
10.1776
10.8895
Max
||
94.0000
92.0000
1.0000
1.3889
Min
||
72.0000
73.0000
-20.0000
-21.2766
StdErr
||
4.9728
4.7346
5.0888
5.4447
t
||
17.3442
17.2136
-0.9334
-0.9132
Prob>|t| ||
0.0004
0.0004
0.4195
0.4285
||
Ndif
||
2
50.000%
DifMeans ||
-5.507%
-5.828%
-4.7500
r, rsq
||
0.451
0.204
__________________________________________________________

Note: If you use a wide line size with PRINTALL, then PROC COMPARE prints the
value comparison result for character variables next to the result for numeric variables.
In that case, PROC COMPARE calculates only NDIF for the character variables. 4

Comparison Results for Observations (Using the TRANSPOSE Option)

The TRANSPOSE option prints the comparison results by observation instead of by
variable. The comparison results precede the observation summary report. By default,
the source of the values for each row of the table is indicated by the following label:
_OBS_1=number-1

_OBS_2=number-2

The COMPARE Procedure

ODS Table Names

253

where number-1 is the number of the observation in the base data set for which the
value of the variable is shown, and number-2 is the number of the observation in the
comparison data set.
Output 9.8 shows the differences in PROCLIB.ONE and PROCLIB.TWO by
observation instead of by variable.

Output 9.8 Partial Output

Comparison Results for Observations
_OBS_1=1 _OBS_2=1:
Variable
Base Value
gr1
85.0

Compare
84.00

_OBS_1=2 _OBS_2=2:
Variable
Base Value
state
MD

Compare
MA

_OBS_1=3 _OBS_2=3:
Variable
Base Value
gr1
78.0
gr2
72.000000
_OBS_1=4 _OBS_2=4:
Variable
Base Value
gr2
94.000000
state
MA

Diff.
-1.000000

% Diff
-1.176471

Compare
79.00
73.000000

Diff.
1.000000
1.000000

% Diff
1.282051
1.388889

Compare
74.000000
MD

Diff.
-20.000000

% Diff
-21.276596

If you use an ID statement, then the identifying label has the following form:
ID-1=ID-value-1 ...

ID-n=ID-value-n

where ID is the name of an ID variable and ID-value is the value of the ID variable.
Note: When you use the TRANSPOSE option, PROC COMPARE prints only the
rst 12 characters of the value. 4

ODS Table Names

The COMPARE procedure assigns a name to each table that it creates. You can use
these names to reference the table when using the Output Delivery System (ODS) to
select tables and create output data sets. For more information, see SAS Output
Delivery System: Users Guide.
Table 9.2 ODS Tables Produced by the COMPARE Procedure
Table Name

Description

Generated...

CompareDatasets

Information about the data set

or data sets

by default, unless
NOSUMMARY or NOVALUES
option is specied

CompareDetails (Comparison
Results for Observations)

A listing of observations that

the base data set and the
compare data set do not have
in common

if PRINTALL option is specied

254

Output Data Set (OUT=)

Chapter 9

Table Name

Description

Generated...

CompareDetails (ID variable

notes and warnings)

A listing of notes and warnings

concerning duplicate ID
variable values

if ID statement is specied and

duplicate ID variable values
exist in either data set

CompareDifferences

A report of variable value

differences

by default unless NOVALUES

option is specied

CompareSummary

Summary report of
observations, values, and
variables with unequal values

by default

CompareVariables

A listing of differences in
variable types or attributes
between the base data set and
the compare data set

by default, unless the variables

are identical or the
NOSUMMARY option is
specied

Output Data Set (OUT=)

By default, the OUT= data set contains an observation for each pair of matching
observations. The OUT= data set contains the following variables from the data sets
you are comparing:

3 all variables named in the BY statement

3 all variables named in the ID statement
3 all matching variables or, if you use the VAR statement, all variables listed in the
VAR statement.
In addition, the data set contains two variables created by PROC COMPARE to
identify the source of the values for the matching variables: _TYPE_ and _OBS_.
_TYPE_
is a character variable of length 8. Its value indicates the source of the values for
the matching (or VAR) variables in that observation. (For ID and BY variables,
which are not compared, the values are the values from the original data sets.)
_TYPE_ has the label Type of Observation. The four possible values of this
variable are as follows:
BASE
The values in this observation are from an observation in the base data set.
PROC COMPARE writes this type of observation to the OUT= data set when
you specify the OUTBASE option.
COMPARE
The values in this observation are from an observation in the comparison
data set. PROC COMPARE writes this type of observation to the OUT= data
set when you specify the OUTCOMP option.
DIF
The values in this observation are the differences between the values in the
base and comparison data sets. For character variables, PROC COMPARE
uses a period (.) to represent equal characters and an X to represent unequal
characters. PROC COMPARE writes this type of observation to the OUT=
data set by default. However, if you request any other type of observation
with the OUTBASE, OUTCOMP, or OUTPERCENT option, then you must
specify the OUTDIF option to generate observations of this type in the OUT=
data set.

The COMPARE Procedure

Output Statistics Data Set (OUTSTATS=)

255

PERCENT
The values in this observation are the percent differences between the values
in the base and comparison data sets. For character variables the values in
observations of type PERCENT are the same as the values in observations of
type DIF.
_OBS_
is a numeric variable that contains a number further identifying the source of the
OUT= observations.
For observations with _TYPE_ equal to BASE, _OBS_ is the number of the
observation in the base data set from which the values of the VAR variables were
copied. Similarly, for observations with _TYPE_ equal to COMPARE, _OBS_ is the
number of the observation in the comparison data set from which the values of the
VAR variables were copied.
For observations with _TYPE_ equal to DIF or PERCENT, _OBS_ is a sequence
number that counts the matching observations in the BY group.
_OBS_ has the label Observation Number.
The COMPARE procedure takes variable names and attributes for the OUT= data
set from the base data set except for the lengths of ID and VAR variables, for which it
uses the longer length regardless of which data set that length is from. This behavior
has two important repercussions:

3 If you use the VAR and WITH statements, then the names of the variables in the
OUT= data set come from the VAR statement. Thus, observations with _TYPE_
equal to BASE contain the values of the VAR variables, while observations with
_TYPE_ equal to COMPARE contain the values of the WITH variables.

3 If you include a variable more than once in the VAR statement in order to compare
it with more than one variable, then PROC COMPARE can include only the rst
comparison in the OUT= data set because each variable must have a unique name.
Other comparisons produce warning messages.
For an example of the OUT= option, see Example 6 on page 270.

Output Statistics Data Set (OUTSTATS=)

When you use the OUTSTATS= option, PROC COMPARE calculates the same
summary statistics as the ALLSTATS option for each pair of numeric variables
compared (see Table of Summary Statistics on page 250). The OUTSTATS= data set
contains an observation for each summary statistic for each pair of variables. The data
set also contains the BY variables used in the comparison and several variables created
by PROC COMPARE:
_VAR_
is a character variable that contains the name of the variable from the base data
set for which the statistic in the observation was calculated.
_WITH_
is a character variable that contains the name of the variable from the comparison
data set for which the statistic in the observation was calculated. The _WITH_
variable is not included in the OUTSTATS= data set unless you use the WITH
statement.
_TYPE_
is a character variable that contains the name of the statistic contained in the
observation. Values of the _TYPE_ variable are N, MEAN, STD, MIN, MAX, STDERR, T,
PROBT, NDIF, DIFMEANS, and R, RSQ.

256

Examples: COMPARE Procedure

Chapter 9

_BASE_
is a numeric variable that contains the value of the statistic calculated from the
values of the variable named by _VAR_ in the observations in the base data set
with matching observations in the comparison data set.
_COMP_
is a numeric variable that contains the value of the statistic calculated from the
values of the variable named by the _VAR_ variable (or by the _WITH_ variable if
you use the WITH statement) in the observations in the comparison data set with
matching observations in the base data set.
_DIF_
is a numeric variable that contains the value of the statistic calculated from the
differences of the values of the variable named by the _VAR_ variable in the base
data set and the matching variable (named by the _VAR_ or _WITH_ variable) in
the comparison data set.
_PCTDIF_
is a numeric variable that contains the value of the statistic calculated from the
percent differences of the values of the variable named by the _VAR_ variable in
the base data set and the matching variable (named by the _VAR_ or _WITH_
variable) in the comparison data set.
Note: For both types of output data sets, PROC COMPARE assigns one of the
following data set labels:
Comparison of base-SAS-data-set
with comparison-SAS-data-set
Comparison of variables in base-SAS-data-set

4
Labels are limited to 40 characters.
See Example 7 on page 273 for an example of an OUTSTATS= data set.

Examples: COMPARE Procedure

Example 1: Producing a Complete Report of the Differences

Procedure features:

PROC COMPARE statement options

BASE=
PRINTALL
COMPARE=
Data sets:

PROCLIB.ONE, PROCLIB.TWO on page 226

This example shows the most complete report that PROC COMPARE produces as
procedure output.

The COMPARE Procedure

Program

257

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Create a complete report of the differences between two data sets. BASE= and
COMPARE= specify the data sets to compare. PRINTALL prints a full report of the differences.
proc compare base=proclib.one compare=proclib.two printall;
title Comparing Two Data Sets: Full Report;
run;

258

Output

Chapter 9

Output

A > in the output marks information that is in the full report but not in the default report. The
additional information includes a listing of variables found in one data set but not the other, a
listing of observations found in one data set but not the other, a listing of variables with all
equal values, and summary statistics. For an explanation of the statistics, see Table of
Summary Statistics on page 250.

Comparing Two Data Sets: Full Report

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Data Set Summary
Dataset

Created

PROCLIB.ONE
PROCLIB.TWO

Modified

NVar

NObs

11SEP97:16:19:59
11SEP97:16:20:01

11SEP97:16:20:01
11SEP97:16:20:01

5
6

4
5

Label
First Data Set
Second Data Set

Variables Summary
Number
Number
Number
Number

of
of
of
of

Variables
Variables
Variables
Variables

in Common: 5.
in PROCLIB.TWO but not in PROCLIB.ONE: 1.
with Conflicting Types: 1.
with Differing Attributes: 3.

Listing of Variables in PROCLIB.TWO but not in PROCLIB.ONE

Variable
>

major

Type

Length

Char

Listing of Common Variables with Conflicting Types

Variable

Dataset

Type

Length

student

PROCLIB.ONE
PROCLIB.TWO

Num
Char

8
8

The COMPARE Procedure

Comparing Two Data Sets: Full Report

Output

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Listing of Common Variables with Differing Attributes
Variable

Dataset

Type

Length

year

PROCLIB.ONE
PROCLIB.TWO
PROCLIB.ONE
PROCLIB.TWO
PROCLIB.ONE
PROCLIB.TWO

Char
Char
Char
Char
Num
Num

8
8
8
8
8
8

state
gr1

Format

Label
Year of Birth

Home State
4.1
5.2

Comparison Results for Observations

Observation 5 in PROCLIB.TWO not found in PROCLIB.ONE.

Observation Summary
Observation
First
First
Last
Last
Last

Base
1
1
4
4
.

Obs
Unequal
Unequal
Match
Obs

Compare
1
1
4
4
5

Number of Observations in Common: 4.

Comparing Two Data Sets: Full Report

Variables with All Equal Values

Variable
year

Type
CHAR

Len
8

Label
Year of Birth

Variables with Unequal Values

Variable

Type

Len

state
gr1
gr2

CHAR
NUM
NUM

8
8
8

Compare Label
Home State

Ndif

MaxDif

2
2
2

1.000
20.000

259

260

Output

Chapter 9

Comparing Two Data Sets: Full Report

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Value Comparison Results for Variables
__________________________________________________________
|| Year of Birth
|| Base Value
Compare Value
Obs || year
year
________ || ________
________
||
1 || 1970
1970
2 || 1971
1971
3 || 1969
1969
4 || 1970
1970
__________________________________________________________

__________________________________________________________
|| Home State
|| Base Value
Compare Value
Obs || state
state
________ || ________
________
||
1 || NC
NC
2 || MD
MA
3 || PA
PA
4 || MA
MD
__________________________________________________________

Comparing Two Data Sets: Full Report

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Value Comparison Results for Variables

__________________________________________________________
||
Base
Compare
Obs ||
gr1
gr1
Diff.
% Diff
________ || _________ _________ _________ _________
||
1 ||
85.0
84.00
-1.0000
-1.1765
2 ||
92.0
92.00
0
0
3 ||
78.0
79.00
1.0000
1.2821
4 ||
87.0
87.00
0
0
________ || _________ _________ _________ _________
||
N
||
4
4
4
4
Mean
||
85.5000
85.5000
0
0.0264
Std
||
5.8023
5.4467
0.8165
1.0042
Max
||
92.0000
92.0000
1.0000
1.2821
Min
||
78.0000
79.0000
-1.0000
-1.1765
StdErr
||
2.9011
2.7234
0.4082
0.5021
t
||
29.4711
31.3951
0.0000
0.0526
Prob>|t| ||
<.0001
<.0001
1.0000
0.9614
||
Ndif
||
2
50.000%
DifMeans ||
0.000%
0.000%
0
r, rsq
||
0.991
0.983
__________________________________________________________

The COMPARE Procedure

Comparing Two Data Sets: Full Report

Program

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
Value Comparison Results for Variables
__________________________________________________________
||
Base
Compare
Obs ||
gr2
gr2
Diff.
% Diff
________ || _________ _________ _________ _________
||
1 ||
87.0000
87.0000
0
0
2 ||
92.0000
92.0000
0
0
3 ||
72.0000
73.0000
1.0000
1.3889
4 ||
94.0000
74.0000
-20.0000
-21.2766
________ || _________ _________ _________ _________
||
N
||
4
4
4
4
Mean
||
86.2500
81.5000
-4.7500
-4.9719
Std
||
9.9457
9.4692
10.1776
10.8895
Max
||
94.0000
92.0000
1.0000
1.3889
Min
||
72.0000
73.0000
-20.0000
-21.2766
StdErr
||
4.9728
4.7346
5.0888
5.4447
t
||
17.3442
17.2136
-0.9334
-0.9132
Prob>|t| ||
0.0004
0.0004
0.4195
0.4285
||
Ndif
||
2
50.000%
DifMeans ||
-5.507%
-5.828%
-4.7500
r, rsq
||
0.451
0.204
__________________________________________________________

Example 2: Comparing Variables in Different Data Sets

Procedure features:

PROC COMPARE statement option

NOSUMMARY
VAR statement
WITH statement
Data sets:

PROCLIB.ONE, PROCLIB.TWO on page 226.

This example compares a variable from the base data set with a variable in the
comparison data set. All summary reports are suppressed.

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

261

262

Output

Chapter 9

Suppress all summary reports of the differences between two data sets. BASE=
species the base data set and COMPARE= species the comparison data set. NOSUMMARY
suppresses all summary reports.
proc compare base=proclib.one compare=proclib.two nosummary;

Specify one variable from the base data set to compare with one variable from the
comparison data set. The VAR and WITH statements specify the variables to compare. This
example compares GR1 from the base data set with GR2 from the comparison data set.
var gr1;
with gr2;
title Comparison of Variables in Different Data Sets;
run;

Output
Comparison of Variables in Different Data Sets
COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
NOTE: Data set PROCLIB.TWO contains 1 observations not in PROCLIB.ONE.
NOTE: Values of the following 1 variables compare unequal: gr1^=gr2

Value Comparison Results for Variables

The COMPARE Procedure

Program

263

Example 3: Comparing a Variable Multiple Times

Procedure features:

VAR statement
WITH statement
Data sets:

PROCLIB.ONE, PROCLIB.TWO on page 226.

This example compares one variable from the base data set with two variables in the
comparison data set.

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Specify one variable from the base data set to compare with two variables from the
comparison data set. The VAR and WITH statements specify the variables to compare. This
example compares GR1 from the base data set with GR1 and GR2 from the comparison data set.
var gr1 gr1;
with gr1 gr2;
title Comparison of One Variable with Two Variables;
run;

264

Output

Chapter 9

Output

The Value Comparison Results section shows the result of the comparison.

Comparison of One Variable with Two Variables

COMPARE Procedure
Comparison of PROCLIB.ONE with PROCLIB.TWO
(Method=EXACT)
NOTE: Data set PROCLIB.TWO contains 1 observations not in PROCLIB.ONE.
NOTE: Values of the following 2 variables compare unequal: gr1^=gr1 gr1^=gr2

Value Comparison Results for Variables

Example 4: Comparing Variables That Are in the Same Data Set

Procedure features:

PROC COMPARE statement options

ALLSTATS
BRIEFSUMMARY
VAR statement
WITH statement
Data set:

PROCLIB.ONE on page 226.

This example shows that PROC COMPARE can compare two variables that are in
the same data set.

The COMPARE Procedure

Program

265

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Create a short summary report of the differences within one data set. ALLSTATS prints
summary statistics. BRIEFSUMMARY prints only a short comparison summary.
proc compare base=proclib.one allstats briefsummary;

Specify two variables from the base data set to compare. The VAR and WITH statements
specify the variables in the base data set to compare. This example compares GR1 with GR2.
Because there is no comparison data set, the variables GR1 and GR2 must be in the base data
set.
var gr1;
with gr2;
title Comparison of Variables in the Same Data Set;
run;

266

Output

Chapter 9

Output
Comparison of Variables in the Same Data Set
COMPARE Procedure
Comparisons of variables in PROCLIB.ONE
(Method=EXACT)
NOTE: Values of the following 1 variables compare unequal: gr1^=gr2

Value Comparison Results for Variables

__________________________________________________________
||
Base
Compare
Obs ||
gr1
gr2
Diff.
% Diff
________ || _________ _________ _________ _________
||
1 ||
85.0
87.0000
2.0000
2.3529
3 ||
78.0
72.0000
-6.0000
-7.6923
4 ||
87.0
94.0000
7.0000
8.0460
________ || _________ _________ _________ _________
||
N
||
4
4
4
4
Mean
||
85.5000
86.2500
0.7500
0.6767
Std
||
5.8023
9.9457
5.3774
6.5221
Max
||
92.0000
94.0000
7.0000
8.0460
Min
||
78.0000
72.0000
-6.0000
-7.6923
StdErr
||
2.9011
4.9728
2.6887
3.2611
t
||
29.4711
17.3442
0.2789
0.2075
Prob>|t| ||
<.0001
0.0004
0.7984
0.8489
||
Ndif
||
3
75.000%
DifMeans ||
0.877%
0.870%
0.7500
r, rsq
||
0.898
0.807
__________________________________________________________

Example 5: Comparing Observations with an ID Variable

Procedure features:

ID statement

In this example, PROC COMPARE compares only the observations that have
matching values for the ID variable.

The COMPARE Procedure

Program

267

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Create the PROCLIB.EMP95 and PROCLIB.EMP96 data sets. PROCLIB.EMP95 and

PROCLIB.EMP96 contain employee data. IDNUM works well as an ID variable because it has
unique values. A DATA step on page 1391 creates PROCLIB.EMP95. A DATA step on page 1392
creates PROCLIB.EMP96.
data proclib.emp95;
input #1 idnum $4. @6 name $15.
#2 address $42.
#3 salary 6.;
datalines;
2388 James Schmidt
100 Apt. C Blount St. SW Raleigh NC 27693
92100
2457 Fred Williams
99 West Lane Garner NC 27509
33190
... more data lines...
3888 Kim Siu
5662 Magnolia Blvd Southeast Cary NC 27513
77558
;
data proclib.emp96;
input #1 idnum $4. @6 name $15.
#2 address $42.
#3 salary 6.;
datalines;
2388 James Schmidt
100 Apt. C Blount St. SW Raleigh NC 27693
92100
2457 Fred Williams
99 West Lane Garner NC 27509
33190
...more data lines...
6544 Roger Monday
3004 Crepe Myrtle Court Raleigh NC 27604
47007
;

268

Program

Chapter 9

Sort the data sets by the ID variable. Both data sets must be sorted by the variable that will
be used as the ID variable in the PROC COMPARE step. OUT= species the location of the
sorted data.
proc sort data=proclib.emp95 out=emp95_byidnum;
by idnum;
run;
proc sort data=proclib.emp96 out=emp96_byidnum;
by idnum;
run;

Create a summary report that compares observations with matching values for the ID
variable. The ID statement species IDNUM as the ID variable.
proc compare base=emp95_byidnum compare=emp96_byidnum;
id idnum;
title Comparing Observations that Have Matching IDNUMs;
run;

The COMPARE Procedure

Output

269

Output

PROC COMPARE identies specic observations by the value of IDNUM. In the

Value Comparison Results for Variables section, PROC COMPARE prints the
nonmatching addresses and nonmatching salaries. For salaries, PROC COMPARE computes the
numerical difference and the percent difference. Because ADDRESS is a character variable,
PROC COMPARE displays only the rst 20 characters. For addresses where the observation
has an IDNUM of 0987, 2776, or 3888, the differences occur after the 20th character and the
differences do not appear in the output. The plus sign in the output indicates that the full value
is not shown. To see the entire value, create an output data set. See Example 6 on page 270.

Comparing Observations that Have Matching IDNUMs

COMPARE Procedure
Comparison of WORK.EMP95_BYIDNUM with WORK.EMP96_BYIDNUM
(Method=EXACT)
Data Set Summary
Dataset

Created

WORK.EMP95_BYIDNUM
WORK.EMP96_BYIDNUM

Modified

NVar

NObs

13MAY98:16:03:36
13MAY98:16:03:36

4
4

10
12

Variables Summary
Number of Variables in Common: 4.
Number of ID Variables: 1.

Observation Summary
Observation
First
First
Last
Last

Obs
Unequal
Unequal
Obs

Base

Compare

1
1
10
10

1
1
12
12

ID
idnum=0987
idnum=0987
idnum=9857
idnum=9857

Number of Observations in Common: 10.

Number of Observations in WORK.EMP96_BYIDNUM but not in WORK.EMP95_BYIDNUM: 2.
Total Number of Observations Read from WORK.EMP95_BYIDNUM: 10.
Total Number of Observations Read from WORK.EMP96_BYIDNUM: 12.
Number of Observations with Some Compared Variables Unequal: 5.
Number of Observations with All Compared Variables Equal: 5.
Comparing Observations that Have Matching IDNUMs
COMPARE Procedure
Comparison of WORK.EMP95_BYIDNUM with WORK.EMP96_BYIDNUM
(Method=EXACT)
Values Comparison Summary
Number of Variables Compared with All Observations Equal: 1.
Number of Variables Compared with Some Observations Unequal: 2.
Total Number of Values which Compare Unequal: 8.
Maximum Difference: 2400.

270

Example 6: Comparing Values of Observations Using an Output Data Set (OUT=)

Chapter 9

Variables with Unequal Values

Variable

Type

Len

Ndif

MaxDif

address
salary

CHAR
NUM

42
8

4
4

2400

Value Comparison Results for Variables

_______________________________________________________
|| Base Value
Compare Value
idnum || address
address
_____ || ___________________+ ___________________+
||
0987
|| 2344 Persimmons Bran 2344 Persimmons Bran
2776
|| 12988 Wellington Far 12988 Wellington Far
3888
|| 5662 Magnolia Blvd S 5662 Magnolia Blvd S
9857
|| 1000 Taft Ave. Morri 100 Taft Ave. Morris
_______________________________________________________
Comparing Observations that Have Matching IDNUMs

COMPARE Procedure
Comparison of WORK.EMP95_BYIDNUM with WORK.EMP96_BYIDNUM
(Method=EXACT)
Value Comparison Results for Variables
_______________________________________________________
||
Base
Compare
idnum ||
salary
salary
Diff.
% Diff
_____ || _________ _________ _________ _________
||
0987
||
44010
45110
1100
2.4994
3286
||
87734
89834
2100
2.3936
3888
||
77558
79958
2400
3.0945
9857
||
38756
40456
1700
4.3864
_______________________________________________________

Example 6: Comparing Values of Observations Using an Output Data Set

(OUT=)
Procedure features:

PROC COMPARE statement options:

NOPRINT
OUT=
OUTBASE
OUTBASE
OUTCOMP
OUTDIF
OUTNOEQUAL
Other features: PRINT procedure
Data sets: PROCLIB.EMP95 and PROCLIB.EMP96 on page 267

This example creates and prints an output data set that shows the differences
between matching observations.

The COMPARE Procedure

Program

271

In Example 5 on page 266, the output does not show the differences past the 20th
character. The output data set in this example shows the full values. Further, it shows
the observations that occur in only one of the data sets.

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Specify the data sets to compare. BASE= and COMPARE= specify the data sets to compare.
proc compare base=emp95_byidnum compare=emp96_byidnum

Create the output data set RESULT and include all unequal observations and their
differences. OUT= names and creates the output data set. NOPRINT suppresses the printing
of the procedure output. OUTNOEQUAL includes only observations that are judged unequal.
OUTBASE writes an observation to the output data set for each observation in the base data
set. OUTCOMP writes an observation to the output data set for each observation in the
comparison data set. OUTDIF writes an observation to the output data set that contains the
differences between the two observations.
out=result outnoequal outbase outcomp outdif
noprint;

272

Output

Chapter 9

Specify the ID variable. The ID statement species IDNUM as the ID variable.

id idnum;
run;

Print the output data set RESULT and use the BY and ID statements with the ID
variable. PROC PRINT prints the output data set. Using the BY and ID statements with the
same variable makes the output easy to read. See Chapter 34, The PRINT Procedure, on page
703 for more information on this technique.
proc print data=result noobs;
by idnum;
id idnum;
title The Output Data Set RESULT;
run;

Output

The differences for character variables are noted with an X or a period (.). An X shows that the characters do
not match. A period shows that the characters do match. For numeric variables, an E means that there is no
difference. Otherwise, the numeric difference is shown. By default, the output data set shows that two
observations in the comparison data set have no matching observation in the base data set. You do not have to
use an option to make those observations appear in the output data set.

The Output Data Set RESULT

idnum

_TYPE_

0987

BASE
COMPARE
DIF

name

address

Dolly Lunford

2344 Persimmons Branch

1
1

Dolly Lunford
...............

2344 Persimmons Branch Trail Apex NC 27505

.......................XXXXX.XXXXXXXXXXXXX

45110
1100

BASE
COMPARE

5
5

Robert Jones
Robert Jones

12988 Wellington Farms Ave. Cary NC 27512

12988 Wellington Farms Ave. Cary NC 27511

29025
29025

DIF

...............

........................................X.

3278

COMPARE

Mary Cravens

211 N. Cypress St. Cary NC 27512

35362

3286

BASE
COMPARE

6
7

Hoa Nguyen
Hoa Nguyen

2818 Long St. Cary NC 27513

87734
89834

DIF

...............

..........................................

2100

BASE

Kim Siu

5662 Magnolia Blvd Southeast Cary NC 27513

77558

COMPARE
DIF

8
7

Kim Siu
...............

5662 Magnolia Blvd Southwest Cary NC 27513

........................XX................

79958
2400

6544

COMPARE

Roger Monday

3004 Crepe Myrtle Court Raleigh NC 27604

47007

9857

BASE
COMPARE
DIF

10
12
10

Kathy Krupski
Kathy Krupski
...............

1000 Taft Ave. Morrisville NC 27508

100 Taft Ave. Morrisville NC 27508
...XXXXXXXXXXXXXX.XXXXX.XXXXXXXXXXX.......

38756
40456
1700

2776

3888

_OBS_

1
salary
Apex NC 27505

44010

The COMPARE Procedure

Program

273

Example 7: Creating an Output Data Set of Statistics (OUTSTATS=)

Procedure features:

PROC COMPARE statement options:

NOPRINT
OUTSTATS=
Data sets: PROCLIB.EMP95, PROCLIB.EMP96 on page 267

This example creates an output data set that contains summary statistics for the
numeric variables that are compared.

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Create the output data set of statistics and compare observations that have matching
values for the ID variable. BASE= and COMPARE= specify the data sets to compare.
OUTSTATS= creates the output data set DIFFSTAT. NOPRINT suppresses the procedure
output. The ID statement species IDNUM as the ID variable. PROC COMPARE uses the
values of IDNUM to match observations.
proc compare base=emp95_byidnum compare=emp96_byidnum
outstats=diffstat noprint;
id idnum;
run;

274

Output

Chapter 9

Print the output data set DIFFSTAT. PROC PRINT prints the output data set DIFFSTAT.
proc print data=diffstat noobs;
title The DIFFSTAT Data Set;
run;

Output

The variables are described in Output Statistics Data Set (OUTSTATS=) on page 255.

The DIFFSTAT Data Set

_VAR_

_TYPE_

salary
salary
salary
salary
salary
salary
salary
salary
salary
salary
salary

N
MEAN
STD
MAX
MIN
STDERR
T
PROBT
NDIF
DIFMEANS
R,RSQ

_BASE_

_COMP_

_DIF_

_PCTDIF_

10.00
52359.00
24143.84
92100.00
29025.00
7634.95
6.86
0.00
4.00
1.39
1.00

10.00
53089.00
24631.01
92100.00
29025.00
7789.01
6.82
0.00
40.00
1.38
1.00

10.00
730.00
996.72
2400.00
0.00
315.19
2.32
0.05
.
730.00
.

10.0000
1.2374
1.6826
4.3864
0.0000
0.5321
2.3255
0.0451
.
.
.

275

CHAPTER

10
The CONTENTS Procedure
Overview: CONTENTS Procedure 275
Syntax: CONTENTS Procedure 275

Overview: CONTENTS Procedure

The CONTENTS procedure shows the contents of a SAS data set and prints the
directory of the SAS data library.
Generally, the CONTENTS procedure functions the same as the CONTENTS
statement in the DATASETS procedure. The differences between the CONTENTS
procedure and the CONTENTS statement in PROC DATASETS are as follows:
3 The default for libref in the DATA= option in PROC CONTENTS is either WORK
or USER. For the CONTENTS statement, the default is the libref of the procedure
input library.
3 PROC CONTENTS can read sequential les. The CONTENTS statement cannot.

Syntax: CONTENTS Procedure

Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Names: See: ODS Table Names on page 369
Reminder: You can use the ATTRIB, FORMAT, and LABEL statements. See Chapter 3,
Statements with the Same Function in Multiple Procedures, on page 57 for details.
You can also use any global statements. See Global Statements on page 18 for a list.
Reminder: You can use data set options with the DATA= and OUT= options. See Data
Set Options on page 18 for a list.
Reminder: Complete documentation for the CONTENTS statement and the CONTENTS
procedure is in CONTENTS Statement on page 323.
Tip:

See:

CONTENTS Procedure in the documentation for your operating environment.

PROC CONTENTS <option(s)>;

276

Syntax: CONTENTS Procedure

Chapter 10

To do this

Use this option

Print centiles information for indexed variables

CENTILES

Specify the input data set

DATA=

Include information in the output about the

number of observations, number of variables, and
data set labels

DETAILS|NODETAILS

Print a list of the SAS les in the SAS data library

Print the length of a variables informat or format

FMTLEN

Restrict processing to one or more types of SAS le

MEMTYPE=

Suppress the printing of individual les

NODS

Suppress the printing of the output

NOPRINT

Print a list of variables in alphabetical order even

if they include mixed case names

ORDER=IGNORECASE

Specify the output data set

OUT=

Specify an output data set that contains

information about constraints

OUT2=

Print abbreviated output

SHORT

Print a list of the variables by their logical position

in the data set

VARNUM

277

CHAPTER

11
The COPY Procedure
Overview: COPY Procedure 277
Syntax: COPY Procedure 277
Concepts: COPY Procedure 278
Transporting SAS Data Sets between Hosts 278
Example: COPY Procedure 279
Example 1: Copying SAS Data Sets between Hosts

279

Overview: COPY Procedure

The COPY procedure copies one or more SAS les from a SAS data library.
Generally, the COPY procedure functions the same as the COPY statement in the
DATASETS procedure. The two differences are as follows:

3 The IN= argument is required with PROC COPY. In the COPY statement, IN= is
optional. If IN= is omitted, the default value is the libref of the procedure input
library.
3 PROC DATASETS cannot work with libraries that allow only sequential data
access.
Note: The MIGRATE procedure is available specically for migrating a SAS data
library from a previous release to the most recent release. For migration, PROC
MIGRATE offers benets that PROC COPY does not. For documentation on PROC
MIGRATE, see the Migration Community at http://support.sas.com/rnd/
migration. 4

Syntax: COPY Procedure

See Chapter 3, Statements with the Same Function in Multiple Procedures,
on page 57 for details. You can also use any global statements. See Global Statements
on page 18 for a list.
Reminder: Complete documentation for the COPY statement and the COPY procedure is
in COPY Statement on page 327.
Restriction: PROC COPY ignores explicit concatenations with catalogs. Use PROC
CATALOG COPY to copy concatenated catalogs.
Reminder:

278

Concepts: COPY Procedure

Chapter 11

PROC COPY OUT=libref-1 IN=libref-2

<CLONE|NOCLONE>
<CONSTRAINT=YES|NO>
<DATECOPY>
<INDEX=YES|NO>
<MEMTYPE=(mtype(s))>
<MOVE <ALTER=alter-password>>;
EXCLUDE SAS-le(s) </ MEMTYPE=mtype>;
SELECT SAS-le(s) </ <MEMTYPE=mtype>
<ALTER=alter-password>>;

Concepts: COPY Procedure

Transporting SAS Data Sets between Hosts

The COPY procedure, along with the XPORT engine and the XML engine, can create
and read transport les that can be moved from one host to another. PROC COPY can
create transport les only with SAS data sets, not with catalogs or other types of SAS
les.
Transporting is a three-step process:
1 Use PROC COPY to copy one or more SAS data sets to a le that is created with

either the transport (XPORT) engine or the XML engine. This le is referred to as
a transport le and is always a sequential le.
2 After the le is created, you can move it to another operating environment via

communications software, such as FTP, or tape. If you use communications

software, be sure to move the le in binary format to avoid any type of conversion.
If you are moving the le to a mainframe, the le must have certain attributes.
Consult the SAS documentation for your operating environment and the SAS
Technical Support Web page for more information.
3 After you have successfully moved the le to the receiving host, use PROC COPY

to copy the data sets from the transport le to a SAS data library.
For an example, see Example 1 on page 279.
For details on transporting les, see Moving and Accessing SAS Files across
Operating Environments.
The CPORT and CIMPORT procedures also provide a way to transport SAS les. For
information, see Chapter 8, The CIMPORT Procedure, on page 215 and Chapter 13,
The CPORT Procedure, on page 285.

The COPY Procedure

Program

279

Example: COPY Procedure

Example 1: Copying SAS Data Sets between Hosts

Features:

PROC COPY statement options:

IN=
MEMTYPE=
OUT=
Other features:

XPORT engine

This example illustrates how to create a transport le on a host and read it on

another host.
In order for this example to work correctly, the transport le must have certain
characteristics, as described in the SAS documentation for your operating environment.
In addition, the transport le must be moved to the receiving operating system in
binary format.

Program

Assign library references. Assign a libref, such as SOURCE, to the SAS data library that
contains the SAS data set that you want to transport. Also, assign a libref to the transport le
and use the XPORT keyword to specify the XPORT engine.
libname source SAS-data-library-on-sending-host;
libname xptout xport filename-on-sending-host;

Copy the SAS data sets to the transport le. Use PROC COPY to copy the SAS data sets
from the IN= library to the transport le. MEMTYPE=DATA species that only SAS data sets
are copied. SELECT selects the data sets that you want to copy.
proc copy in=source out=xptout memtype=data;
select bonus budget salary;
run;

280

SAS Log

Chapter 11

SAS Log

SAS Log on Sending Host

1
libname source SAS-data-library-on-sending-host ;
NOTE: Libref SOURCE was successfully assigned as follows:
Engine:
V9
Physical Name: SAS-data-library-on-sending-host
2
libname xptout xport filename-on-sending-host;
NOTE: Libref XPTOUT was successfully assigned as follows:
Engine:
XPORT
Physical Name: filename-on-sending-host
3
proc copy in=source out=xptout memtype=data;
4
select bonus budget salary;
5
run;
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:

Copying SOURCE.BONUS to XPTOUT.BONUS (memtype=DATA).

The data set XPTOUT.BONUS has 1 observations and 3 variables.
Copying SOURCE.BUDGET to XPTOUT.BUDGET (memtype=DATA).
The data set XPTOUT.BUDGET has 1 observations and 3 variables.
Copying SOURCE.SALARY to XPTOUT.SALARY (memtype=DATA).
The data set XPTOUT.SALARY has 1 observations and 3 variables.

Enable the procedure to read data from the transport le. The XPORT engine in the
LIBNAME statement enables the procedure to read the data from the transport le.
libname insource xport filename-on-receiving-host;

Copy the SAS data sets to the receiving host. After you copy the les (for example, by using
FTP in binary mode to the Windows NT host), use PROC COPY to copy the SAS data sets to the
WORK data library on the receiving host.
proc copy in=insource out=work;
run;

The COPY Procedure

SAS Log on Receiving Host

1
libname insource xport filename-on-receiving-host;
NOTE: Libref INSOURCE was successfully assigned as follows:
Engine:
XPORT
Physical Name: filename-on-receiving-host
2
proc copy in=insource out=work;
3
run;
NOTE: Input library INSOURCE is sequential.
NOTE: Copying INSOURCE.BUDGET to WORK.BUDGET (memtype=DATA).
NOTE: BUFSIZE is not cloned when copying across different engines.
System Option for BUFSIZE was used.
NOTE: The data set WORK.BUDGET has 1 observations and 3 variables.
NOTE: Copying INSOURCE.BONUS to WORK.BONUS (memtype=DATA).
NOTE: BUFSIZE is not cloned when copying across different engines.
System Option for BUFSIZE was used.
NOTE: The data set WORK.BONUS has 1 observations and 3 variables.
NOTE: Copying INSOURCE.SALARY to WORK.SALARY (memtype=DATA).
NOTE: BUFSIZE is not cloned when copying across different engines.
System Option for BUFSIZE was used.
NOTE: The data set WORK.SALARY has 1 observations and 3 variables.

SAS Log

281

282

283

CHAPTER

12
The CORR Procedure
Information about the CORR Procedure

283

Information about the CORR Procedure

See:

The documentation for the CORR procedure has moved to Volume 3 of this book.

284

285

CHAPTER

13
The CPORT Procedure
Overview: CPORT Procedure 285
What Does the CPORT Procedure Do? 285
General File Transport Process 286
Syntax: CPORT Procedure 286
PROC CPORT Statement 286
EXCLUDE Statement 292
SELECT Statement 293
TRANTAB Statement 294
Concepts: CPORT Procedure 294
Results: CPORT Procedure 294
Examples: CPORT Procedure 295
Example 1: Exporting Multiple Catalogs 295
Example 2: Exporting Individual Catalog Entries 296
Example 3: Exporting a Single SAS Data Set 296
Example 4: Applying a Translation Table 297
Example 5: Exporting Entries Based on Modication Date

298

Overview: CPORT Procedure

What Does the CPORT Procedure Do?
The CPORT procedure writes SAS data sets, SAS catalogs, or SAS data libraries to
sequential le formats (transport les). Use PROC CPORT with the CIMPORT
procedure to move les from one environment to another. Transport les are sequential
les that each contain a SAS data library, a SAS catalog, or a SAS data set in transport
format. The transport format that PROC CPORT writes is the same for all
environments and for many releases of SAS. In PROC CPORT, export means to put a
SAS data library, a SAS catalog, or a SAS data set into transport format. PROC
CPORT exports catalogs and data sets, either singly or as a SAS data library. PROC
CIMPORT restores (imports) the transport le to its original form as a SAS catalog,
SAS data set, or SAS data library.
Only PROC CIMPORT can read the transport les that PROC CPORT creates. For
information on the transport les that the transport engine creates, see the section on
SAS les in SAS Language Reference: Concepts.
PROC CPORT also converts SAS les, which means that it changes the format of a
SAS le from the format appropriate for one version of SAS to the format appropriate
for another version. For example, you can use PROC CPORT and PROC CIMPORT to
move les from earlier releases of SAS to more recent releases. In such cases, PROC
CIMPORT automatically converts the contents of the transport le as it imports it.

286

General File Transport Process

Chapter 13

PROC CPORT produces no output (other than the transport les), but it does write
notes to the SAS log.

General File Transport Process

To export and import les, follow these steps:
1 Use PROC CPORT to export the SAS les that you want to transport.
2 If you are changing operating environments, move the transport le to the new

machine by using either communications software or a magnetic medium.

Note: If you use communications software to move the transport le, be sure that
it treats the transport le as a binary le and that it modies neither the
attributes nor the contents of the le.

3 Use PROC CIMPORT to translate the transport le into the format appropriate

for the new operating environment or release.

Syntax: CPORT Procedure

See:

CPORT Procedure in the documentation for your operating environment.

PROC CPORT source-type=libref | <libref.>member-name< option(s)>;

EXCLUDE SAS le(s) | catalog entry(s)</ MEMTYPE=mtype></
ENTRYTYPE=entry-type>;
SELECT SAS le(s) | catalog entry(s) </ MEMTYPE=mtype></
ENTRYTYPE=entry-type>;
TRANTAB NAME=translation-table-name
<option(s)>;

PROC CPORT Statement

PROC CPORT source-type=libref | <libref.>member-name< option(s)>;

To do this

Use this option

Identify the transport le

Specify the transport le to write to

FILE=

The CPORT Procedure

To do this

PROC CPORT Statement

Use this option

Direct the output from PROC CPORT to a
tape

TAPE

Select les to export

Export copies of all data sets or catalog
entries that have a modication date equal to
or later than the date you specify

AFTER=

Exclude specied entry types from the

transport le

EET=

Include specied entry types in the transport

ET=

Specify whether to export all generations of a

data set

GENERATION=

Specify that only data sets, only catalogs, or

both, be moved when a library is exported

MEMTYPE=

Control the contents of the transport le

Suppress the conversion of displayed
character data to transport format

ASIS

Control the exportation of integrity

constraints

CONSTRAINT

Copy the created and modied date and time

to the transport le

DATECOPY

Control the exportation of indexes with

indexed SAS data sets

INDEX

Suppress the compression of binary zeros and

blanks in the transport le

NOCOMPRESS

Write all alphabetic characters to the

transport le in uppercase

OUTTYPE=
UPCASE

Translate specied characters from one

ASCII or EBCDIC value to another

TRANSLATE

Export SAS/AF PROGRAM and SCL entries without

edit capability when you import them

NOEDIT

Specify that exported catalog entries contain compiled

SCL code, but not the source code

NOSRC

Specify a libref associated with a SAS data library

OUTLIB=

Required Arguments

source-type=libref | < libref.>member-name

identies the type of le to export and species the catalog, SAS data set, or SAS
data library to export.

287

288

PROC CPORT Statement

Chapter 13

source-type
identies the le(s) to export as a single catalog, as a single SAS data set, or as
the members of a SAS data library. The source-type argument can be one of the
following:
CATALOG | CAT | C
DATA | DS | D
LIBRARY | LIB | L
libref | <libref.>member-name
species the specic catalog, SAS data set, or SAS data library to export. If
source-type is CATALOG or DATA, you can specify both a libref and a member
name. If the libref is omitted, PROC CPORT uses the default library as the libref,
which is usually the WORK library. If the source-type argument is LIBRARY,
specify only a libref. If you specify a library, PROC CPORT exports only data sets
and catalogs from that library. You cannot export other types of les.

Options
AFTER=date

exports copies of all data sets or catalog entries that have a modication date later
than or equal to the date you specify. The modication date is the most recent date
when the contents of the data set or catalog entry changed. Specify date as a SAS
date literal or as a numeric SAS date value.
Tip: You can determine the modication date of a catalog entry by using the
CATALOG procedure.
Featured in: Example 5 on page 298.
ASIS

suppresses the conversion of displayed character data to transport format. Use this
option when you move les that contain DBCS (double-byte character set) data from
one operating environment to another if both operating environments use the same
type of DBCS data.
Interaction: The ASIS option invokes the NOCOMPRESS option.
Interaction: You cannot use both the ASIS option and the OUTTYPE= options in
the same PROC CPORT step.
CONSTRAINT=YES | NO

controls the exportation of integrity constraints that have been dened on a data set.
When you specify CONSTRAINT=YES, all types of integrity constraints are exported
for a library; only general integrity constraints are exported for a single data set.
When you specify CONTRAINT=NO, indexes created without integrity constraints
are ported, but neither integrity constraints nor any indexes created with integrity
constraints are ported. For more information on integrity constraints, see the section
on SAS les in SAS Language Reference: Concepts.
Alias: CON=
Default: YES
Interaction: You cannot specify both CONSTRAINT= and INDEX= in the same
PROC CPORT step.
Interaction: If you specify INDEX=NO, no integrity constraints are exported.
DATECOPY

copies the SAS internal date and time when the SAS le was created and the date
and time when it was last modied to the resulting transport le. Note that the
operating environment date and time are not preserved.

The CPORT Procedure

PROC CPORT Statement

289

Restriction: DATECOPY can be used only when the destination le uses the V8 or

V9 engine.
You can alter the le creation date and time with the DTC= option on the
MODIFY statementMODIFY Statement on page 348 in a PROC DATASETS step.

Chapter 13

INTYPE=DBCS-type

species the type of DBCS data stored in the SAS les to be exported. Double-byte
character set (DBCS) data uses up to two bytes for each character in the set.
DBCS-type must be one of the following values:
IBM | HITAC |
FACOM

for z/OS

IBM

for VSE

DEC | SJIS

for OpenVMS

PCIBM | SJIS

for OS/2

Restriction The INTYPE= option is allowed only if SAS is built with Double-Byte

Character Set (DBCS) extensions. Because these extensions require signicant

computing resources, there is a special distribution for those sites that require it.
An error is reported if this option is used at a site for which DBCS extensions are
not enabled.
Default: If the INTYPE= option is not used, the DBCS type defaults to the value of

the SAS system option DBCSTYPE=.

Interaction: Use the INTYPE= option in conjunction with the OUTTYPE= option to

change from one type of DBCS data to another.

Interaction: The INTYPE= option invokes the NOCOMRPESS option.
Interaction: You cannot use the INTYPE= option and the ASIS option in the same

PROC CPORT step.

You can set the value of the SAS system option DBCSTYPE= in your
conguration le.

Tip:

MEMTYPE=mtype

restricts the type of SAS le that PROC CPORT writes to the transport le.
MEMTYPE= restricts processing to one member type. Values for mtype can be
ALL
both catalogs and data sets
CATALOG | CAT
catalogs
DATA | DS
SAS data sets
Alias: MT=
Default: ALL
Featured in:

Example 1 on page 295.

NOCOMPRESS

suppresses the compression of binary zeros and blanks in the transport le.
Alias:

NOCOMP

Default: By default, PROC CPORT compresses binary zeros and blanks to conserve

space.
Interaction: The ASIS, INTYPE=, and OUTTYPE= options invoke the

NOCOMPRESS option.
Note: Compression of the transport le does not alter the ag in each catalog and
data set that indicates whether the original le was compressed. 4
NOEDIT

The CPORT Procedure

PROC CPORT Statement

291

exports SAS/AF PROGRAM and SCL entries without edit capability when you
import them.
The NOEDIT option produces the same results as when you create a new catalog
to contain SCL code by using the MERGE statement with the NOEDIT option in the
BUILD procedure of SAS/AF software.
Note: The NOEDIT option affects only SAS/AF PROGRAM and SCL entries. It
does not affect FSEDIT SCREEN or FSVIEW FORMULA entries. 4
Alias:

NEDIT

NOSRC

species that exported catalog entries contain compiled SCL code but not the source
code.
The NOSRC option produces the same results as when you create a new catalog to
contain SCL code by using the MERGE statement with the NOSOURCE option in
the BUILD procedure of SAS/AF software.
Alias:

NSRC

OUTLIB=libref

species a libref associated with a SAS data library. If you specify the OUTLIB=
option, PROC CIMPORT is invoked automatically to re-create the input data library,
data set, or catalog in the specied library.
Alias: OUT=
Tip: Use the OUTLIB= option when you change SAS les from one DBCS type to
another within the same operating environment if you want to keep the original
data intact.
OUTTYPE=UPCASE

writes all displayed characters to the transport le and to the OUTLIB= le in

uppercase.
Interaction: The OUTTYPE= option invokes the NOCOMPRESS option.
TAPE

directs the output from PROC CPORT to a tape.

Default: The output from PROC CPORT is sent to disk.
TRANSLATE=(translation-list)

translates specied characters from one ASCII or EBCDIC value to another. Each
element of translation-list has the form
ASCII-value-1 TO ASCII-value-2
EBCDIC-value-1 TO EBCDIC-value-2
You can use hexadecimal or decimal representation for ASCII values. If you use
the hexadecimal representation, values must begin with a digit and end with an x.
Use a leading zero if the hexadecimal value begins with an alphabetic character.
For example, to translate all left brackets to left braces, specify the TRANSLATE=
option as follows (for ASCII characters):
translate=(5bx to 7bx)

The following example translates all left brackets to left braces and all right
brackets to right braces:
translate=(5bx to 7bx 5dx to 7dx)

292

EXCLUDE Statement

Chapter 13

EXCLUDE Statement
Excludes specied les or entries from the transport le.
Tip: There is no limit to the number of EXCLUDE statements you can use in one
invocation of PROC CPORT.
Interaction: You can use either EXCLUDE statements or SELECT statements in a PROC
CPORT step, but not both.

EXCLUDE SAS le(s) | catalog entry(s)< / MEMTYPE=mtype></

ENTRYTYPE=entry-type>;

Required Arguments
SAS le(s) | catalog entry(s)

species either the name(s) of one or more SAS les or the names of one or more
catalog entries to be excluded from the transport le. Specify SAS lenames when
you export a SAS data library; specify catalog entry names when you export an
individual SAS catalog. Separate multiple lenames or entry names with a space.
You can use shortcuts to list many like-named les in the EXCLUDE statement. For
more information, see Shortcuts for Specifying Lists of Variable Names on page 24.

Options
ENTRYTYPE=entry-type

species a single entry type for the catalog entries listed in the EXCLUDE statement.
See SAS Language Reference: Concepts for a complete list of catalog entry types.
Restriction: ENTRYTYPE= is valid only when you export an individual SAS
catalog.
Alias: ETYPE=, ET=
MEMTYPE=mtype

species a single member type for the SAS le(s) listed in the EXCLUDE statement.
Valid values are CATALOG or CAT, DATA, or ALL. If you do not specify the
MEMTYPE= option in the EXCLUDE statement, then processing is restricted to
those member types specied in the MEMTYPE= option in the PROC CPORT
statement.
You can also specify the MEMTYPE= option, enclosed in parentheses, immediately
after the name of a le. In parentheses, MEMTYPE= identies the type of the le
name that just precedes it. When you use this form of the option, it overrides the
MEMTYPE= option that follows the slash in the EXCLUDE statement, but it must
match the MEMTYPE= option in the PROC CPORT statement:
Restriction: MEMTYPE= is valid only when you export a SAS data library.
Restriction: If you specify a member type for MEMTYPE= in the PROC CPORT
statement, it must agree with the member type that you specify for MEMTYPE=
in the EXCLUDE statement.
Alias: MTYPE=, MT=
Default: If you do not specify MEMTYPE= in the PROC CPORT statement or in
the EXCLUDE statement, the default is MEMTYPE=ALL.

The CPORT Procedure

SELECT Statement

293

SELECT Statement
Includes specied les or entries in the transport le.
There is no limit to the number of SELECT statements you can use in one
invocation of PROC CPORT.

Tip:

Interaction: You can use either EXCLUDE statements or SELECT statements in a PROC
CPORT step, but not both.
Featured in: Example 2 on page 296

SELECT SAS le(s) | catalog entry(s)< / MEMTYPE=mtype> </

ENTRYTYPE=entry-type> ;

Required Arguments
SAS le(s) | catalog entry(s)

species either the name(s) of one or more SAS les or the names of one or more
catalog entries to be included in the transport le. Specify SAS lenames when you
export a SAS data library; specify catalog entry names when you export an
individual SAS catalog. Separate multiple lenames or entry names with a space.
You can use shortcuts to list many like-named les in the SELECT statement. For
more information, see Shortcuts for Specifying Lists of Variable Names on page 24.

Options
ENTRYTYPE=entry-type

species a single entry type for the catalog entries listed in the SELECT statement.
See SAS Language Reference: Concepts for a complete list of catalog entry types.
Restriction: ENTRYTYPE= is valid only when you export an individual SAS

catalog.
Alias:

ETYPE=, ET=

MEMTYPE=mtype

species a single member type for the SAS le(s) listed in the SELECT statement.
Valid values are CATALOG or CAT, DATA, or ALL. If you do not specify the
MEMTYPE= option in the SELECT statement, then processing is restricted to those
member types specied in the MEMTYPE= option in the PROC CPORT statement.
You can also specify the MEMTYPE= option, enclosed in parentheses, immediately
after the name of a member. In parentheses, MEMTYPE= identies the type of the
member name that just precedes it. When you use this form of the option, it
overrides the MEMTYPE= option that follows the slash in the SELECT statement,
but it must match the MEMTYPE= option in the PROC CPORT statement.
Restriction: MEMTYPE= is valid only when you export a SAS data library.
Restriction: If you specify a member type for MEMTYPE= in the PROC CPORT

statement, it must agree with the member type that you specify for MEMTYPE=
in the SELECT statement.
Alias:

MTYPE=, MT=

294

TRANTAB Statement

Chapter 13

Default: If you do not specify MEMTYPE= in the PROC CPORT statement or in

the SELECT statement, the default is MEMTYPE=ALL.

TRANTAB Statement
Species translation tables for characters in catalog entries you export.
Tip: You can specify only one table for each TRANTAB statement, but there is no limit
to the number of TRANTAB statements you can use in one invocation of PROC CPORT.
Featured in:

Example 4 on page 297.

The TRANTAB Statement for the CPORT Procedure and the UPLOAD and
DOWNLOAD Procedures in SAS National Language Support (NLS): Users Guide

See:

TRANTAB NAME=translation-table-name
<option(s)>;

Concepts: CPORT Procedure

For password-protected data sets, the password(s) are applied to the destination data
set when it is imported. If the data set is transported as part of a library, it is not
necessary to supply the password. If the data set is transported singly, you must supply
the read password. If you omit the password in the PROC CPORT step, SAS prompts
you for the password. If the target SAS engine does not support passwords, then the
import will fail. For example, the following SAS code transports a password-protected
data set called WORK.ONE:
proc cport data=one(read=hithere) file=bin;

Results: CPORT Procedure

A common problem when you create or import a transport le under the z/OS
environment is a failure to specify the correct Data Control Block (DCB) characteristics.
When you reference a transport le, you must specify the following DCB characteristics:
Another common problem can occur if you use communications software to move les
from another environment to z/OS. In some cases, the transport le does not have the
proper DCB characteristics when it arrives on z/OS. If the communications software
does not allow you to specify le characteristics, try the following approach for z/OS:
1 Create a le under z/OS with the correct DCB characteristics and initialize the le.
2 Move the transport le from the other environment to the newly created le under

z/OS using binary transfer.

The CPORT Procedure

SAS Log

295

Examples: CPORT Procedure

Example 1: Exporting Multiple Catalogs

Procedure features:

PROC CPORT statement options:

FILE=
MEMTYPE=

This example shows how to use PROC CPORT to export entries from all of the SAS
catalogs in the SAS data library you specify.

Program

Specify the library reference for the SAS data library that contains the source les to
be exported and the le reference to which the output transport le is written. The
LIBNAME statement assigns a libref for the SAS data library. The FILENAME statement
assigns a leref and any operating environment options for le characteristics for the transport
le that PROC CPORT creates.
libname source SAS-data-library;
filename tranfile transport-file
host-option(s)-for-file-characteristics;

Create the transport le. The PROC CPORT step executes on the operating environment
where the source library is located. MEMTYPE=CATALOG writes all SAS catalogs in the source
library to the transport le.
proc cport library=source file=tranfile memtype=catalog;
run;

SAS Log
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:

Proc CPORT begins to transport catalog SOURCE.FINANCE

The catalog has 5 entries and its maximum logical record length is 866.
Entry LOAN.FRAME has been transported.
Entry LOAN.HELP has been transported.
Entry LOAN.KEYS has been transported.
Entry LOAN.PMENU has been transported.
Entry LOAN.SCL has been transported.

NOTE:
NOTE:
NOTE:
NOTE:

Proc CPORT begins to transport catalog SOURCE.FORMATS

The catalog has 2 entries and its maximum logical record length is 104.
Entry REVENUE.FORMAT has been transported.
Entry DEPT.FORMATC has been transported.

296

Example 2: Exporting Individual Catalog Entries

Chapter 13

Example 2: Exporting Individual Catalog Entries

Procedure features:

PROC CPORT statement options:

AFTER=
FILE=

This example shows how to use PROC CPORT to transport only the catalog entries
with modication dates equal to or later than the date you specify in the AFTER=
option.

Program

The CPORT Procedure

SAS Log

299

Specify the catalog entries to be written to the transport le. AFTER= species that only
catalog entries with modication dates on or after September 9, 1996, should be written to the
transport le.
proc cport catalog=source.finance file=tranfile
after=09sep1996d;
run;

SAS Log

PROC CPORT writes messages to the SAS log to inform you that it began the export process for
all the entries in the specied catalog. However, PROC CPORT wrote only the entries
LOAN.FRAME and LOAN.HELP in the FINANCE catalog to the transport le because only
those two entries had a modication date equal to or later than September 9, 1996. That is, of
all the entries in the specied catalog, only two met the requirement of the AFTER= option.

NOTE:
NOTE:
NOTE:
NOTE:

Proc CPORT begins to transport catalog SOURCE.FINANCE

The catalog has 5 entries and its maximum logical record length is 866.
Entry LOAN.FRAME has been transported.
Entry LOAN.HELP has been transported.

300

301

CHAPTER

14
The CV2VIEW Procedure
Information about the CV2VIEW Procedure

301

Information about the CV2VIEW Procedure

For complete documentation of the CV2VIEW procedure, see SAS/ACCESS for
Relational Databases: Reference.

See:

302

303

CHAPTER

15
The DATASETS Procedure
Overview: DATASETS Procedure 304
What Does the DATASETS Procedure Do? 304
Sample PROC DATASETS Output 305
Notes 306
Syntax: DATASETS Procedure 307
PROC DATASETS Statement 308
AGE Statement 312
APPEND Statement 313
AUDIT Statement 319
CHANGE Statement 322
CONTENTS Statement 323
COPY Statement 327
DELETE Statement 334
EXCHANGE Statement 338
EXCLUDE Statement 339
FORMAT Statement 339
IC CREATE Statement 340
IC DELETE Statement 343
IC REACTIVATE Statement 343
INDEX CENTILES 344
INDEX CREATE Statement 345
INDEX DELETE Statement 346
INFORMAT Statement 347
LABEL Statement 347
MODIFY Statement 348
RENAME Statement 352
REPAIR Statement 353
SAVE Statement 355
SELECT Statement 356
Concepts: DATASETS Procedure 357
Procedure Execution 357
Execution of Statements 357
RUN-Group Processing 357
Error Handling 358
Password Errors 359
Forcing a RUN Group with Errors to Execute 359
Ending the Procedure 359
Using Passwords with the DATASETS Procedure 359
Restricting Member Types for Processing 360
In the PROC DATASETS Statement 360
In Subordinate Statements 360

304

Overview: DATASETS Procedure

Chapter 15

Member Types 361

Restricting Processing for Generation Data Sets 362
Results: DATASETS Procedure 363
Directory Listing to the SAS Log 363
Directory Listing as SAS Output 364
Procedure Output 364
The CONTENTS Statement 364
Data Set Attributes 364
Engine and Operating Environment-Dependent Information
Alphabetic List of Variables and Attributes 366
Alphabetic List of Indexes and Attributes 367
Sort Information 367
PROC DATASETS and the Output Delivery System (ODS) 368
ODS Table Names 369
Output Data Sets 370
The CONTENTS Statement 370
The OUT= Data Set 370
The OUT2= Data Set 374
Examples: DATASETS Procedure 376
Example 1: Manipulating SAS Files 376
Example 2: Saving SAS Files from Deletion 380
Example 3: Modifying SAS Data Sets 381
Example 4: Describing a SAS Data Set 384
Example 5: Concatenating Two SAS Data Sets 386
Example 6: Aging SAS Data Sets 388
Example 7: PROC CONTENTS ODS Output 389

365

Overview: DATASETS Procedure

What Does the DATASETS Procedure Do?
The DATASETS procedure is a utility procedure that manages your SAS les. With
PROC DATASETS, you can

3 copy SAS les from one SAS library to another

3 rename SAS les
3 repair SAS les
3 delete SAS les
3 list the SAS les that are contained in a SAS library
3 list the attributes of a SAS data set, such as the date when the data was last
modied, whether the data is compressed, whether the data is indexed, and so on

3 manipulate passwords on SAS les

3 append SAS data sets
3 modify attributes of SAS data sets and variables within the data sets
3 create and delete indexes on SAS data sets
3 create and manage audit les for SAS data sets
3 create and delete integrity constraints on SAS data sets.

The DATASETS Procedure

Sample PROC DATASETS Output

The following DATASETS procedure
1 copies all data sets from the CONTROL library to the HEALTH library
2 lists the contents of the HEALTH library
3 deletes the SYNDROME data set from the HEALTH library
4 changes the name of the PRENAT data set to INFANT.

The SAS log is shown in Output 15.1.

libname control SAS-data-library-1;
libname health SAS-data-library-2;

proc datasets memtype=data;

copy in=control out=health;
run;
proc datasets library=health memtype=data details;
delete syndrome;
change prenat=infant;
run;
quit;

305

306

Notes

Output 15.1

Chapter 15

Log from PROC DATASETS

proc datasets library=health memtype=data details;

Directory
Libref
Engine
Physical Name
File Name

Member
#
1
2
3
4

external-file
external-file

Obs, Entries

File

Type

or Indexes

Vars

ALL

DATA

13312

29JAN2002:08:06:46

BODYFAT
CONFOUND
CORONARY

DATA
DATA
DATA

1
8
39

2
4
4

5120
5120
5120

29JAN2002:08:06:46
29JAN2002:08:06:46
29JAN2002:08:06:46

5
6
7

DRUG1
DRUG2
DRUG3

DATA
DATA
DATA

6
13
11

2
2
2

JAN95 Data
MAY95 Data
JUL95 Data

5120
5120
5120

29JAN2002:08:06:46
29JAN2002:08:06:46
29JAN2002:08:06:46

8
9

DRUG4
DRUG5

DATA
DATA

7
1

2
2

JAN92 Data
JUL92 Data

5120
5120

29JAN2002:08:06:46
29JAN2002:08:06:46

10
11
12

GROUP
MLSCL
NAMES

DATA
DATA
DATA

148
32
7

11
4
4

25600
5120
5120

29JAN2002:08:06:46
29JAN2002:08:06:46
29JAN2002:08:06:46

13
14
15

OXYGEN
PERSONL
PHARM

DATA
DATA
DATA

31
148
6

7
11
3

9216
25600
5120

29JAN2002:08:06:46
29JAN2002:08:06:46
29JAN2002:08:06:46

16
17
18

POINTS
PRENAT
RESULTS

DATA
DATA
DATA

6
149
10

6
6
5

5120
17408
5120

29JAN2002:08:06:46
29JAN2002:08:06:46
29JAN2002:08:06:46

19
20

SLEEP
SYNDROME

DATA
DATA

108
46

6
8

9216
9216

29JAN2002:08:06:46
29JAN2002:08:06:46

21
22
23

TENSION
TEST2
TRAIN

DATA
DATA
DATA

4
15
7

3
5
2

5120
5120
5120

29JAN2002:08:06:46
29JAN2002:08:06:46
29JAN2002:08:06:47

24
25
26
60
61
62

Name

HEALTH
V9

Label

VISION
WEIGHT
WGHT

DATA
DATA
DATA

16
83
83

3
13
13

5120
13312
13312

29JAN2002:08:06:47
29JAN2002:08:06:47
29JAN2002:08:06:47

Multiple Sclerosis Data

Sugar Study

California Results
California Results

Size

Last Modified

delete syndrome;
change prenat=infant;
run;

NOTE: Deleting HEALTH.SYNDROME (memtype=DATA).

NOTE: Changing the name HEALTH.PRENAT to HEALTH.INFANT (memtype=DATA).
63
quit;

Notes
3 Although the DATASETS procedure can perform some operations on catalogs,
generally the CATALOG procedure is the best utility to use for managing catalogs.
For documentation of PROC CATALOG, see Overview: CATALOG Procedure on
page 153.
3 The term member often appears as a synonym for SAS le. If you are unfamiliar
with SAS les and SAS libraries, refer to SAS Files Concepts in SAS Language
Reference: Concepts.
3 PROC DATASETS cannot work with sequential data libraries.

The DATASETS Procedure

Syntax: DATASETS Procedure

307

Syntax: DATASETS Procedure

Supports RUN-group processing.
Tip: Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Names: See: ODS Table Names on page 369
Reminder: See Chapter 3, Statements with the Same Function in Multiple Procedures,
on page 57 for details. You can also use any global statements. See Global Statements
on page 18 for a list.
See: DATASETS Procedure in the documentation for your operating environment.
Tip:

PROC DATASETS <option(s)>;

AGE current-name related-SAS-le(s)
</ <ALTER=alter-password>
<MEMTYPE=mtype>>;
APPEND BASE=<libref.>SAS-data-set
<APPENDVER=V6>
<DATA=< libref.>SAS-data-set>
<FORCE>;
AUDIT SAS-le <(SAS-password)>;
INITIATE
<AUDIT_ALL=NO|YES>;
<LOG <ADMIN_IMAGE=YES|NO>
<BEFORE_IMAGE=YES|NO>
<DATA_IMAGE=YES|NO>
<ERROR_IMAGE=YES|NO>>;
<USER_VAR variable-1 < variable-n>>;
AUDIT SAS-le <(<SAS-password> <GENNUM= integer>)>;
SUSPEND|RESUME|TERMINATE;
CHANGE old-name-1=new-name-1
<old-name-n=new-name-n >
</ <ALTER=alter-password>
< GENNUM=ALL|integer>
<MEMTYPE=mtype>>;
CONTENTS<option(s)>;
COPY OUT=libref-1
<CLONE|NOCLONE>
<CONSTRAINT=YES|NO>
<DATECOPY>
<FORCE>
<IN=libref-2>
<INDEX=YES|NO>
<MEMTYPE=(mtype(s))>
<MOVE <ALTER=alter-password>>;
EXCLUDE SAS-le(s) < / MEMTYPE=mtype>;
SELECT SAS-le(s)
< / <ALTER=alter-password>
< MEMTYPE= mtype>>;
DELETE SAS-le(s)
</ <ALTER=alter-password>

308

PROC DATASETS Statement

Chapter 15

<GENNUM=ALL|HIST|REVERT|integer>
<MEMTYPE=mtype>>;
EXCHANGE name-1=other-name-1
<name-n=other-name-n>
</ < ALTER=alter-password>
<MEMTYPE=mtype> >;
MODIFY SAS-le <(option(s))>
</ < CORRECTENCODING=encoding-value>
<DTC=SAS-date-time>
<GENNUM=integer>
<MEMTYPE=mtype>>;
FORMAT variable-list-1 <format-1>
<variable-list-n <format-n>>;
IC CREATE <constraint-name=> constraint
<MESSAGE=message-string < MSGTYPE=USER>>;
IC DELETE constraint-name(s)| _ALL_;
IC REACTIVATE foreign-key-name REFERENCES libref;
INDEX CENTILES index(s)
</ <REFRESH>
<UPDATECENTILES= ALWAYS|NEVER|integer>>;
INDEX CREATE index-specication(s)
</ <NOMISS>
<UNIQUE>
<UPDATECENTILES=ALWAYS|NEVER|integer>>;
INDEX DELETE index(s) | _ALL_;
INFORMAT variable-list-1 <informat-1>
<variable-list-n <informat-n>>;
LABEL variable-1=<label-1| >
<variable-n=<label-n| >>;
RENAME old-name-1=new-name-1
<old-name-n=new-name-n>;
REPAIR SAS-le(s)
</ < ALTER=alter-password>
<GENNUM=integer>
<MEMTYPE=mtype>>;
SAVE SAS-le(s) </ MEMTYPE=mtype>;

PROC DATASETS Statement

PROC DATASETS <option(s)>;

To do this

Use this option

Specify the procedure input library

LIBRARY=

Provide alter access to any alter-protected SAS

le in the SAS data library

ALTER=

The DATASETS Procedure

PROC DATASETS Statement

To do this

Use this option

Include information in the log about the number

of observations, number of variables, number of
indexes, and data set labels

DETAILS|NODETAILS

Force a RUN group to execute even when there

are errors

FORCE

Force an append operation

FORCE

Restrict processing for generation data sets

GENNUM=

Delete SAS les

KILL

Restrict processing to a certain type of SAS le

MEMTYPE=

Suppress the printing of the directory

NOLIST

Suppress error processing

NOWARN

Provide read, write, or alter access

PW=

Provide read access

309

READ=

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les in the SAS data library.
See also: Using Passwords with the DATASETS Procedure on page 359
DETAILS|NODETAILS

determines whether the following columns are written to the log:

Obs, Entries, or Indexes
gives the number of observations for SAS les of type AUDIT, DATA, and VIEW;
the number of entries for type CATALOG; and the number of les of type INDEX
that are associated with a data le, if any. If SAS cannot determine the number of
observations in a SAS data set, the value in this column is set to missing. For
example, in a very large data set, if the number of observations or deleted
observations exceeds the number that can be stored in a double-precision integer,
the count will show as missing. The value for type CATALOG is the total number
of entries. For other types, this column is blank.
Tip: The value for les of type INDEX includes both user-dened indexes and

indexes created by integrity constraints. To view index ownership and attribute

information, use PROC DATASETS with the CONTENTS statement and the
OUT2 option.
Vars
gives the number of variables for types AUDIT, DATA and VIEW. If SAS cannot
determine the number of variables in the SAS data set, the value in this column is
set to missing. For other types, this column is blank.
Label
contains the label associated with the SAS data set. This column prints a label
only for the type DATA.
The DETAILS option affects output only when a directory is specied and requires
read access to all read-protected SAS les in the SAS data library. If you do not

310

PROC DATASETS Statement

Chapter 15

supply the read password, the directory listing contains missing values for the
columns produced by the DETAILS option.
Default: If neither DETAILS or NODETAILS is specied, the default is the system

option setting. The default system option setting is NODETAILS.

If you are using the SAS windowing environment and specify the DETAILS
option for a library that contains read-protected SAS les, a requestor window
prompts you for each read password that you do not specify in the PROC
DATASETS statement. Therefore, you may want to assign the same read
password to all SAS les in the same SAS data library.

Tip:

Featured in:

Example 1 on page 376

FORCE

performs two separate actions:

3 forces a RUN group to execute even if errors are present in one or more
statements in the RUN group. See RUN-Group Processing on page 357 for a
discussion of RUN-group processing and error handling.

3 forces all APPEND statements to concatenate two data sets even when the
variables in the data sets are not exactly the same. The APPEND statement
drops the extra variables and issues a warning message. Without the FORCE
option, the procedure issues an error message and stops processing if you try to
perform an append operation with two SAS data sets whose variables are not
exactly the same. Refer to APPEND Statement on page 313 for more
information on the FORCE option.
GENNUM=ALL|HIST|REVERT|integer

restricts processing for generation data sets. Valid values are as follows:
ALL
for subordinate CHANGE and DELETE statements, refers to the base version and
all historical versions in a generation group.
HIST
for a subordinate DELETE statement, refers to all historical versions, but excludes
the base version in a generation group.
REVERT|0
for a subordinate DELETE statement, refers to the base version in a generation
group and changes the most current historical version, if it exists, to the base
version.
integer
for subordinate AUDIT, CHANGE, MODIFY, DELETE, and REPAIR statements,
refers to a specic version in a generation group. Specifying a positive number is
an absolute reference to a specic generation number that is appended to a data set
name; that is, gennum=2 species MYDATA#002. Specifying a negative number is
a relative reference to a historical version in relation to the base version, from the
youngest to the oldest; that is, gennum=-1 refers to the youngest historical version.
See also: Restricting Processing for Generation Data Sets on page 362
See also: Understanding Generation Data Sets in SAS Language Reference:

Concepts
KILL

deletes all SAS les in the SAS data library that are available for processing. The
MEMTYPE= option subsets the member types that the statement deletes.
CAUTION:

The KILL option deletes the SAS les immediately after you submit the statement.

The DATASETS Procedure

PROC DATASETS Statement

311

LIBRARY=libref

names the library that the procedure processes. This library is the procedure input
library.
Aliases: DDNAME=, DD=, LIB=
Default: WORK or USER. See Temporary and Permanent SAS Data Sets on page

16 for more information on the WORK and USER libraries.

Restriction: A SAS library that is accessed via a sequential engine (such as a tape
format engine) cannot be specied as the value of the LIBRARY= option.
Featured in: Example 1 on page 376
MEMTYPE=(mtype(s))

restricts processing to one or more member types and restricts the listing of the data
library directory to SAS les of the specied member types. For example, the
following PROC DATASETS statement limits processing to SAS data sets in the
default data library and limits the directory listing in the SAS log to SAS les of
member type DATA:
proc datasets memtype=data;

MTYPE=, MT=
Default: ALL
Aliases:

See the description of DETAILS on page 309.

NOLIST

suppresses the printing of the directory of the SAS les in the SAS log.
Featured in: Example 3 on page 381
Note: If you specify the ODS RTF destination, PROC DATASETS output will go
to both the SAS log and the ODS output area. The NOLIST option will suppress
output to both. To see the output only in the SAS log, use the ODS EXCLUDE
statement by specifying the member directory as the exclusion. 4
NOWARN

suppresses the error processing that occurs when a SAS le that is specied in a
SAVE, CHANGE, EXCHANGE, REPAIR, DELETE, or COPY statement or listed as
the rst SAS le in an AGE statement is not in the procedure input library. When an
error occurs and the NOWARN option is in effect, PROC DATASETS continues
processing that RUN group. If NOWARN is not in effect, PROC DATASETS stops
processing that RUN group and issues a warning for all operations except DELETE,
for which it does not stop processing.
PW= password

provides the password for any protected SAS les in the SAS data library. PW= can
act as an alias for READ=, WRITE=, or ALTER=.
See also: Using Passwords with the DATASETS Procedure on page 359
READ=read-password

provides the read-password for any read-protected SAS les in the SAS data library.
See also: Using Passwords with the DATASETS Procedure on page 359

312

AGE Statement

Chapter 15

AGE Statement
Renames a group of related SAS les in a library.
Featured in:

Example 6 on page 388

AGE current-name related-SAS-le(s)

</ <ALTER=alter-password>
<MEMTYPE=mtype>>;

Required Arguments
current-name

is a SAS le that the procedure renames. current-name receives the name of the rst
name in related-SAS-le(s).
related-SAS-le(s)

is one or more SAS les in the SAS data library.

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les named in the AGE
statement. Because an AGE statement renames and deletes SAS les, you need alter
access to use the AGE statement. You can use the option either in parentheses after
the name of each SAS le or after a forward slash.
See also: Using Passwords with the DATASETS Procedure on page 359
MEMTYPE=mtype

restricts processing to one member type. All of the SAS les that you name in the
AGE statement must be the same member type. You can use the option either in
parentheses after the name of each SAS le or after a forward slash.
Aliases:

MTYPE=, MT=

Default: If you do not specify MEMTYPE= in the PROC DATASETS statement, the

default is DATA.
See also: Restricting Member Types for Processing on page 360

Details
3 The AGE statement renames current-name to the name of the rst name in
related-SAS-le(s), renames the rst name in related-SAS-le(s) to the second
name in related-SAS-le(s), and so on until it changes the name of the next-to-last
SAS le in related-SAS-le(s) to the last name in related-SAS-le(s). The AGE
statement then deletes the last le in related-SAS-le(s).

3 If the rst SAS le named in the AGE statement does not exist in the SAS data
library, PROC DATASETS stops processing the RUN group containing the AGE
statement and issues an error message. The AGE statement does not age any of

The DATASETS Procedure

APPEND Statement

313

the related-SAS-le(s). To override this behavior, use the NOWARN option in the
PROC DATASETS statement.
If one of the related-SAS-le(s) does not exist, the procedure prints a warning
message to the SAS log but continues to age the SAS les that it can.

3 If you age a data set that has an index, the index continues to correspond to the
data set.

3 You can age only entire generation groups. For example, if data sets A and B have
generation groups, then the following statement deletes generation group B and
ages (renames) generation group A to the name B:
age a b;

For example, suppose the generation group for data set A has 3 historical versions
and the generation group for data set B has 2 historical versions. Then aging A to
B has this effect:
Old Name

Version

New Name

Version

base

is deleted

APPEND Statement
Adds the observations from one SAS data set to the end of another SAS data set.
Reminder: You can use data set options with the BASE= and DATA= options. See Data

Set Options on page 18 for a list. You can also use any global statements as well. See
Global Statements on page 18.
Requirement: The BASE= data set must be a member of a SAS library that supports
update processing.

If the BASE= data set is accessed through a SAS server and if no other user
has the data set open at the time the APPEND statement begins processing, the
BASE= data set defaults to CNTLLEV=MEMBER (member-level locking). When this
happens, no other user can update the le while the data set is processed.
Tip: If a failure occurs during processing, the data set is marked as damaged and is
reset to its pre-append condition at the next REPAIR statement. If the data set has an
index, the index is not updated with each observation but is updated once at the end.
(This is Version 7 and later behavior, as long as APPENDVER=V6 is not set.)
Default:

Featured in: Example 5 on page 386

APPEND BASE=< libref.>SAS-data-set

<APPENDVER=V6>

314

APPEND Statement

Chapter 15

<DATA=<libref.>SAS-data-set>
<FORCE>;

Required Arguments
BASE=<libref.> SAS-data-set

names the data set to which you want to add observations.

libref
species the library that contains the SAS data set. If you omit the libref, the
default is the libref for the procedure input library. If you are using PROC
APPEND, the default for libref is either WORK or USER.
SAS-data-set
names a SAS data set. If the APPEND statement cannot nd an existing data set
with this name, it creates a new data set in the library. That is, you can use the
APPEND statement to create a data set by specifying a new data set name in the
BASE= argument.
The BASE= data set is the current SAS data set after all append operations
regardless of whether you are creating a new data set or appending to an existing
data set.
Alias: OUT=
Featured in: Example 5 on page 386

Options
APPENDVER=V6

uses the Version 6 behavior for appending observations to the BASE= data set, which
is to append one observation at a time. Beginning in Version 7, to improve
performance, the default behavior changed so that all observations are appended
after the data set is processed.
See also: Appending to an Indexed Data Set Fast-Append Method on page 316
DATA=<libref.> SAS-data-set

names the SAS data set containing observations that you want to append to the end
of the SAS data set specied in the BASE= argument.
libref
species the library that contains the SAS data set. If you omit libref, the default
is the libref for the procedure input library. The DATA= data set can be from any
SAS data library, but you must use the two-level name if the data set resides in a
library other than the procedure input library.
SAS-data-set
names a SAS data set. If the APPEND statement cannot nd an existing data set
with this name, it stops processing.
Alias: NEW=
Default: the most recently created SAS data set, from any SAS data library
See also: Appending with Generation Groups on page 318
Featured in: Example 5 on page 386
FORCE

forces the APPEND statement to concatenate data sets when the DATA= data set
contains variables that either

The DATASETS Procedure

APPEND Statement

315

3 are not in the BASE= data set

3 do not have the same type as the variables in the BASE= data set
3 are longer than the variables in the BASE= data set.
See also: Appending to Data Sets with Different Variables on page 317
See also: Appending to Data Sets That Contain Variables with Different

Attributes on page 317

Featured in: Example 5 on page 386
Tip: You can use the GENNUM= data set option to append to or from a specic

version in a generation group. Here are some examples:

/* appends historical version to base A */
proc datasets;
append base=a
data=a (gennum=2);
/* appends current version of A to historical version */
proc datasets;
append base=a (gennum=1)
data=a;

Restricting the Observations That Are Appended

You can use the WHERE= data set option with the DATA= data set in order to
restrict the observations that are appended. Likewise, you can use the WHERE
statement in order to restrict the observations from the DATA= data set. The WHERE
statement has no effect on the BASE= data set. If you use the WHERE= data set option
with the BASE= data set, WHERE= has no effect.
CAUTION:

For an existing BASE= data set: If there is a WHERE statement on the BASE= data
set, it will take effect only if the WHEREUP= option is set to YES. 4
CAUTION:

For the non-existent BASE= data set: If there is a WHERE statement on the
non-existent BASE= data set, regardless of the WHEREUP option setting, you use
the WHERE statement. 4
Note: You cannot append a data set to itself by using the WHERE= data set
option. 4

Choosing between the SET Statement and the APPEND Statement

If you use the SET statement in a DATA step to concatenate two data sets, SAS must
process all the observations in both data sets to create a new one. The APPEND
statement bypasses the processing of data in the original data set and adds new
observations directly to the end of the original data set. Using the APPEND statement
can be more efcient than using a SET statement if
3 the BASE= data set is large
3 all variables in the BASE= data set have the same length and type as the
variables in the DATA= data set and if all variables exist in both data sets.
Note: You can use the CONTENTS statement to see the variable lengths and
types.

The APPEND statement is especially useful if you frequently add observations to a

SAS data set (for example, in production programs that are constantly appending data
to a journal-type data set).

316

APPEND Statement

Chapter 15

Appending Password-Protected SAS Data Sets

In order to use the APPEND statement, you need read access to the DATA= data set
and write access to the BASE= data set. To gain access, use the READ= and WRITE=
data set options in the APPEND statement the way you would use them in any other
SAS statement, which is in parentheses immediately after the data set name. When
you are appending password-protected data sets, use the following guidelines:

3 If you do not give the read password for the DATA= data set in the APPEND
statement, by default the procedure looks for the read password for the DATA=
data set in the PROC DATASETS statement. However, the procedure does not
look for the write password for the BASE= data set in the PROC DATASETS
statement. Therefore, you must specify the write password for the BASE= data set
in the APPEND statement.

3 If the BASE= data set is read-protected only, you must specify its read password in
the APPEND statement.

Appending to a Compressed Data Set

You can concatenate compressed SAS data sets. Either or both of the BASE= and
DATA= data sets can be compressed. If the BASE= data set allows the reuse of space
from deleted observations, the APPEND statement may insert the observations into the
middle of the BASE= data set to make use of available space.
For information on the COMPRESS= and REUSE= data set and system options, see
SAS Language Reference: Dictionary.

Appending to an Indexed Data Set Fast-Append Method

Beginning with Version 7, the behavior of appending to an indexed data set changed
to improve performance.

3 In Version 6, when you appended to an indexed data set, the index was updated
for each added observation. Index updates tend to be random; therefore, disk I/O
could have been high.

3 Currently, SAS does not update the index until all observations are added to the
data set. After the append, SAS internally sorts the observations and inserts the
data into the index in sequential order, which reduces most of the disk I/O and
results in a faster append method.
The fast-append method is used by default when the following requirements are met;
otherwise, the Version 6 method is used:

3 The BASE= data set is open for member-level locking. If CNTLLEV= is set to
record, then the fast-append code is not used.

3 The BASE= data set does not contain referential integrity constraints.
3 The BASE= data set is not accessed using the Cross Environment Data Access
(CEDA) facility.

3 The BASE= data set is not using a WHERE= data set option.
To display information in the SAS log about the append method that is being used,
you can specify the MSGLEVEL= system option as follows:
options msglevel=i;

Either a message displays if the fast-append method is in use or a message or messages

display as to why the fast-append method is not in use.
The current append method initially adds observations to the BASE= data set
regardless of the restrictions that are determined by the index. For example, a variable
that has an index that was created with the UNIQUE option does not have its values

The DATASETS Procedure

APPEND Statement

317

validated for uniqueness until the index is updated. Then, if a nonunique value is
detected, the offending observation is deleted from the data set. This means that after
observations are appended, some of them may subsequently be deleted.
For a simple example, consider that the BASE= data set has ten observations
numbered from 1 to 10 with a UNIQUE index for the variable ID. You append a data
set that contains ve observations numbered from 1 to 5, and observations 3 and 4 both
contain the same value for ID. The following occurs
1 After the observations are appended, the BASE= data set contains 15 observations
numbered from 1 to 15.
2 SAS updates the index for ID, validates the values, and determines that
observations 13 and 14 contain the same value for ID.
3 SAS deletes one of the observations from the BASE= data set, resulting in 14
observations that are numbered from 1 to 15. For example, observation 13 is
deleted. Note that you cannot predict which observation will be deleted, because
the internal sort may place either observation rst. (In Version 6, you could
predict that observation 13 would be added and observation 14 would be rejected.)
If you do not want the current behavior (which could result in deleted observations)
or if you want to be able to predict which observations are appended, request the
Version 6 append method by specifying the APPENDVER=V6 option:
proc datasets;
append base=a data=b appendver=v6;
run;

Note: In Version 6, deleting the index and then recreating it after the append could
improve performance. The current method may eliminate the need to do that. However,
the performance depends on the nature of your data. 4

Appending to Data Sets with Different Variables

If the DATA= data set contains variables that are not in the BASE= data set, use the
FORCE option in the APPEND statement to force the concatenation of the two data
sets. The APPEND statement drops the extra variables and issues a warning message.
If the BASE= data set contains a variable that is not in the DATA= data set, the
APPEND statement concatenates the data sets, but the observations from the DATA=
data set have a missing value for the variable that was not present in the DATA= data
set. The FORCE option is not necessary in this case.

Appending to Data Sets That Contain Variables with Different Attributes

3 If a variable has different attributes in the BASE= data set than it does in the
DATA= data set, the attributes in the BASE= data set prevail.
3 If formats in the DATA= data set are different from those in the BASE= data set,
then the formats in the BASE= data set are used. However, SAS does not convert
the data from the DATA= data set in order to be consistent with the formats in the
BASE= data set. The result could be data that appears to be incorrect. A warning
message is displayed in the SAS log. The following example illustrates appending
data by using different formats:
data format1;
input Date date9.;
format Date date9.;
datalines;
24sep1975

318

APPEND Statement

Chapter 15

22may1952
;
data format2;
input Date datetime20.;
format Date datetime20.;
datalines;
25aug1952:11:23:07.4
;
proc append base=format1 data=format2;
run;

The following messages are displayed in the SAS log.

Output 15.2

Warning Message in SAS Log

NOTE: Appending WORK.FORMAT2 to WORK.FORMAT1.

WARNING: Variable Date has format DATE9. on the BASE data set
and format DATETIME20. on the DATA data set. DATE9. used.
NOTE: There were 1 observations read from the data set WORK.FORMAT2.
NOTE: 1 observations added.
NOTE: The data set WORK.FORMAT1 has 3 observations and 1 variables.

3 If the length of a variable is longer in the DATA= data set than in the BASE= data
set, or if the same variable is a character variable in one data set and a numeric
variable in the other, use the FORCE option. Using FORCE has these
consequences:
3 The length of the variables in the BASE= data set takes precedence. SAS
truncates values from the DATA= data set to t them into the length that is
specied in the BASE= data set.
3 The type of the variables in the BASE= data set takes precedence. The
APPEND statement replaces values of the wrong type (all values for the
variable in the DATA= data set) with missing values.

Appending Data Sets That Contain Integrity Constraints

If the DATA= data set contains integrity constraints and the BASE= data set does
not exist, the APPEND statement copies the general constraints. Note that the
referential constraints are not copied. If the BASE= data set exists, the APPEND action
copies only observations.

Appending with Generation Groups

You can use the GENNUM= data set option to append to a specic version in a
generation group. Here are examples:

The DATASETS Procedure

AUDIT Statement

SAS Statements

Result

proc datasets;
append base=a
data=b(gennum=2);

appends historical version B#002 to base A

proc datasets;
append base=a(gennum=2)
data=b(gennum=2);

319

appends historical version B#002 to

historical version A#002

Using the APPEND Procedure instead of the APPEND Statement

The only difference between the APPEND procedure and the APPEND statement in
PROC DATASETS, is the default for libref in the BASE= and DATA= arguments. For
PROC APPEND, the default is either WORK or USER. For the APPEND statement,
the default is the libref of the procedure input library.

System Failures
If a system failure or some other type of interruption occurs while the procedure is
executing, the append operation may not be successful; it is possible that not all,
perhaps none, of the observations will be added to the BASE= data set. In addition, the
BASE= data set may suffer damage. The APPEND operation performs an update in
place, which means that it does not make a copy of the original data set before it begins
to append observations. If you want to be able to restore the original observations, you
can initiate an audit trail for the base data le and select to store a before-update
image of the observations. Then you can write a DATA step to extract and reapply the
original observations to the data le. For information about initiating an audit trail,
see the PROC DATASETS AUDIT Statement on page 319.

AUDIT Statement
Initiates and controls event logging to an audit le as well as suspends, resumes, or terminates
event logging in an audit le.
See also: Understanding an Audit Trail in SAS Language Reference: Concepts

The AUDIT statement takes one of two forms, depending on whether you are
initiating the audit trail or suspending, resuming, or terminating event logging in an
audit le.

Tip:

AUDIT SAS-le < (SAS-password)>;

320

AUDIT Statement

Chapter 15

AUDIT SAS-le <(<SAS-password> <GENNUM= integer>)>;

SUSPEND|RESUME|TERMINATE;

Required Arguments and Statements

SAS-le

species the SAS data le in the procedure input library that you want to audit.
INITIATE

creates an audit le that has the same name as the SAS data le and a data set type
of AUDIT. The audit le logs additions, deletions, and updates to the SAS data le.
You must initiate an audit trail before you can suspend, resume, or terminate it.

Options
SAS-password

species the password for the SAS data le, if one exists. The parentheses are
required.
GENNUM=integer

species that the SUSPEND, RESUME, or TERMINATE action be performed on the

audit trail of a generation le. You cannot initiate an audit trail on a generation le.
Valid values for GENNUM= are integer, which is a number that references a specic
version from a generation group. Specifying a positive number is an absolute
reference to a specic generation number that is appended to a data sets name; that
is, gennum=2 species MYDATA#002. Specifying a negative number is a relative
reference to a historical version in relation to the base version, from the youngest to
the oldest; that is, gennum=-1 refers to the youngest historical version. Specifying 0,
which is the default, refers to the base version. The parentheses are required.
AUDIT_ALL=NO|YES

species whether logging can be suspended and audit settings can be changed.
AUDIT_ALL=YES species that all images are logged and cannot be suspended.
That is, you cannot use the LOG statement to turn off logging of particular images,
and you cannot suspend event logging by using the SUSPEND statement. To turn off
logging, you must use the TERMINATE statement, which terminates event logging
and deletes the audit le.
Default: NO
LOG

species the audit settings:

ADMIN_IMAGE=YES|NO
controls the logging of administrative events to the audit le (that is, the
SUSPEND and RESUME actions).
BEFORE_IMAGE=YES|NO
controls the storage of before-update record images.
DATA_IMAGE=YES|NO
controls the storage of added, deleted, and after-update record images.
ERROR_IMAGE=YES|NO
controls the storage of unsuccessful after-update record images.
Default: All images are logged by default; that is, all four are set to YES.

The DATASETS Procedure

AUDIT Statement

321

If you do not want to log a particular image, specify NO for the image type.
For example, the following code turns off logging the error images, but the
administrative, before, and data images continue to be logged:

Tip:

log error_image=no;

USER_VAR variable-1 < variable-n>

denes optional variables to be logged in the audit le with each update to an

observation. The syntax for dening variables is
USER_VAR variable-name-1 <$> <length> < LABEL=variable-label >
< variable-name-n <$> <length> <LABEL=variable-label > >
where
variable-name
is a name for the variable.
$
indicates that the variable is a character variable.
length
species the length of the variable. If a length is not specied, the default is 8.
LABEL=variable-label
species a label for the variable.
You can dene attributes such as format and informat for the user variables in the
data le by using the PROC DATASETS MODIFY statement.
SUSPEND

suspends event logging to the audit le, but does not delete the audit le.
RESUME

resumes event logging to the audit le, if it was suspended.

TERMINATE

terminates event logging and deletes the audit le.

Creating an Audit File

The following example creates the audit le MYLIB.MYFILE.AUDIT to log updates
to the data le MYLIB.MYFILE.DATA, storing all available record images:
proc datasets library=MyLib;
audit MyFile (alter=MyPassword);
initiate;
run;

The following example creates the same audit le but stores only error record images:
proc datasets library=MyLib;
audit MyFile (alter=MyPassword);
initiate
log data_image=NO before_image=NO;
run;

322

CHANGE Statement

Chapter 15

CHANGE Statement
Renames one or more SAS les in the same SAS data library.
Featured in:

Example 1 on page 376

CHANGE old-name-1=new-name-1
<old-name-n=new-name-n >
</ <ALTER=alter-password>
<GENNUM=ALL|integer>
<MEMTYPE=mtype>>;

Required Arguments
old-name=new-name

changes the name of a SAS le in the input data library. old-name must be the name
of an existing SAS le in the input data library.
Featured in:

Example 1 on page 376

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les named in the CHANGE
statement. Because a CHANGE statement changes the names of SAS les, you need
alter access to use the CHANGE statement for new-name. You can use the option
either in parentheses after the name of each SAS le or after a forward slash.
See also: Using Passwords with the DATASETS Procedure on page 359
GENNUM=ALL|integer

restricts processing for generation data sets. You can use the option either in
parentheses after the name of each SAS le or after a forward slash. Valid values are
ALL | 0
refers to the base version and all historical versions of a generation group.
integer
refers to a specic version from a generation group. Specifying a positive number
is an absolute reference to a specic generation number that is appended to a data
sets name; that is, gennum=2 species MYDATA#002. Specifying a negative
number is a relative reference to a historical version in relation to the base
version, from the youngest to the oldest; that is, gennum=-1 refers to the youngest
historical version.
For example, the following statements change the name of version A#003 to base B:
proc datasets;
change A=B / gennum=3;
proc datasets;
change A(gennum=3)=B;

The following CHANGE statement produces an error:

The DATASETS Procedure

CONTENTS Statement

323

proc datasets;
change A(gennum=3)=B(gennum=3);

See also: Restricting Processing for Generation Data Sets on page 362
See also: Understanding Generation Data Sets in SAS Language Reference:

Concepts
MEMTYPE=mtype

restricts processing to one member type. You can use the option either in
parentheses after the name of each SAS le or after a forward slash.
Aliases: MTYPE=, MT=
Default: If you do not specify MEMTYPE= in the PROC DATASETS statement, the
default is MEMTYPE=ALL.
See also: Restricting Member Types for Processing on page 360

Details
3 The CHANGE statement changes names by the order that the old-names occur in
the directory listing, not in the order that you list the changes in the CHANGE
statement.
3 If the old-name SAS le does not exist in the SAS data library, PROC DATASETS
stops processing the RUN group containing the CHANGE statement and issues an
error message. To override this behavior, use the NOWARN option in the PROC
DATASETS statement.
3 If you change the name of a data set that has an index, the index continues to
correspond to the data set.

CONTENTS Statement
Describes the contents of one or more SAS data sets and prints the directory of the SAS data
library.
Reminder: You can use data set options with the DATA=, OUT=, and OUT2= options.

See Data Set Options on page 18 for a list. You can use any global statements as well.
See Global Statements on page 18.
Featured in: Example 4 on page 384

CONTENTS <option(s)>;

To do this

Use this option

Specify the input data set

DATA=

Specify the name for an output data set

OUT=

Specify the name of an output data set to contain

information about indexes and integrity constraints

OUT2=

Include information in the output about the number

of observations, number of variables, number of
indexes, and data set labels

DETAILS|NODETAILS

324

CONTENTS Statement

Chapter 15

To do this

Use this option

Print a list of the SAS les in the SAS data library

Print the length of a variables informat or format

FMTLEN

Restrict processing to one or more types of SAS les

MEMTYPE=

Suppress the printing of individual les

NODS

Suppress the printing of the output

NOPRINT

Print a list of the variables by their position in the

data set. By default, the CONTENTS statement lists
the variables alphabetically.

VARNUM

Print a list of variables in alphabetical order even if

they include mixed-case names

ORDER=IGNORECASE

Print abbreviated output

SHORT

Print centiles information for indexed variables

CENTILES

Options
CENTILES

prints centiles information for indexed variables.

The following additional elds are printed in the default report of PROC
CONTENTS when the CENTILES option is selected and an index exists on the data
set. Note that the additional elds depend on whether the index is simple or complex.
#

number of the index on the data set.

Index

name of the index.

Update Centiles

percent of the data values that must be changed before the

CENTILES for the indexed variables are automatically updated.

Current Update
Percent

percent of index updated since CENTILES were refreshed.

# of Unique
Values

number of unique indexed values.

Variables

names of the variables used to make up the index. Centile

information is listed below the variables.

DATA=SAS-le-specication

species an entire library or a specic SAS data set within a library.

SAS-le-specication can take one of the following forms:
<libref.>SAS-data-set
names one SAS data set to process. The default for libref is the libref of the
procedure input library. For example, to obtain the contents of the SAS data set
HTWT from the procedure input library, use the following CONTENTS statement:
contents data=HtWt;

To obtain the contents of a specic version from a generation group, use the
GENNUM= data set option as shown in the following CONTENTS statement:
contents data=HtWt(gennum=3);

The DATASETS Procedure

CONTENTS Statement

325

<libref.>_ALL_
gives you information about all SAS data sets that have the type or types specied
by the MEMTYPE= option. libref refers to the SAS data library. The default for
libref is the libref of the procedure input library.

3 If you are using the _ALL_ keyword, you need read access to all
read-protected SAS data sets in the SAS data library.

3 DATA=_ALL_ automatically prints a listing of the SAS les that are

contained in the SAS library. Note that for SAS views, all librefs that are
associated with the views must be assigned in the current session in order for
them to be processed for the listing.
Default: most recently created data set in your job or session, from any SAS data

library.
If you specify a read-protected data set in the DATA= option but do not give
the read password, by default the procedure looks in the PROC DATASETS
statement for the read password. However, if you do not specify the DATA= option
and the default data set (last one created in the session) is read protected, the
procedure does not look in the PROC DATASETS statement for the read password.

Tip:

Featured in:

Example 4 on page 384

DETAILS|NODETAILS

DETAILS includes these additional columns of information in the output, but only if
DIRECTORY is also specied.
Default: If neither DETAILS or NODETAILS is specied, the defaults are as

follows: for the CONTENTS procedure, the default is the system option setting,
which is NODETAILS; for the CONTENTS statement, the default is whatever is
specied on the PROC DATASETS statement, which also defaults to the system
option setting.
See also: description of the additional columns in Options in PROC DATASETS

Statement on page 308

restricts processing to one or more member types. The CONTENTS statement

produces output only for member types DATA, VIEW, and ALL, which includes DATA
and VIEW.
MEMTYPE= in the CONTENTS statement differs from MEMTYPE= in most of
the other statements in the DATASETS procedure in the following ways:

3 A slash does not precede the option.

3 You cannot enclose the MEMTYPE= option in parentheses to limit its effect to
only the SAS le immediately preceding it.
MEMTYPE= results in a directory of the library in which the DATA= member is
located. However, MEMTYPE= does not limit the types of members whose contents
are displayed unless the _ALL_ keyword is used in the DATA= option. For example,

326

CONTENTS Statement

Chapter 15

the following statements produce the contents of only the SAS data sets with the
member type DATA:
proc datasets memtype=data;
contents data=_all_;
run;

Aliases: MT=, MTYPE=

Default: DATA
NODS

suppresses printing the contents of individual les when you specify _ALL_ in the
DATA= option. The CONTENTS statement prints only the SAS data library
directory. You cannot use the NODS option when you specify only one SAS data set
in the DATA= option.
NODETAILS

See the description of DETAILS|NODETAILS on page 325.

NOPRINT

suppresses printing the output of the CONTENTS statement.

ORDER= IGNORECASE | VARNUM

IGNORECASE

prints a list of variables in alphabetical order even if they include

mixed-case names.

VARNUM

is the same as the VARNUM option. See VARNUM.

OUT=SAS-data-set

names an output SAS data set.

Tip: OUT= does not suppress the printed output from the statement. If you want to
suppress the printed output, you must use the NOPRINT option.
See: The OUT= Data Set on page 370 for a description of the variables in the
OUT= data set.
See also: Example 7 on page 389 for an example of how to get the CONTENTS
output into an ODS data set for processing.
OUT2=SAS-data-set

names the output data set to contain information about indexes and integrity
constraints.
Tip: If UPDATECENTILES was not specied in the index denition, then the
default value of 5 is used in the RECREATE variable of the OUT2 data set.
Tip: OUT2= does not suppress the printed output from the statement. To suppress
the printed output, use the NOPRINT option.
See also: The OUT2= Data Set on page 374 for a description of the variables in
the OUT2= data set.
SHORT

prints only the list of variable names, the index information, and the sort
information for the SAS data set.
VARNUM

prints a list of the variable names in the order of their logical position in the data
set. By default, the CONTENTS statement lists the variables alphabetically. The
physical position of the variable in the data set is engine-dependent.

Details
The CONTENTS statement prints an alphabetical listing of the variables by default,
except for variables in the form of a numbered range list. Numbered range lists, such

The DATASETS Procedure

COPY Statement

327

as x1x100, are printed in incrementing order, that is, x1x100. For more information,
see Alphabetic List of Variables and Attributes on page 366.

Requesting CONTENTS Output for a Password-Protected File with Integrity

Constraints
For a SAS data le with dened referential integrity constraints that is also password
protected, some SAS requests require that both les be open in order to process the
request. If both les are password protected, then both passwords must be provided.
For example, suppose you want to execute the CONTENTS procedure for a data le
with a primary key that is referenced by a foreign key. You must provide the password
for the primary key data le as well as the password for the referential data le,
because in order to obtain the information for the CONTENTS output for the primary
key data le, SAS must open both les.
For an example, see Understanding Integrity Constraints in SAS Language
Reference: Concepts.

Using the CONTENTS Procedure instead of the CONTENTS Statement

The only difference between the CONTENTS procedure and the CONTENTS
statement in PROC DATASETS is the default for libref in the DATA= option. For
PROC CONTENTS, the default is either WORK or USER. For the CONTENTS
statement, the default is the libref of the procedure input library.

COPY Statement
Copies all or some of the SAS les in a SAS library.
Featured in: Example 1 on page 376

COPY OUT=libref-1
<CLONE|NOCLONE>
<CONSTRAINT=YES|NO>
<DATECOPY>
<FORCE>
<IN=libref-2>
<INDEX=YES|NO>
<MEMTYPE=(mtype(s))>
<MOVE < ALTER=alter-password>> ;

Required Arguments
OUT=libref-1

names the SAS library to copy SAS les to.

Aliases:

OUTLIB= and OUTDD=

Featured in:

Example 1 on page 376

328

COPY Statement

Chapter 15

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les that you are moving
from one data library to another. Because the MOVE option deletes the SAS le from
the original data library, you need alter access to move the SAS le.
See also: Using Passwords with the DATASETS Procedure on page 359
CLONE|NOCLONE

species whether to copy the following data set attributes:

3 size of input/output buffers
3 whether the data set is compressed
3 whether free space is reused

3 data representation of input data set, library, or operating environment

3 encoding value.
These attributes are specied with data set options, SAS system options, and
LIBNAME statement options:

3
3
3
3
3

BUFSIZE= value for the size of the input/output buffers

COMPRESS= value for whether the data set is compressed
REUSE= value for whether free space is reused
OUTREP= value for data representation
ENCODING= or INENCODING= for encoding value.

For the BUFSIZE= attribute, the following table summarizes how the COPY
statement works:
Table 15.1

CLONE and the Buffer Page Size Attribute

If you use

the COPY statement

CLONE

uses the BUFSIZE= value from the input data set for the output data
set.

NOCLONE

uses the current setting of the SAS system option BUFSIZE= for the
output data set.

neither

determines the type of access method, sequential or random, used by

the COPY statement

CLONE

results in a copy that uses the encoding of the input data set or, if
specied, the value of the INENCODING= option in the LIBNAME
statement for the input library.

NOCLONE

results in a copy that uses the encoding of the current session

encoding or, if specied, the value of the OUTENCODING= option in
the LIBNAME statement for the output library.

neither

default is CLONE.

All data that is stored, transmitted, or processed by a computer is in an encoding.

An encoding maps each character to a unique numeric representation. An encoding
is a combination of a character set with an encoding method. A character set is the
repertoire of characters and symbols that are used by a language or group of
languages. An encoding method is the set of rules that are used to assign the
numbers to the set of characters that will be used in a encoding.

330

COPY Statement

Chapter 15

CONSTRAINT=YES|NO

species whether to copy all integrity constraints when copying a data set.
Default: NO
Tip: For data sets with integrity constraints that have a foreign key, the COPY
statement copies the general and referential constraints if CONSTRAINT=YES is
specied and the entire library is copied. If you use the SELECT or EXCLUDE
statement to copy the data sets, then the referential integrity constraints are not
copied. For more information, see Understanding Integrity Constraints in SAS
Language Reference: Concepts.
DATECOPY

copies the SAS internal date and time when the SAS le was created and the date
and time when it was last modied to the resulting copy of the le. Note that the
operating environment date and time are not preserved.
Restriction: DATECOPY cannot be used with encrypted les or catalogs.
Restriction: DATECOPY can be used only when the resulting SAS le uses the V8
or V9 engine.
Tip: You can alter the le creation date and time with the DTC= option on the
MODIFY statement. See MODIFY Statement on page 348.
Tip: If the le that you are copying has attributes that require additional
processing, the last modied date is changed to the current date. For example,
when you copy a data set that has an index, the index must be rebuilt, and this
changes the last modied date to the current date. Other attributes that require
additional processing and that could affect the last modied date include integrity
constraints and a sort indicator.
FORCE

allows you to use the MOVE option for a SAS data set on which an audit trail exists.
Note:

The AUDIT le is not moved with the audited data set.

IN=libref-2

names the SAS library containing SAS les to copy.

Aliases: INLIB= and INDD=
Default: the libref of the procedure input library
To copy only selected members, use the SELECT or EXCLUDE statements.
INDEX=YES|NO

species whether to copy all indexes for a data set when copying the data set to
another SAS data library.
Default: YES
MEMTYPE=(mtype(s))

restricts processing to one or more member types.

Aliases: MT=, MTYPE=
Default: If you omit MEMTYPE= in the PROC DATASETS statement, the default
is MEMTYPE=ALL.
See also: Specifying Member Types When Copying or Moving SAS Files on page
331
See also: Member Types on page 361
Featured in: Example 1 on page 376
MOVE

moves SAS les from the input data library (named with the IN= option) to the
output data library (named with the OUT= option) and deletes the original les from
the input data library.

The DATASETS Procedure

COPY Statement

331

Restriction: The MOVE option can be used to delete a member of a SAS library

only if the IN= engine supports the deletion of tables. A tape format engine does
not support table deletion. If you use a tape format engine, SAS suppresses the
MOVE operation and prints a warning.
Featured in:

Example 1 on page 376

NOCLONE

See the description of CLONE.

Copying an Entire Library

To copy an entire SAS data library, simply specify an input data library and an output
data library following the COPY statement. For example, the following statements copy
all the SAS les in the SOURCE data library into the DEST data library:
proc datasets library=source;
copy out=dest;
run;

Copying Selected SAS Files

To copy selected SAS les, use a SELECT or EXCLUDE statement. For more
discussion of using the COPY statement with a SELECT or an EXCLUDE statement,
see Specifying Member Types When Copying or Moving SAS Files on page 331 and
see Example 1 on page 376 for an example. Also, see EXCLUDE Statement on page
339 and SELECT Statement on page 356.
You can also select or exclude an abbreviated list of members. For example, the
following statement selects members TABS, TEST1, TEST2, and TEST3:
select tabs test1-test3;

Also, you can select a group of members whose names begin with the same letter or
letters by entering the common letters followed by a colon (:). For example, you can
select the four members in the previous example and all other members having names
that begin with the letter T by specifying the following statement:
select t:;

You specify members to exclude in the same way that you specify those to select.
That is, you can list individual member names, use an abbreviated list, or specify a
common letter or letters followed by a colon (:). For example, the following statement
excludes the members STATS, TEAMS1, TEAMS2, TEAMS3, TEAMS4 and all the
members that begin with the letters RBI from the copy operation:
exclude stats teams1-teams4 rbi:;

Note that the MEMTYPE= option affects which types of members are available to be
selected or excluded.
When a SELECT or EXCLUDE statement is used with CONSTRAINT=YES, only the
general integrity constraints on the data sets are copied. Any referential integrity
constraints are not copied. For more information, see Understanding Integrity
Constraints in SAS Language Reference: Concepts.

Specifying Member Types When Copying or Moving SAS Files

The MEMTYPE= option in the COPY statement differs from the MEMTYPE= option
in other statements in the procedure in several ways:

3 A slash does not precede the option.

332

COPY Statement

Chapter 15

3 You cannot limit its effect to the member immediately preceding it by enclosing
the MEMTYPE= option in parentheses.

3 The SELECT and EXCLUDE statements and the IN= option (in the COPY
statement) affect the behavior of the MEMTYPE= option in the COPY statement
according to the following rules:
1 MEMTYPE= in a SELECT or EXCLUDE statement takes precedence over

the MEMTYPE= option in the COPY statement. The following statements

copy only VISION.CATALOG and NUTR.DATA from the default data library
to the DEST data library; the MEMTYPE= value in the rst SELECT
statement overrides the MEMTYPE= value in the COPY statement.
proc datasets;
copy out=dest memtype=data;
select vision(memtype=catalog) nutr;
run;

2 If you do not use the IN= option, or you use it to specify the library that

happens to be the procedure input library, the value of the MEMTYPE=

option in the PROC DATASETS statement limits the types of SAS les that
are available for processing. The procedure uses the order of precedence
described in rule 1 to further subset the types available for copying. The
following statements do not copy any members from the default data library
to the DEST data library; instead, the procedure issues an error message
because the MEMTYPE= value specied in the SELECT statement is not one
of the values of the MEMTYPE= option in the PROC DATASETS statement.
/* This step fails! */
proc datasets memtype=(data program);
copy out=dest;
select apples / memtype=catalog;
run;

3 If you specify an input data library in the IN= option other than the

procedure input library, the MEMTYPE= option in the PROC DATASETS

statement has no effect on the copy operation. Because no subsetting has yet
occurred, the procedure uses the order of precedence described in rule 1 to
subset the types available for copying. The following statements successfully
copy BODYFAT.DATA to the DEST data library because the SOURCE library
specied in the IN= option in the COPY statement is not effected by the
MEMTYPE= option in the PROC DATASETS statement.
proc datasets library=work memtype=catalog;
copy in=source out=dest;
select bodyfat / memtype=data;
run;

Copying Password-Protected SAS Files

You can copy a password-protected SAS le without specifying the password. In
addition, because the password continues to correspond to the SAS le, you must know
the password in order to access and manipulate the SAS le after you copy it.

Copying Data Sets with Long Variable Names

If the VALIDVARNAME=V6 system option is set and the data set has long variable
names, the long variable names are truncated, unique variables names are generated,
and the copy succeeds. The same is true for index names. If VALIDVARNAME=ANY or

The DATASETS Procedure

COPY Statement

333

MIXEDCASE, the copy fails with an error if the OUT= engine does not support long
variable names.
When a variable name is truncated, the variable name is shortened to eight bytes. If
this name has already been dened in the data set, the name is shortened and a digit is
added, starting with the number 2. The process of truncation and adding a digit
continues until the variable name is unique. For example, a variable named
LONGVARNAME becomes LONGVARN, provided that a variable with that name does
not already exist in the data set. In that case, the variable name becomes LONGVAR2.
CAUTION:

Truncated variable names can collide with names already dened in the input data set.
This is possible when the variable name that is already dened is exactly eight bytes
long and ends in a digit. In that case, the truncated name is dened in the output
data set and the name from the input data set is changed. For example,
options validvarname=mixedcase;
data test;
lonvar10=aLongVariableName;
retain longvar1-longvar5 0;
run;
options validvarname=v6;
proc copy in=work out=sasuser;
select test;
run;

In this example, LONGVAR10 is truncated to LONVAR1 and placed in the output

data set. Next, the original LONGVAR1 is copied. Its name is no longer unique and
so it is renamed LONGVAR2. The other variables in the input data set are also
renamed according to the renaming algorithm. 4

Using the COPY Procedure instead of the COPY Statement

Generally, the COPY procedure functions the same as the COPY statement in the
DATASETS procedure. The differences are
3 The IN= argument is required with PROC COPY. In the COPY statement, IN= is
optional. If omitted, the default value is the libref of the procedure input library.
3 PROC DATASETS cannot work with libraries that allow only sequential data
access.
3 The COPY statement honors the NOWARN option but PROC COPY does not.

Copying Generation Groups

You can use the COPY statement to copy an entire generation group. However, you
cannot copy a specic version in a generation group.

Transporting SAS Data Sets between Hosts

You use the COPY procedure, along with the XPORT engine, to transport SAS data
sets between hosts. See Moving and Accessing SAS Files for more information and an
example.

334

DELETE Statement

Chapter 15

DELETE Statement
Deletes SAS les from a SAS data library.
Featured in:

Example 1 on page 376

DELETE SAS-le(s)
</ <ALTER=alter-password>
<GENNUM=ALL|HIST|REVERT|integer>
<MEMTYPE=mtype>>;

Required Arguments
SAS-le(s)

species one or more SAS les that you want to delete.

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les that you want to delete.
You can use the option either in parentheses after the name of each SAS le or after
a forward slash.
See also: Using Passwords with the DATASETS Procedure on page 359
GENNUM=ALL|HIST|REVERT|integer

restricts processing for generation data sets. You can use the option either in
parentheses after the name of each SAS le or after a forward slash. Valid values are
ALL
refers to the base version and all historical versions in a generation group.
HIST
refers to all historical versions, but excludes the base version in a generation group.
REVERT|0
deletes the base version and changes the most current historical version, if it
exists, to the base version.
integer
is a number that references a specic version from a generation group. Specifying
a positive number is an absolute reference to a specic generation number that is
appended to a data sets name; that is, gennum=2 species MYDATA#002.
Specifying a negative number is a relative reference to a historical version in
relation to the base version, from the youngest to the oldest; that is, gennum=-1
refers to the youngest historical version.
See also: Restricting Processing for Generation Data Sets on page 362
See also: Understanding Generation Data Sets in SAS Language Reference:
Concepts
MEMTYPE=mtype

restricts processing to one member type. You can use the option either in
parentheses after the name of each SAS le or after a forward slash.

The DATASETS Procedure

DELETE Statement

335

MT=, MTYPE=
Default: DATA
Aliases:

Example 1 on page 376

Details
3 SAS immediately deletes SAS les when the RUN group executes. You do not
have an opportunity to verify the delete operation before it begins.
3 If you attempt to delete a SAS le that does not exist in the procedure input
library, PROC DATASETS issues a message and continues processing. If
NOWARN is used, no message is issued.
3 When you use the DELETE statement to delete a data set that has indexes
associated with it, the statement also deletes the indexes.

3 You cannot use the DELETE statement to delete a data le that has a foreign key
integrity constraint or a primary key with foreign key references. For data les
that have foreign keys, you must remove the foreign keys before you delete the
data le. For data les that have primary keys with foreign key references, you
must remove the foreign keys that reference the primary key before you delete the
data le.

Working with Generation Groups

When you are working with generation groups, you can use the DELETE statement to

3 delete the base version and all historical versions

3 delete the base version and rename the youngest historical version to the base
version

3 delete an absolute version

3 delete a relative version
3 delete all historical versions and leave the base version.
Deleting the Base Version and All Historical Versions
The following statements delete the base version and all historical versions where the
data set name is A:
proc datasets;
delete A(gennum=all);
proc datasets;
delete A / gennum=all;
proc datasets gennum=all;
delete A;

336

DELETE Statement

Chapter 15

The following statements delete the base version and all historical versions where
the data set name begins with the letter A:
proc datasets;
delete A:(gennum=all);
proc datasets;
delete A: / gennum=all;
proc datasets gennum=all;
delete A:;

Deleting the Base Version and Renaming the Youngest Historical Version to the Base
Version
The following statements delete the base version and rename the youngest historical
version to the base version, where the data set name is A:
proc datasets;
delete A(gennum=revert);
proc datasets;
delete A / gennum=revert;
proc datasets gennum=revert;
delete A;

The following statements delete the base version and rename the youngest historical
version to the base version, where the data set name begins with the letter A:
proc datasets;
delete A:(gennum=revert);
proc datasets;
delete A: / gennum=revert;
proc datasets gennum=revert;
delete A:;

Deleting a Version with an Absolute Number

The following statements use an absolute number to delete the rst historical version:
proc datasets;
delete A(gennum=1);
proc datasets;
delete A / gennum=1;
proc datasets gennum=1;
delete A;

The following statements delete a specic historical version, where the data set name
begins with the letter A:
proc datasets;
delete A:(gennum=1);
proc datasets;
delete A: / gennum=1;

The DATASETS Procedure

DELETE Statement

proc datasets gennum=1;

delete A:;

Deleting a Version with a Relative Number

The following statements use a relative number to delete the youngest historical
version, where the data set name is A:
proc datasets;
delete A(gennum=-1);
proc datasets;
delete A / gennum=-1;
proc datasets gennum=-1;
delete A;

The following statements use a relative number to delete the youngest historical
version, where the data set name begins with the letter A:
proc datasets;
delete A:(gennum=-1);
proc datasets;
delete A: / gennum=-1;
proc datasets gennum=-1;
delete A:;

Deleting All Historical Versions and Leaving the Base Version

The following statements delete all historical versions and leave the base version,
where the data set name is A:
proc datasets;
delete A(gennum=hist);
proc datasets;
delete A / gennum=hist;
proc datasets gennum=hist;
delete A;

The following statements delete all historical versions and leave the base version,
where the data set name begins with the letter A:
proc datasets;
delete A:(gennum=hist);
proc datasets;
delete A: / gennum=hist;
proc datasets gennum=hist;
delete A:;

337

338

EXCHANGE Statement

Chapter 15

EXCHANGE Statement
Exchanges the names of two SAS les in a SAS library.
Featured in:

Example 1 on page 376

EXCHANGE name-1=other-name-1
<name-n=other-name-n>
</ <ALTER=alter-password>
<MEMTYPE=mtype>>;

Required Arguments
name=other-name

exchanges the names of SAS les in the procedure input library. Both name and
other-name must already exist in the procedure input library.

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les whose names you want
to exchange. You can use the option either in parentheses after the name of each
SAS le or after a forward slash.
See also: Using Passwords with the DATASETS Procedure on page 359
MEMTYPE=mtype

restricts processing to one member type. You can only exchange the names of SAS
les of the same type. You can use the option either in parentheses after the name of
each SAS le or after a forward slash.
Default: If you do not specify MEMTYPE= in the PROC DATASETS statement, the
default is ALL.
See also: Restricting Member Types for Processing on page 360

Details
3 When you exchange more than one pair of names in one EXCHANGE statement,
PROC DATASETS performs the exchanges in the order that the names of the SAS
les occur in the directory listing, not in the order that you list the exchanges in
the EXCHANGE statement.
3 If the name SAS le does not exist in the SAS data library, PROC DATASETS
stops processing the RUN group that contains the EXCHANGE statement and
issues an error message. To override this behavior, specify the NOWARN option in
the PROC DATASETS statement.
3 The EXCHANGE statement also exchanges the associated indexes so that they
correspond with the new name.
3 The EXCHANGE statement only allows two existing generation groups to
exchange names. You cannot exchange a specic generation number with either an
existing base version or another generation number.

The DATASETS Procedure

FORMAT Statement

339

EXCLUDE Statement
Excludes SAS les from copying.
Restriction:

Must follow a COPY statement

Restriction:

Cannot appear in the same COPY step with a SELECT statement

Featured in: Example 1 on page 376

EXCLUDE SAS-le(s) </ MEMTYPE=mtype>;

Required Arguments
SAS-le(s)

species one or more SAS les to exclude from the copy operation. All SAS les you
name in the EXCLUDE statement must be in the library that is specied in the IN=
option in the COPY statement. If the SAS les are generation groups, the EXCLUDE
statement allows only selection of the base versions.

Options
MEMTYPE=mtype

restricts processing to one member type. You can use the option either in
parentheses after the name of each SAS le or after a forward slash.
Aliases:

MTYPE=, MT=

Default: If you do not specify MEMTYPE= in the PROC DATASETS statement, the

COPY statement, or in the EXCLUDE statement, the default is MEMTYPE=ALL.

See also: Restricting Member Types for Processing on page 360
See also: Specifying Member Types When Copying or Moving SAS Files on page

331

Excluding Many Like-Named Files

You can use shortcuts for listing many SAS les in the EXCLUDE statement. For
more information, see Shortcuts for Specifying Lists of Variable Names on page 24.

FORMAT Statement
Permanently assigns, changes, and removes variable formats in the SAS data set specied in the
MODIFY statement.
Restriction:

Must appear in a MODIFY RUN group

Featured in: Example 3 on page 381

340

IC CREATE Statement

Chapter 15

FORMAT variable-list-1 <format-1>

<variable-list-n <format-n>>;

Required Arguments
variable-list

species one or more variables whose format you want to assign, change, or remove.
If you want to disassociate a format with a variable, list the variable last in the list
with no format following. For example:
format x1-x3 4.1 time hhmm2.2 age;

Options
format

species a format to apply to the variable or variables listed before it. If you do not
specify a format, the FORMAT statement removes any format associated with the
variables in variable-list.
Note: You can use shortcut methods for specifying variables, such as the keywords
_NUMERIC, _CHARACTER_, and _ALL_. See Shortcuts for Specifying Lists of
Variable Names on page 24 for more information. 4

IC CREATE Statement
Creates an integrity constraint.
Must be in a MODIFY RUN group
See also: Understanding Integrity Constraints in SAS Language Reference: Concepts
Restriction:

IC CREATE <constraint-name=> constraint <MESSAGE=message-string

<MSGTYPE=USER>>;

Required Arguments
constraint

is the type of constraint. Valid values are as follows:

NOT NULL (variable)
species that variable does not contain a SAS missing value, including special
missing values.
UNIQUE (variables)
species that the values of variables must be unique. This constraint is identical
to DISTINCT.
DISTINCT (variables)
species that the values of variables must be unique. This constraint is identical
to UNIQUE.

The DATASETS Procedure

IC CREATE Statement

341

CHECK (WHERE-expression)
limits the data values of variables to a specic set, range, or list of values. This is
accomplished with a WHERE expression.
PRIMARY KEY (variables)
species a primary key, that is, a set of variables that do not contain missing
values and whose values are unique.
Interaction: A primary key affects the values of an individual data le until it

has a foreign key referencing it.

Requirement: When dening overlapping primary key and foreign key

constraints, which means that variables in a data le are part of both a primary
key and a foreign key denition, if you use exactly the same variables, then the
variables must be dened in a different order.
FOREIGN KEY (variables) REFERENCES table-name
<ON DELETE referential-action> <ON UPDATE referential-action>
species a foreign key, that is, a set of variables whose values are linked to the
values of the primary key variables in another data le. The referential actions
are enforced when updates are made to the values of a primary key variable that
is referenced by a foreign key.
There are three types of referential actions: RESTRICT, SET NULL, and
CASCADE:
For a RESTRICT referential action,
a delete operation
deletes the primary key row, but only if no foreign key values match the deleted
value.
an update operation
updates the primary key value, but only if no foreign key values match the
current value to be updated.
For a SET NULL referential action,
a delete operation
deletes the primary key row and sets the corresponding foreign key values to
NULL.
an update operation
modies the primary key value and sets all matching foreign key values to
NULL.
For a CASCADE referential action,
an update operation
modies the primary key value, and additionally modies any matching foreign
key values to the same value. CASCADE is not supported for delete operations.
Default: RESTRICT is the default action if no referential action is specied.
Interaction: Before it will enforce a SET NULL or CASCADE referential action,

SAS checks to see if there are other foreign keys that reference the primary key
and that specify RESTRICT for the intended operation. If RESTRICT is specied,
or if the constraint reverts to the default values, then RESTRICT is enforced for all
foreign keys, unless no foreign key values match the values to updated or deleted.
When dening overlapping primary key and foreign key constraints,
which means that variables in a data le are part of both a primary key and a
foreign key denition,

Requirement:

3 if you use exactly the same variables, then the variables must be dened in a
different order.

342

IC CREATE Statement

Chapter 15

3 the foreign keys update and delete referential actions must both be
RESTRICT.

Options

<constraint-name=>

is an optional name for the constraint. The name must be a valid SAS name. When
you do not supply a constraint name, a default name is generated. This default
constraint name has the following form
Default name

Constraint type

_NMxxxx_

Not Null

_UNxxxx_

Unique

_CKxxxx_

Check

_PKxxxx_

Primary key

_FKxxxx_

Foreign key

where xxxx is a counter beginning at 0001.

Note: The names PRIMARY, FOREIGN, MESSAGE, UNIQUE, DISTINCT,
CHECK, and NOT cannot be used as values for constraint-name. 4
<MESSAGE=message-string <MSGTYPE=USER>>

message-string is the text of an error message to be written to the log when the data
fails the constraint. For example,
ic create not null(socsec)
message=Invalid Social Security number;

Length:

The maximum length of the message is 250 characters.

<MSGTYPE=USER> controls the format of the integrity constraint error message.

By default when the MESSAGE= option is specied, the message you dene is
inserted into the SAS error message for the constraint, separated by a space.
MSGTYPE=USER suppresses the SAS portion of the message.
The following examples show how to create integrity constraints:
ic
ic
ic
ic
ic
ic

create a = not null(x);

create Unique_D = unique(d);
create Distinct_DE = distinct(d e);
create E_less_D = check(where=(e < d or d = 99));
create primkey = primary key(a b);
create forkey = foreign key (a b) references table-name
on update cascade on delete set null;
ic create not null (x);

Note that for a referential constraint to be established, the foreign key must specify the
same number of variables as the primary key, in the same order, and the variables
must be of the same type (character/numeric) and length.

The DATASETS Procedure

IC REACTIVATE Statement

343

IC DELETE Statement
Deletes an integrity constraint.
Restriction:

Must be in a MODIFY RUN group

IC DELETE constraint-name(s) | _ALL_;

Arguments
constraint-name(s)

names one or more constraints to delete. For example, to delete the constraints
Unique_D and Unique_E, use this statement:
ic delete Unique_D Unique_E;

_ALL_

deletes all constraints for the SAS data le specied in the preceding MODIFY
statement.

IC REACTIVATE Statement
Reactivates a foreign key integrity constraint that is inactive.
Restriction:

Must be in a MODIFY RUN group

IC REACTIVATE foreign-key-name REFERENCES libref;

Arguments
foreign-key-name

is the name of the foreign key to reactivate.

libref

refers to the SAS library containing the data set that contains the primary key that
is referenced by the foreign key.

344

INDEX CENTILES

Chapter 15

For example, suppose that you have the foreign key FKEY dened in data set
MYLIB.MYOWN and that FKEY is linked to a primary key in data set
MAINLIB.MAIN. If the integrity constraint is inactivated by a copy or move operation,
you can reactivate the integrity constraint by using the following code:
proc datasets library=mylib;
modify myown;
ic reactivate fkey references mainlib;
run;

INDEX CENTILES
Updates centiles statistics for indexed variables.
Restriction:

Must be in a MODIFY RUN group

INDEX CENTILES index(s)

</ <REFRESH>
<UPDATECENTILES= ALWAYS|NEVER|integer>>;

Required Arguments
index(s)

names one or more indexes.

Options
REFRESH

updates centiles immediately, regardless of the value of UPDATECENTILES.

UPDATECENTILES=ALWAYS|NEVER|integer

species when centiles are to be updated. It is not practical to update centiles after
every data set update. Therefore, you can specify as the value of
UPDATECENTILES the percent of the data values that can be changed before
centiles for the indexed variables are updated.
Valid values for UPDATECENTILES are
ALWAYS|0
updates centiles when the data set is closed if any changes have been made to the
data set index.
NEVER|101
does not update centiles.
integer
is the percent of values for the indexed variable that can be updated before
centiles are refreshed.
Alias: UPDCEN
Default 5 (percent)

The DATASETS Procedure

INDEX CREATE Statement

345

INDEX CREATE Statement

Creates simple or composite indexes for the SAS data set specied in the MODIFY statement.
Restriction:

Must be in a MODIFY RUN group

INDEX CREATE index-specication(s)

</ < NOMISS>
<UNIQUE>
<UPDATECENTILES= ALWAYS|NEVER|integer>>;

Required Arguments
index-specication(s)

can be one or both of the following forms:

variable
creates a simple index on the specied variable.
index=(variables)
creates a composite index. The name you specify for index is the name of the
composite index. It must be a valid SAS name and cannot be the same as any
variable name or any other composite index name. You must specify at least two
variables.
Note: The index name must follow the same rules as a SAS variable name,
including avoiding the use of reserved names for automatic variables, such as _N_,
and special variable list names, such as _ALL_. For more information, refer to Rules
for Words and Names in the SAS Language in SAS Language Reference: Concepts. 4

Options
NOMISS

excludes from the index all observations with missing values for all index variables.
When you create an index with the NOMISS option, SAS uses the index only for
WHERE processing and only when missing values fail to satisfy the WHERE
expression. For example, if you use the following WHERE statement, SAS does not
use the index, because missing values satisfy the WHERE expression:
where dept ne 01;

Refer to SAS Language Reference: Concepts.

Note: BY-group processing ignores indexes that are created with the NOMISS
option. 4
Featured in: Example 3 on page 381
UNIQUE

species that the combination of values of the index variables must be unique. If you
specify UNIQUE and multiple observations have the same values for the index
variables, the index is not created.

346

INDEX DELETE Statement

Chapter 15

Featured in:

Example 3 on page 381

UPDATECENTILES=ALWAYS|NEVER|integer

species when centiles are to be updated. It is not practical to update centiles after
every data set update. Therefore, you can specify the percent of the data values that
can be changed before centiles for the indexed variables are updated. Valid values for
UPDATECENTILES are as follows:
ALWAYS|0
updates centiles when the data set is closed if any changes have been made to the
data set index.
NEVER|101
does not update centiles.
integer
species the percent of values for the indexed variable that can be updated before
centiles are refreshed.
Alias: UPDCEN
Default: 5% (percent)

INDEX DELETE Statement

Deletes one or more indexes associated with the SAS data set specied in the MODIFY statement.
Restriction:

Must appear in a MODIFY RUN group

INDEX DELETE index(s) | _ALL_;

Required Arguments
index(s)

names one or more indexes to delete. The index(es) must be for variables in the SAS
data set that is named in the preceding MODIFY statement. You can delete both
simple and composite indexes.
_ALL_

deletes all indexes, except for indexes that are owned by an integrity constraint.
When an index is created, it is marked as owned by the user, by an integrity
constraint, or by both. If an index is owned by both a user and an integrity
constraint, the index is not deleted until both an IC DELETE statement and an
INDEX DELETE statement are processed.
Note: You can use the CONTENTS statement to produce a list of all indexes for a
data set. 4

The DATASETS Procedure

LABEL Statement

347

INFORMAT Statement
Permanently assigns, changes, and removes variable informats in the data set specied in the
MODIFY statement.
Restriction:

Must appear in a MODIFY RUN group

Featured in: Example 3 on page 381

INFORMAT variable-list-1 <informat-1>

<variable-list-n <informat-n>>;

Required Arguments

variable-list

species one or more variables whose informats you want to assign, change, or
remove. If you want to disassociate an informat with a variable, list the variable last
in the list with no informat following. For example:
informat a b 2. x1-x3 4.1 c;

Options

informat

species an informat for the variables immediately preceding it in the statement. If

you do not specify an informat, the INFORMAT statement removes any existing
informats for the variables in variable-list.
Note: You can use shortcut methods for specifying variables, such as the keywords
_NUMERIC, _CHARACTER_, and _ALL_. See Shortcuts for Specifying Lists of
Variable Names on page 24 for more information. 4

LABEL Statement
Assigns, changes, and removes variable labels for the SAS data set specied in the MODIFY
statement.
Restriction:

Must appear in a MODIFY RUN group

Featured in: Example 3 on page 381

LABEL variable-1=< label-1| >

<variable-n=< label-n| >>;

348

MODIFY Statement

Chapter 15

Required Arguments
variable=<label>

assigns a label to a variable. If a single quotation mark appears in the label, write it
as two single quotation marks in the LABEL statement. Specifying variable= or
variable= removes the current label.
Range:

1-256 characters

MODIFY Statement
Changes the attributes of a SAS le and, through the use of subordinate statements, the attributes
of variables in the SAS le.
Featured in:

Example 3 on page 381

MODIFY SAS-le <(option(s))>

</ <CORRECTENCODING=encoding-value>
<DTC=SAS-date-time>
<GENNUM=integer>
<MEMTYPE=mtype>>;

To do this

Use this option

Restrict processing to a certain type of SAS le

MEMTYPE=

Specify attributes
Change the character-set encoding

CORRECTENCODING=

Specify a creation date and time

DTC=

Assign or change a data set label

LABEL=

Specify how the data are currently sorted

SORTEDBY=

Assign or change a special data set type

TYPE=

Modify passwords
Modify an alter password

ALTER=

Modify a read, write, or alter password

PW=

Modify a read password

READ=

Modify a write password

WRITE=

Modify generation groups

Modify the maximum number of versions for a
generation group

GENMAX=

Modify a historical version

GENNUM=

The DATASETS Procedure

MODIFY Statement

349

Required Arguments
SAS-le

species a SAS le that exists in the procedure input library.

Options
ALTER=password-modication

assigns, changes, or removes an alter password for the SAS le named in the
MODIFY statement. password-modication is one of the following:

3
3
3
3
3

new-password
old-password / new-password
/ new-password
old-password /
/

and time the le was actually created.

Restriction: DTC= cannot be used with encrypted les or sequential les.
Restriction: DTC= can be used only when the resulting SAS le uses the V8 or V9

engine.
GENMAX=number-of-generations

species the maximum number of versions. You can use this option either in
parentheses after the name of each SAS le or after a forward slash.
Range: 0 to 1,000
Default: 0
GENNUM=integer

restricts processing for generation data sets. You can specify GENNUM= either in
parentheses after the name of each SAS le or after a forward slash. Valid value is
integer, which is a number that references a specic version from a generation group.
Specifying a positive number is an absolute reference to a specic generation number
that is appended to a data sets name; that is, gennum=2 species MYDATA#002.

350

MODIFY Statement

Chapter 15

Specifying a negative number is a relative reference to a historical version in relation

to the base version, from the youngest to the oldest; that is, gennum=-1 refers to the
youngest historical version. Specifying 0, which is the default, refers to the base
version.
See also: Understanding Generation Data Sets in SAS Language Reference:
Concepts
LABEL=data-set-label |

assigns, changes, or removes a data set label for the SAS data set named in the
MODIFY statement. If a single quotation mark appears in the label, write it as two
single quotation marks. LABEL= or LABEL= removes the current label.
Range: 1-40 characters
Featured in: Example 3 on page 381
MEMTYPE=mtype

restricts processing to one member type. You cannot specify MEMTYPE= in

parentheses after the name of each SAS le; you must specify MEMTYPE= after a
forward slash.
Aliases: MTYPE= and MT=
Default: If you do not specify the MEMTYPE= option in the PROC DATASETS
statement or in the MODIFY statement, the default is MEMTYPE=DATA.
PW=password-modication

assigns, changes, or removes a read, write, or alter password for the SAS le named
in the MODIFY statement. password-modication is one of the following:
3 new-password
3 old-password / new-password

3 / new-password
3 old-password /
3 /
See also: Manipulating Passwords on page 351
READ=password-modication

assigns, changes, or removes a read password for the SAS le named in the MODIFY
statement. password-modication is one of the following:

3
3
3
3
3

new-password
old-password / new-password
/ new-password
old-password /
/

Example 3 on page 381

SORTEDBY=sort-information

species how the data are currently sorted. SAS stores the sort information with the
le but does not verify that the data are sorted the way you indicate.
sort-information can be one of the following:
by-clause </ collate-name>
indicates how the data are currently sorted. Values for by-clause are the variables
and options you can use in a BY statement in a PROC SORT step. collate-name
names the collating sequence used for the sort. By default, the collating sequence
is that of your host operating environment.

The DATASETS Procedure

MODIFY Statement

351

_NULL_
removes any existing sort information.
Restriction: The data must be sorted in the order that you specify. If the data is

not in the specied order, SAS will not sort it for you.
Featured in:

Example 3 on page 381

TYPE=special-type

assigns or changes the special data set type of a SAS data set. SAS does not verify

3 the SAS data set type you specify in the TYPE= option (except to check if it has
a length of eight or fewer characters).

3 that the SAS data sets structure is appropriate for the type you have
designated.
Note: Do not confuse the TYPE= option with the MEMTYPE= option. The
TYPE= option species a type of special SAS data set. The MEMTYPE= option
species one or more types of SAS les in a SAS data library. 4
Most SAS data sets have no special type. However, certain SAS procedures,
like the CORR procedure, can create a number of special SAS data sets. In
addition, SAS/STAT software and SAS/EIS software support special data set types.

Tip:

WRITE=password-modication

assigns, changes, or removes a write password for the SAS le named in the
MODIFY statement. password-modication is one of the following:

3
3
3
3
3

new-password
old-password / new-password
/ new-password
old-password /
/

/* assigns an alter password to an already read-protected SAS data set */

modify colors (read=green alter=red);

Changing Passwords
/* changes the write password from YELLOW to BROWN */
modify cars (write=yellow/brown);

/* uses alter access to change unknown read password to BLUE */

modify colors (read=/blue alter=red);

352

RENAME Statement

Chapter 15

Removing Passwords
/* removes the alter password RED from STATES */
modify states (alter=red/);

/* uses alter access to remove the read password */

modify zoology (read=green/ alter=red);

/* uses PW= as an alias for either WRITE= or ALTER= to remove unknown

read password */
modify biology (read=/ pw=red);

Working with Generation Groups

Changing the Number of Generations

/* changes the number of generations on data set A to 99 */
modify A (genmax=99);

Removing Passwords
/* removes the alter password RED from STATES#002 */
modify states (alter=red/) / gennum=2;

RENAME Statement
Renames variables in the SAS data set specied in the MODIFY statement.
Restriction:

Must appear in a MODIFY RUN group

Featured in:

Example 3 on page 381

RENAME old-name-1=new-name-1
<old-name-n=new-name-n>;

Required Arguments

old-name=new-name

changes the name of a variable in the data set specied in the MODIFY statement.
old-name must be a variable that already exists in the data set. new-name cannot be
the name of a variable that already exists in the data set or the name of an index,
and the new name must be a valid SAS name. See Rules for SAS Variable Names
in SAS Language Reference: Concepts.

The DATASETS Procedure

REPAIR Statement

353

Details
3 If old-name does not exist in the SAS data set or new-name already exists, PROC
DATASETS stops processing the RUN group containing the RENAME statement
and issues an error message.

3 When you use the RENAME statement to change the name of a variable for which
there is a simple index, the statement also renames the index.

3 If the variable that you are renaming is used in a composite index, the composite
index automatically references the new variable name. However, if you attempt to
rename a variable to a name that has already been used for a composite index, you
receive an error message.

REPAIR Statement
Attempts to restore damaged SAS data sets or catalogs to a usable condition.
REPAIR SAS-le(s)
</ < ALTER=alter-password>
<GENNUM=integer>
<MEMTYPE=mtype>>;

Required Arguments
SAS-le(s)

species one or more SAS data sets or catalogs in the procedure input library.

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les that are named in the
REPAIR statement. You can use the option either in parentheses after the name of
each SAS le or after a forward slash.
See also: Using Passwords with the DATASETS Procedure on page 359
GENNUM=integer

restricts processing for generation data sets. You can use the option either in
parentheses after the name of each SAS le or after a forward slash. Valid value is
integer, which is a number that references a specic version from a generation group.
Specifying a positive number is an absolute reference to a specic generation number
that is appended to a data sets name; that is, gennum=2 species MYDATA#002.
Specifying a negative number is a relative reference to a historical version in relation
to the base version, from the youngest to the oldest; that is, gennum=-1 refers to the
youngest historical version. Specifying 0, which is the default, refers to the base
version.
See also: Restricting Processing for Generation Data Sets on page 362
See also: Understanding Generation Data Sets in SAS Language Reference:

Concepts

354

REPAIR Statement

Chapter 15

MEMTYPE=mtype

restricts processing to one member type.

Aliases: MT=, MTYPE=
Default: If you do not specify the MEMTYPE= option in the PROC DATASETS
statement or in the REPAIR statement, the default is MEMTYPE=ALL.
See also: Restricting Member Types for Processing on page 360

Details
The most common situations that require the REPAIR statement are as follows:
3 A system failure occurs while you are updating a SAS data set or catalog.
3 The device on which a SAS data set or an associated index resides is damaged. In
this case, you can restore the damaged data set or index from a backup device, but
the data set and index no longer match.
3 The disk that stores the SAS data set or catalog becomes full before the le is
completely written to disk. You may need to free some disk space. PROC
DATASETS requires free space when repairing SAS data sets with indexes and
when repairing SAS catalogs.
3 An I/O error occurs while you are writing a SAS data set or catalog entry.
When you use the REPAIR statement for SAS data sets, it recreates all indexes for
the data set. It also attempts to restore the data set to a usable condition, but the
restored data set may not include the last several updates that occurred before the
system failed. You cannot use the REPAIR statement to recreate indexes that were
destroyed by using the FORCE option in a PROC SORT step.
When you use the REPAIR statement for a catalog, you receive a message stating
whether the REPAIR statement restored the entry. If the entire catalog is potentially
damaged, the REPAIR statement attempts to restore all the entries in the catalog. If
only a single entry is potentially damaged, for example when a single entry is being
updated and a disk-full condition occurs, on most systems only the entry that is open
when the problem occurs is potentially damaged. In this case, the REPAIR statement
attempts to repair only that entry. Some entries within the restored catalog may not
include the last updates that occurred before a system crash or an I/O error. The
REPAIR statement issues warning messages for entries that may have truncated data.
To repair a damaged catalog, the version of SAS that you use must be able to update
the catalog. Whether a SAS version can update a catalog (or just read it) is determined
by the SAS version that created the catalog:
3 A damaged Version 6 catalog can be repaired with Version 6 only.
3 A damaged Version 8 catalog can be repaired with either Version 8 or SAS System
9, but not with Version 6.
3 A damaged SAS System 9 catalog can be repaired with SAS System 9 only.
If the REPAIR operation is not successful, try to restore the SAS data set or catalog
from your systems backup les.
If you issue a REPAIR statement for a SAS le that does not exist in the specied
library, PROC DATASETS stops processing the run group that contains the REPAIR
statement, and issues an error message. To override this behavior and continue
processing, use the NOWARN option in the PROC DATASETS statement.
If you are using Cross-Environment Data Access (CEDA) to process a damaged
foreign SAS data set, CEDA cannot repair it. CEDA does not support update processing,
which is required in order to repair a damaged data set. To repair the foreign le, you
must move it back to its native environment. Note that observations may be lost during
the repair process. For more information about CEDA, refer to Processing Data Using
Cross-Environment Data Access in SAS Language Reference: Concepts.

The DATASETS Procedure

SAVE Statement

355

SAVE Statement
Deletes all the SAS les in a library except the ones listed in the SAVE statement.
Featured in: Example 2 on page 380

SAVE SAS-le(s) </ MEMTYPE=mtype>;

Required Arguments
SAS-le(s)

species one or more SAS les that you do not want to delete from the SAS data
library.

Options
MEMTYPE=mtype

restricts processing to one member type. You can use the option either in
parentheses after the name of each SAS le or after a forward slash.
Aliases:

MTYPE= and MT=

Default: If you do not specify the MEMTYPE= option in the PROC DATASETS

statement or in the SAVE statement, the default is MEMTYPE=ALL.

See also: Restricting Member Types for Processing on page 360
Featured in:

Example 2 on page 380

Details
3 If one of the SAS les in SAS-le does not exist in the procedure input library,
PROC DATASETS stops processing the RUN group containing the SAVE
statement and issues an error message. To override this behavior, specify the
NOWARN option in the PROC DATASETS statement.

3 When the SAVE statement deletes SAS data sets, it also deletes any indexes
associated with those data sets.
CAUTION:

SAS immediately deletes libraries and library members when you submit a RUN
group. You are not asked to verify the delete operation before it begins. Because
the SAVE statement deletes many SAS les in one operation, be sure that you
understand how the MEMTYPE= option affects which types of SAS les are
saved and which types are deleted.

3 When you use the SAVE statement with generation groups, the SAVE statement
treats the base version and all historical versions as a unit. You cannot save a
specic version.

356

SELECT Statement

Chapter 15

SELECT Statement
Selects SAS les for copying.
Restriction:

Must follow a COPY statement

Restriction:

Cannot appear with an EXCLUDE statement in the same COPY step

Featured in:

Example 1 on page 376

SELECT SAS-le(s)
</ <ALTER=alter-password>
<MEMTYPE= mtype>>;

Required Arguments
SAS-le(s)

species one or more SAS les that you want to copy. All of the SAS les that you
name must be in the data library that is referenced by the libref named in the IN=
option in the COPY statement. If the SAS les have generation groups, the SELECT
statement allows only selection of the base versions.

Options
ALTER=alter-password

provides the alter password for any alter-protected SAS les that you are moving
from one data library to another. Because you are moving and thus deleting a SAS
le from a SAS data library, you need alter access. You can use the option either in
parentheses after the name of each SAS le or after a forward slash.
See also: Using Passwords with the DATASETS Procedure on page 359
MEMTYPE=mtype

restricts processing to one member type. You can use the option either in
parentheses after the name of each SAS le or after a forward slash.
Aliases:

MTYPE= and MT=

Default: If you do not specify the MEMTYPE= option in the PROC DATASETS

statement, in the COPY statement, or in the SELECT statement, the default is

MEMTYPE=ALL.
See also: Specifying Member Types When Copying or Moving SAS Files on page

331
See also: Restricting Member Types for Processing on page 360
Featured in:

Example 1 on page 376

Selecting Many Like-Named Files

You can use shortcuts for listing many SAS les in the SELECT statement. For more
information, see Shortcuts for Specifying Lists of Variable Names on page 24.

The DATASETS Procedure

Procedure Execution

357

Concepts: DATASETS Procedure

Procedure Execution
Execution of Statements
When you start the DATASETS procedure, you specify the procedure input library in
the PROC DATASETS statement. If you omit a procedure input library, the procedure
processes the current default SAS data library (usually the WORK data library). To
specify a new procedure input library, issue the DATASETS procedure again.
Statements execute in the order they are written. For example, if you want to see
the contents of a data set, copy a data set, and then visually compare the contents of
the second data set with the rst, the statements that perform those tasks must appear
in that order (that is, CONTENTS, COPY, CONTENTS).

RUN-Group Processing
PROC DATASETS supports RUN-group processing. RUN-group processing enables
you to submit RUN groups without ending the procedure.
The DATASETS procedure supports four types of RUN groups. Each RUN group is
dened by the statements that compose it and by what causes it to execute.
Some statements in PROC DATASETS act as implied RUN statements because they
cause the RUN group preceding them to execute.
The following list discusses what statements compose a RUN group and what causes
each RUN group to execute:

3 The PROC DATASETS statement always executes immediately. No other

statement is necessary to cause the PROC DATASETS statement to execute.
Therefore, the PROC DATASETS statement alone is a RUN group.

3 The MODIFY statement, and any of its subordinate statements, form a RUN
group. These RUN groups always execute immediately. No other statement is
necessary to cause a MODIFY RUN group to execute.

3 The APPEND, CONTENTS, and COPY statements (including EXCLUDE and

SELECT, if present), form their own separate RUN groups. Every APPEND
statement forms a single-statement RUN group; every CONTENTS statement
forms a single-statement RUN group; and every COPY step forms a RUN group.
Any other statement in the procedure, except those that are subordinate to either
the COPY or MODIFY statement, causes the RUN group to execute.

3 One or more of the following statements form a RUN group:

3 AGE
3 CHANGE
3 DELETE
3 EXCHANGE
3 REPAIR
3 SAVE
If any of these statements appear in sequence in the PROC step, the sequence
forms a RUN group. For example, if a REPAIR statement appears immediately
after a SAVE statement, the REPAIR statement does not force the SAVE

358

Procedure Execution

Chapter 15

statement to execute; it becomes part of the same RUN group. To execute the
RUN group, submit one of the following statements:

3
3
3
3
3
3
3
3

PROC DATASETS
APPEND
CONTENTS
COPY
MODIFY
QUIT
RUN
another DATA or PROC step.

SAS reads the program statements that are associated with one task until it reaches
a RUN statement or an implied RUN statement. It executes all of the preceding
statements immediately, then continues reading until it reaches another RUN
statement or implied RUN statement. To execute the last task, you must use a RUN
statement or a statement that stops the procedure.
The following PROC DATASETS step contains ve RUN groups:
libname dest SAS-data-library;
/* RUN group */
proc datasets;
/* RUN group */
change nutr=fatg;
delete bldtest;
exchange xray=chest;
/* RUN group */
copy out=dest;
select report;
/* RUN group */
modify bp;
label dias=Taken at Noon;
rename weight=bodyfat;
/* RUN group */
append base=tissue data=newtiss;
quit;

Note: If you are running in interactive line mode, you can receive messages that
statements have already executed before you submit a RUN statement. Plan your tasks
carefully if you are using this environment for running PROC DATASETS. 4

Error Handling
Generally, if an error occurs in a statement, the RUN group containing the error does
not execute. RUN groups preceding or following the one containing the error execute
normally. The MODIFY RUN group is an exception. If a syntax error occurs in a
statement subordinate to the MODIFY statement, only the statement containing the
error fails. The other statements in the RUN group execute.
Note that if the rst word of the statement (the statement name) is in error and the
procedure cannot recognize it, the procedure treats the statement as part of the
preceding RUN group.

The DATASETS Procedure

Using Passwords with the DATASETS Procedure

359

Password Errors
If there is an error involving an incorrect or omitted password in a statement, the
error affects only the statement containing the error. The other statements in the RUN
group execute.

Forcing a RUN Group with Errors to Execute

The FORCE option in the PROC DATASETS statement forces execution of the RUN
group even if one or more of the statements contain errors. Only the statements that
are error-free execute.

Ending the Procedure

To stop the DATASETS procedure, you must issue a QUIT statement, a RUN
CANCEL statement, a new PROC statement, or a DATA statement. Submitting a
QUIT statement executes any statements that have not executed. Submitting a RUN
CANCEL statement cancels any statements that have not executed.

Using Passwords with the DATASETS Procedure

Several statements in the DATASETS procedure support options that manipulate
passwords on SAS les. These options, ALTER=, PW=, READ=, and WRITE=, are also
data set options.* If you do not know how passwords affect SAS les, refer to SAS
Language Reference: Concepts.
When you are working with password-protected SAS les in the AGE, CHANGE,
DELETE, EXCHANGE, REPAIR, or SELECT statement, you can specify password
options in the PROC DATASETS statement or in the subordinate statement.
Note: The ALTER= option works slightly different for the COPY (when moving a
le) and MODIFY statements. Refer to COPY Statement on page 327 and MODIFY
Statement on page 348. 4
SAS searches for passwords in the following order:
1 in parentheses after the name of the SAS le in a subordinate statement. When

used in parentheses, the option only refers to the name immediately preceding the
option. If you are working with more than one SAS le in a data library and each
SAS le has a different password, you must specify password options in
parentheses after individual names.
In the following statement, the ALTER= option provides the password RED for
the SAS le BONES only:
delete xplant bones(alter=red);

2 after a forward slash (/) in a subordinate statement. When you use a password

option following a slash, the option refers to all SAS les named in the statement
unless the same option appears in parentheses after the name of a SAS le. This
method is convenient when you are working with more than one SAS le and they
all have the same password.
* In the APPEND and CONTENTS statements, you use these options just as you use any SAS data set option, in parentheses
after the SAS data set name.

360

Restricting Member Types for Processing

Chapter 15

In the following statement, the ALTER= option in parentheses provides the

password RED for the SAS le CHEST, and the ALTER= option after the slash
provides the password BLUE for the SAS le VIRUS:
delete chest(alter=red) virus / alter=blue;

3 in the PROC DATASETS statement. Specifying the password in the PROC

DATASETS statement can be useful if all the SAS les you are working with in
the library have the same password. Do not specify the option in parentheses.
In the following PROC DATASETS step, the PW= option provides the password
RED for the SAS les INSULIN and ABNEG:
proc datasets pw=red;
delete insulin;
contents data=abneg;
run;

Note: For the password for a SAS le in a SELECT statement, SAS looks in
the COPY statement before it looks in the PROC DATASETS statement. 4

Restricting Member Types for Processing

In the PROC DATASETS Statement

If you name a member type or several member types in the PROC DATASETS
statement, in most subsequent statements (except the CONTENTS and COPY
statements), you can name only a subset of the list of member types included in the
PROC DATASETS statement. The directory listing that the PROC DATASETS
statement writes to the SAS log includes only those SAS les of the type specied in the
MEMTYPE= option.

In Subordinate Statements
Use the MEMTYPE= option in the following subordinate statements to limit the
member types that are available for processing:
AGE
CHANGE
DELETE
EXCHANGE
EXCLUDE
REPAIR
SAVE
SELECT
Note: The MEMTYPE= option works slightly differently for the CONTENTS, COPY,
and MODIFY statements. Refer to CONTENTS Statement on page 323, COPY
Statement on page 327, and MODIFY Statement on page 348 for more information. 4

The DATASETS Procedure

Restricting Member Types for Processing

361

The procedure searches for MEMTYPE= in the following order:

1 in parentheses immediately after the name of a SAS le. When used in
parentheses, the MEMTYPE= option refers only to the SAS le immediately
preceding the option. For example, the following statement deletes HOUSE.DATA,
LOT.CATALOG, and SALES.DATA because the default member type for the
DELETE statement is DATA. (Refer to Table 15.5 on page 362 for the default
types for each statement.)
delete house lot(memtype=catalog) sales;

2 after a slash (/) at the end of the statement. When used following a slash, the

MEMTYPE= option refers to all SAS les named in the statement unless the option
appears in parentheses after the name of a SAS le. For example, the following
statement deletes LOTPIX.CATALOG, REGIONS.DATA, and APPL.CATALOG:
delete lotpix regions(memtype=data) appl / memtype=catalog;

3 in the PROC DATASETS statement. For example, this DATASETS procedure

deletes APPL.CATALOG:
proc datasets memtype=catalog;
delete appl;
run;

Note: When you use the EXCLUDE and SELECT statements, the procedure
looks in the COPY statement for the MEMTYPE= option before it looks in the
PROC DATASETS statement. For more information, see Specifying Member
Types When Copying or Moving SAS Files on page 331. 4
4 for the default value. If you do not specify a MEMTYPE= option in the subordinate

statement or in the PROC DATASETS statement, the default value for the
subordinate statement determines the member type available for processing.

Member Types
The following list gives the possible values for the MEMTYPE= option:
ACCESS
access descriptor les (created by SAS/ACCESS software)
ALL
all member types
CATALOG
SAS catalogs
DATA
SAS data les
FDB
nancial database
MDDB
multidimensional database
PROGRAM
stored compiled SAS programs
VIEW
SAS views

362

Restricting Processing for Generation Data Sets

Chapter 15

Table 15.5 on page 362 shows the member types that you can use in each statement:
Table 15.5

Subordinate Statements and Appropriate Member Types

Statement

Appropriate member types

Default
member type

AGE

ACCESS, CATALOG, DATA, FDB, MDDB,

PROGRAM, VIEW

DATA

CHANGE

ACCESS, ALL, CATALOG, DATA, FDB, MDDB,

PROGRAM, VIEW

ALL

CONTENTS

ALL, DATA, VIEW

DATA1

COPY

ACCESS, ALL, CATALOG, DATA, FDB, MDDB,

PROGRAM, VIEW

ALL

DELETE

ACCESS, ALL, CATALOG, DATA, FDB, MDDB,

PROGRAM, VIEW

DATA

Directory Listing as SAS Output

The CONTENTS statement lists the directory of the procedure input library if you
use the DIRECTORY option or specify DATA=_ALL_.
If you want only a directory, use the NODS option and the _ALL_ keyword in the
DATA= option. The NODS option suppresses the description of the SAS data sets; only
the directory appears in the output.
Note: The CONTENTS statement does not put a directory in an output data set. If
you try to create an output data set using the NODS option, you receive an empty
output data set. Use the SQL procedure to create a SAS data set that contains
information about a SAS data library. 4
Note: If you specify the ODS RTF destination, the PROC DATASETS output will go
to both the SAS log and the ODS output area. The NOLIST option will suppress output
to both. To see the output only in the SAS log, use the ODS EXCLUDE statement by
specifying the member directory as the exclusion. 4

Procedure Output

The CONTENTS Statement

The only statement in PROC DATASETS that produces procedure output is the
CONTENTS statement. This section shows the output from the CONTENTS statement
for the GROUP data set, which is shown in Output 15.3.
Only the items in the output that require explanation are discussed.

Data Set Attributes

Here are descriptions of selected elds shown in Output 15.3:
Member Type

is the type of library member (DATA or VIEW).

Protection

indicates whether the SAS data set is READ, WRITE, or ALTER password
protected.
Data Set Type

names the special data set type (such as CORR, COV, SSPC, EST, or FACTOR), if
any.
Observations

is the total number of observations currently in the le. Note that for a very large
data set, if the number of observations exceeds the number that can be stored in a
double-precision integer, the count will show as missing.
Deleted Observations

is the number of observations marked for deletion. These observations are not
included in the total number of observations, shown in the Observations eld.
Note that for a very large data set, if the number of deleted observations exceeds
the number that can be stored in a double-precision integer, the count will show as
missing.

The DATASETS Procedure

Procedure Output

365

Compressed

indicates whether the data set is compressed. If the data set is compressed, the
output includes an additional item, Reuse Space (with a value of YES or NO),
that indicates whether to reuse space that is made available when observations
are deleted.
Sorted

indicates whether the data set is sorted. If you sort the data set with PROC SORT,
PROC SQL, or specify sort information with the SORTEDBY= data set option, a
value of YES appears here, and there is an additional section to the output. See
Sort Information on page 367 for details.
Data Representation

is the format in which data is represented on a computer architecture or in an

operating environment. For example, on an IBM PC, character data is represented
by its ASCII encoding and byte-swapped integers. Native data representation
refers to an environment for which the data representation compares with the
CPU that is accessing the le. For example, a le that is in Windows data
representation is native to the Windows operating environment.
Encoding

is the encoding value. Encoding is a set of characters (letters, logograms, digits,

punctuation, symbols, control characters, and so on) that have been mapped to
numeric values (called code points) that can be used by computers. The code
points are assigned to the characters in the character set when you apply an
encoding method.

Output 15.3

Data Set Attributes for the GROUP Data Set

The SAS System

The DATASETS Procedure

Data Set Name
Member Type
Engine
Created
Last Modified
Protection
Data Set Type
Label
Data Representation
Encoding

HEALTH.GROUP
DATA
V9
Wednesday, February
05, 2003 02:20:56
Wednesday, February
05, 2003 02:20:56
READ

Observations
Variables
Indexes
Observation Length

148
11
1
96

Deleted Observations

Compressed
Sorted

NO
YES

Test Subjects
WINDOWS_32
wlatin1 Western (Windows)

Engine and Operating Environment-Dependent Information

The CONTENTS statement produces operating environment-specic and
engine-specic information. This information differs depending on the operating
environment. The following output is from the Windows operating environment.

366

Procedure Output

Chapter 15

Output 15.4

Engine and Operating Environment Dependent Information Section of CONTENTS Output

Engine/Host Dependent Information

Data Set Page Size

Number of Data Set Pages
First Data Page
Max Obs per Page
Obs in First Data Page
Index File Page Size
Number of Index File Pages
Number of Data Set Repairs
File Name
Release Created
Host Created

8192
4
1
84
62
4096
2
0
c:\Myfiles\health\group.sas7bdat
9.0101B0
XP_PRO

Alphabetic List of Variables and Attributes

Here are descriptions of selected columns in Output 15.5:
#

is the logical position of each variable in the observation. This is the number that
is assigned to the variable when it is dened.
Variable

is the name of each variable. By default, variables appear alphabetically.

Note: Variable names are sorted such that X1, X2, and X10 appear in that
order and not in the true collating sequence of X1, X10, and X2. Variable names
that contain an underscore and digits may appear in a nonstandard sort order.
For example, P25 and P75 appear before P2_5. 4
Type

species the type of variable: character or numeric.

Len

species the variables length, which is the number of bytes used to store each of a
variables values in a SAS data set.
Transcode

species whether a character variable is transcoded. If the attribute is NO, then

transcoding is suppressed. By default, character variables are transcoded when
required. For information on transcoding, see SAS National Language Support
(NLS): Users Guide.
Note: If none of the variables in the SAS data set has a format, informat, or label
associated with it, or if none of the variables are set to no transcoding, then the column
for that attribute does not display. 4

The DATASETS Procedure

Output 15.5

Procedure Output

367

Variable Attributes Section

Alphabetic List of Variables and Attributes

#
9
4
3
10
11
1
7
2
8
6
5

Variable

Type

Len

BIRTH
CITY
FNAME
HIRED
HPHONE
IDNUM
JOBCODE
LNAME
SALARY
SEX
STATE

Num
Char
Char
Num
Char
Char
Char
Char
Num
Char
Char

8
15
15
8
12
4
3
15
8
1
2

Format

Informat

DATE7.
$.
$.
DATE7.
$.
$.
$.
$.
COMMA8.
$.
$.

Label

Transcode

DATE7.
$.
$.
DATE7.
$.
$.
$.
$.
current salary excluding bonus
$.
$.

YES
NO
NO
YES
YES
YES
YES
YES
YES
YES
YES

Alphabetic List of Indexes and Attributes

The section shown in Output 15.6 appears only if the data set has indexes associated
with it.
#

indicates the number of each index. The indexes are numbered sequentially as
they are dened.
Index

displays the name of each index. For simple indexes, the name of the index is the
same as a variable in the data set.
Unique Option

indicates whether the index must have unique values. If the column contains YES,
the combination of values of the index variables is unique for each observation.
Nomiss Option

indicates whether the index excludes missing values for all index variables. If the
column contains YES, the index does not contain observations with missing values
for all index variables.
# of Unique Values

gives the number of unique values in the index.

Variables

names the variables in a composite index.

Output 15.6

Index Attributes Section

Alphabetic List of Indexes and Attributes

Index

Unique
Option

vital

YES

NoMiss
Option
YES

# of
Unique
Values
148

Variables
BIRTH SALARY

Sort Information
The section shown in Output 15.7 appears only if the Sorted eld has a value of YES.

368

PROC DATASETS and the Output Delivery System (ODS)

Chapter 15

Sortedby

indicates how the data are currently sorted. This eld contains either the
variables and options you use in the BY statement in PROC SORT, the column
name in PROC SQL, or the values you specify in the SORTEDBY= option.
Validated

indicates whether PROC SORT or PROC SQL sorted the data. If PROC SORT or
PROC SQL sorted the data set, the value is YES. If you assigned the sort
information with the SORTEDBY= data set option, the value is NO.
Character Set

is the character set used to sort the data. The value for this eld can be ASCII,
EBCDIC, or PASCII.
Collating Sequence

is the collating sequence used to sort the data set. This eld does not appear if you
do not specify a specic collating sequence that is different from the character set.
(not shown)
Sort Option

indicates whether PROC SORT used the NODUPKEY or NODUPREC option

when sorting the data set. This eld does not appear if you did not use one of
these options in a PROC SORT statement. (not shown)

Output 15.7

Sort Information Section

The SAS System

The DATASETS Procedure

Sort Information
Sortedby
Validated
Character Set

LNAME
NO
ANSI

PROC DATASETS and the Output Delivery System (ODS)

Most SAS procedures send their messages to the SAS log and their procedure results
to the output. PROC DATASETS is unique because it sends procedure results to both
the SAS log and the procedure output le. When the interface to ODS was created, it
was decided that all procedure results (from both the log and the procedure output le)
should be available to ODS. In order to implement this feature and maintain
compatibility with earlier releases, the interface to ODS had to be slightly different
from the usual interface.
By default, the PROC DATASETS statement itself produces two output objects:
Members and Directory. These objects are routed to the SAS log. The CONTENTS
statement produces three output objects by default: Attributes, EngineHost, and
Variables. (The use of various options adds other output objects.) These objects are
routed to the procedure output le. If you open an ODS destination (such as HTML,
RTF, or PRINTER), all of these objects are, by default, routed to that destination.
You can use ODS SELECT and ODS EXCLUDE statements to control which objects
go to which destination, just as you can for any other procedure. However, because of
the unique interface between PROC DATASETS and ODS, when you use the keyword

The DATASETS Procedure

ODS Table Names

369

LISTING in an ODS SELECT or ODS EXCLUDE statement, you affect both the log
and the listing.

ODS Table Names

PROC DATASETS and PROC CONTENTS assign a name to each table they create.
You can use these names to reference the table when using the Output Delivery System
(ODS) to select tables and create output data sets. For more information, see SAS
Output Delivery System: Users Guide.
PROC CONTENTS generates the same ODS tables as PROC DATASETS with the
CONTENTS statement.
Table 15.6 ODS Tables Produced by the DATASETS Procedure without the
CONTENTS Statement
ODS Table

Description

Table is generated:

General library information

unless you specify the NOLIST option.

Members

Library member
information

unless you specify the NOLIST option.

Table 15.7 ODS Table Names Produced by PROC CONTENTS and PROC DATASETS with the
CONTENTS Statement
ODS Table

Description

Table is generated:

Attributes

Data set attributes

unless you specify the SHORT option.

General library information

if you specify DATA=<libref.>_ALL_ or the

DIRECTORY option.*

EngineHost

Engine and operating

environment information

unless you specify the SHORT option.

IntegrityConstraints

A detailed listing of integrity

constraints

if the data set has integrtiy constraints and you do

not specify the SHORT option.

IntegrityConstraintsShort A concise listing of integrity

constraints

if the data set has integrity constraints and you

specify the SHORT option

Indexes

A detailed listing of indexes

if the data set is indexed and you do not specify the

SHORT option.

IndexesShort

A concise listing of indexes

if the data set is indexed and you specify the SHORT

option.

Members

Library member information

if you specify DATA=<libref.>_ALL_ or the

DIRECTORY option.*

Position

A detailed listing of variables by

logical position in the data set

if you specify the VARNUM option and you do not

specify the SHORT option.

PositionShort

A concise listing of variables by

logical position in the data set

if you specify the VARNUM option and the SHORT

option.

Sortedby

Detailed sort information

if the data set is sorted and you do not specify the

SHORT option.

SortedbyShort

Concise Sort information

if the data set is sorted and you specify the SHORT

option.

370

Output Data Sets

Chapter 15

ODS Table

Description

Table is generated:

Variables

A detailed listing of variables in

alphabetical order

unless you specify the SHORT option.

VariablesShort

A concise listing of variables in

alphabetical order

if you specify the SHORT option.

* For PROC DATASETS, if both the NOLIST option and either the DIRECTORY option or DATA=<libref.>_ALL_
are specied, then the NOLIST option is ignored.

Output Data Sets

The CONTENTS Statement

The CONTENTS statement is the only statement in the DATASETS procedure that
generates output data sets.

The OUT= Data Set

The OUT= option in the CONTENTS statement creates an output data set. Each
variable in each DATA= data set has one observation in the OUT= data set. These are
the variables in the output data set:
CHARSET
the character set used to sort the data set. The value is ASCII, EBCDIC, or
PASCII. A blank appears if the data set does not have sort information stored with
it.
COLLATE
the collating sequence used to sort the data set. A blank appears if the sort
information for the input data set does not include a collating sequence.
COMPRESS
indicates whether the data set is compressed.
CRDATE
date the data set was created.
DELOBS
number of observations marked for deletion in the data set. (Observations can be
marked for deletion but not actually deleted when you use the FSEDIT procedure
of SAS/FSP software.)
ENCRYPT
indicates whether the data set is encrypted.
ENGINE
name of the method used to read from and write to the data set.
FLAGS
indicates whether an SQL view is protected (P) or contributes (C) to a derived
variable.
P

indicates the variable is protected. The value of the variable

can be displayed but not updated.

indicates whether the variable contributes to a derived variable.

The DATASETS Procedure

Output Data Sets

371

The value of FLAG is blank if P or C does not apply to an SQL view or if it is a

data set view.
FORMAT
variable format. The value of FORMAT is a blank if you do not associate a format
with the variable.
FORMATD
number of decimals you specify when you associate the format with the variable.
The value of FORMATD is 0 if you do not specify decimals in the format.
FORMATL
format length. If you specify a length for the format when you associate the format
with a variable, the length you specify is the value of FORMATL. If you do not
specify a length for the format when you associate the format with a variable, the
value of FORMATL is the default length of the format if you use the FMTLEN
option and 0 if you do not use the FMTLEN option.
GENMAX
maximum number of versions for the generation group.
GENNEXT
the next generation number for a generation group.
GENNUM
the version number.
IDXCOUNT
number of indexes for the data set.
IDXUSAGE
use of the variable in indexes. Possible values are
NONE
the variable is not part of an index.
SIMPLE
the variable has a simple index. No other variables are included in the index.
COMPOSITE
the variable is part of a composite index.
BOTH
the variable has a simple index and is part of a composite index.
INFORMAT
variable informat. The value is a blank if you do not associate an informat with
the variable.
INFORMD
number of decimals you specify when you associate the informat with the variable.
The value is 0 if you do not specify decimals when you associate the informat with
the variable.
INFORML
informat length. If you specify a length for the informat when you associate the
informat with a variable, the length you specify is the value of INFORML. If you
do not specify a length for the informat when you associate the informat with a
variable, the value of INFORML is the default length of the informat if you use
the FMTLEN option and 0 if you do not use the FMTLEN option.
JUST
justication (0=left, 1=right).

372

Output Data Sets

Chapter 15

LABEL
variable label (blank if none given).
LENGTH
variable length.
LIBNAME
libref used for the data library.
MEMLABEL
label for this SAS data set (blank if no label).
MEMNAME
SAS data set that contains the variable.
MEMTYPE
library member type (DATA or VIEW).
MODATE
date the data set was last modied.
NAME
variable name.
NOBS
number of observations in the data set.
NODUPKEY
indicates whether the NODUPKEY option was used in a PROC SORT statement
to sort the input data set.
NODUPREC
indicates whether the NODUPREC option was used in a PROC SORT statement
to sort the input data set.
NPOS
physical position of the rst character of the variable in the data set.
POINTOBS
indicates if the data set can be addressed by observation.
PROTECT
the rst letter of the level of protection. The value for PROTECT is one or more of
the following:
A

indicates the data set is alter-protected.

indicates the data set is read-protected.

indicates the data set is write-protected.

REUSE
indicates whether the space made available when observations are deleted from a
compressed data set should be reused. If the data set is not compressed, the
REUSE variable has a value of NO.
SORTED
the value depends on the sorting characteristics of the input data set. Possible
values are
. (period)

for not sorted.

for sorted but not validated.

for sorted and validated.

The DATASETS Procedure

Output Data Sets

373

SORTEDBY
the value depends on that variables role in the sort. Possible values are
. (period)
if the variable was not used to sort the input data set.
n
where n is an integer that denotes the position of that variable in the sort. A
negative value of n indicates that the data set is sorted by the descending
order of that variable.
TYPE
type of the variable (1=numeric, 2=character).
TYPEMEM
special data set type (blank if no TYPE= value is specied).
VARNUM
variable number in the data set. Variables are numbered in the order they appear.
The output data set is sorted by the variables LIBNAME and MEMNAME.
Note: The variable names are sorted so that the values X1, X2, and X10 are listed
in that order, not in the true collating sequence of X1, X10, X2. Therefore, if you want
to use a BY statement on MEMNAME in subsequent steps, run a PROC SORT step on
the output data set rst or use the NOTSORTED option in the BY statement. 4
The following is an example of an output data set created from the GROUP data set,
which is shown in Example 4 on page 384 and in Procedure Output on page 364.

Output 15.8

The Data Set HEALTH.GRPOUT

An Example of an Output Data Set

OBS LIBNAME
1
2
3
4
5
6
7
8
9
10
11
OBS

HEALTH
HEALTH
HEALTH
HEALTH
HEALTH
HEALTH
HEALTH
HEALTH
HEALTH
HEALTH
HEALTH

MEMNAME
GROUP
GROUP
GROUP
GROUP
GROUP
GROUP
GROUP
GROUP
GROUP
GROUP
GROUP

MEMLABEL
Test
Test
Test
Test
Test
Test
Test
Test
Test
Test
Test

TYPEMEM

Subjects
Subjects
Subjects
Subjects
Subjects
Subjects
Subjects
Subjects
Subjects
Subjects
Subjects

LABEL

1
2
3
4
5
6
7
8
9 current salary excluding bonus
10
11

NAME

TYPE

BIRTH
CITY
FNAME
HIRED
HPHONE
IDNUM
JOBCODE
LNAME
SALARY
SEX
STATE
FORMAT
DATE
$
$
DATE
$
$
$
$
COMMA
$
$

1
LENGTH

VARNUM

1
2
2
1
2
2
2
2
1
2
2

8
15
15
8
12
4
3
15
8
1
2

9
4
3
10
11
1
7
2
8
6
5

FORMATL

FORMATD

INFORMAT

INFORML

7
0
0
7
0
0
0
0
8
0
0

0
0
0
0
0
0
0
0
0
0
0

DATE
$
$
DATE
$
$
$
$

7
0
0
7
0
0
0
0
0
0
0

$
$

374

Output Data Sets

Chapter 15

An Example of an Output Data Set

Obs INFORMD JUST NPOS NOBS ENGINE
1
2
3
4
5
6
7
8
9
10
11

0
0
0
0
0
0
0
0
0
0
0

OBS IDXUSAGE
1
2
3
4
5
6
7
8
9
10
11

COMPOSITE
NONE
NONE
NONE
NONE
NONE
NONE
NONE
COMPOSITE
NONE
NONE

1
0
0
1
0
0
0
0
1
0
0

8
58
43
16
79
24
76
28
0
75
73

148
148
148
148
148
148
148
148
148
148
148

V9
V9
V9
V9
V9
V9
V9
V9
V9
V9
V9

CRDATE
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46
29JAN02:08:06:46

MODATE DELOBS
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36
29JAN02:09:13:36

0
0
0
0
0
0
0
0
0
0
0

MEMTYPE IDXCOUNT PROTECT FLAGS COMPRESS REUSE SORTED SORTEDBY

DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA

1
1
1
1
1
1
1
1
1
1
1

R-R-R-R-R-R-R-R-R-R-R--

-----------------------

NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO

0
0
0
0
0
0
0
0
0
0
0

.
.
.
.
.
.
.
1
.
.
.

An Example of an Output Data Set

OBS CHARSET COLLATE NODUPKEY NODUPREC ENCRYPT POINTOBS GENMAX GENNUM GENNEXT
1
2
3
4
5
6
7
8
9
10
11

ANSI
ANSI
ANSI
ANSI
ANSI
ANSI
ANSI
ANSI
ANSI
ANSI
ANSI

NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO

YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES

0
0
0
0
0
0
0
0
0
0
0

.
.
.
.
.
.
.
.
.
.
.

Note: For information about how to get the CONTENTS output into an ODS data
set for processing, see Example 7 on page 389. 4

The OUT2= Data Set

The OUT2= option in the CONTENTS statement creates an output data set that
contains information about indexes and integrity constraints. These are the variables in
the output data set:
IC_OWN
contains YES if the index is owned by the integrity constraint.
INACTIVE
contains YES if the integrity constraint is inactive.

The DATASETS Procedure

Output Data Sets

375

LIBNAME
libref used for the data library.
MEMNAME
SAS data set that contains the variable.
MG
the value of MESSAGE=, if it is used, in the IC CREATE statement.
MSGTYPE
the value will be blank unless an integrity constraint is violated and you specied
a message.
NAME
the name of the index or integrity constraint.
NOMISS
contains YES if the NOMISS option is dened for the index.
NUMVALS
the number of distinct values in the index (displayed for centiles).
NUMVARS
the number of variables involved in the index or integrity constraint.
ONDELETE
for a foreign key integrity constraint, contains RESTRICT or SET NULL if
applicable (the ON DELETE option in the IC CREATE statement).
ONUPDATE
for a foreign key integrity constraint, contains RESTRICT or SET NULL if
applicable (the ON UPDATE option in the IC CREATE statement).
RECREATE
the SAS statement necessary to recreate the index or integrity constraint.
REFERENCE
for a foreign key integrity constraint, contains the name of the referenced data set.
TYPE
the type. For an index, the value is Index while for an integrity constraint, the
value is the type of integrity constraint (Not Null, Check, Primary Key, etc.).
UNIQUE
contains YES if the UNIQUE option is dened for the index.
UPERC
the percentage of the index that has been updated since the last refresh (displayed
for centiles).
UPERCMX
the percentage of the index update that triggers a refresh (displayed for centiles).
WHERE
for a check integrity constraint, contains the WHERE statement.

376

Examples: DATASETS Procedure

Chapter 15

Examples: DATASETS Procedure

Example 1: Manipulating SAS Files

Procedure features:

PROC DATASETS statement options:

DETAILS
LIBRARY=
CHANGE statement
COPY statement options:
MEMTYPE
MOVE
OUT=
DELETE statement option:
MEMTYPE=
EXCHANGE statement
EXCLUDE statement
SELECT statement option:
MEMTYPE=

This example
3 changes the names of SAS les
3 copies SAS les between SAS data libraries
3 deletes SAS les
3 selects SAS les to copy
3 exchanges the names of SAS les
3 excludes SAS les from a copy operation.

Program

Write the programming statements to the SAS log. The SOURCE system option
accomplishes this.
options pagesize=60 linesize=80 nodate pageno=1 source;

libname dest1 SAS-data-library-1;

libname dest2 SAS-data-library-2;
libname health SAS-data-library-3;

The DATASETS Procedure

Program

377

Specify the procedure input library, and add more details to the directory. DETAILS
prints these additional columns in the directory: Obs, Entries or Indexes, Vars, and
Label. All member types are available for processing because the MEMTYPE= option does not
appear in the PROC DATASETS statement.
proc datasets library=health details;

Delete two les in the library, and modify the names of a SAS data set and a catalog.
The DELETE statement deletes the TENSION data set and the A2 catalog. MT=CATALOG
applies only to A2 and is necessary because the default member type for the DELETE statement
is DATA. The CHANGE statement changes the name of the A1 catalog to POSTDRUG. The
EXCHANGE statement exchanges the names of the WEIGHT and BODYFAT data sets.
MEMTYPE= is not necessary in the CHANGE or EXCHANGE statement because the default is
MEMTYPE=ALL for each statement.
delete tension a2(mt=catalog);
change a1=postdrug;
exchange weight=bodyfat;

Restrict processing to one member type and delete and move data views.
MEMTYPE=VIEW restricts processing to SAS data views. MOVE species that all SAS data
views named in the SELECT statements in this step be deleted from the HEALTH data library
and moved to the DEST1 data library.
copy out=dest1 move memtype=view;

Move the SAS data view SPDATA from the HEALTH data library to the DEST1 data
library.
select spdata;

Move the catalogs to another data library. The SELECT statement species that the
catalogs ETEST1 through ETEST5 be moved from the HEALTH data library to the DEST1 data
library. MEMTYPE=CATALOG overrides the MEMTYPE=VIEW option in the COPY statement.
select etest1-etest5 / memtype=catalog;

Exclude all les with a specied criteria from processing. The EXCLUDE statement
excludes from the COPY operation all SAS les that begin with the letter D and the other SAS
les listed. All remaining SAS les in the HEALTH data library are copied to the DEST2 data
library.
copy out=dest2;
exclude d: mlscl oxygen test2 vision weight;
quit;

378

SAS Log

Chapter 15

SAS Log
1
options pagesize=60 linesize=80 nodate pageno=1 source;
2
libname dest1 c:\Myfiles\dest1;
NOTE: Libref DEST1 was successfully assigned as follows:
Engine:
V9
Physical Name: c:\Myfiles\dest1
3
libname dest2 c:\Myfiles\dest2;
NOTE: Libref DEST2 was successfully assigned as follows:
Engine:
V9
Physical Name: c:\Myfiles\dest2
4
libname health c:\Myfiles\health;
NOTE: Libref HEALTH was successfully assigned as follows:
Engine:
V9
Physical Name: c:\Myfiles\health
5
proc datasets library=health details;
Directory
Libref
Engine
Physical Name
File Name

HEALTH
V9
c:\Myfiles\health
c:\Myfiles\health

Name

Member
Type

Obs, Entries
or Indexes

1
2
3
4
5
6
7

A1
A2
ALL
BODYFAT
CONFOUND
CORONARY
DRUG1

CATALOG
CATALOG
DATA
DATA
DATA
DATA
DATA

23
1
23
1
8
39
6

17
2
4
4
2

DRUG2

DATA

DRUG3

DATA

DRUG4

DATA

DRUG5

DATA

12
13
14
15
16
17
18
19
20
21

ETEST1
ETEST2
ETEST3
ETEST4
ETEST5
ETESTS
FORMATS
GROUP
INFANT
MLSCL

CATALOG
CATALOG
CATALOG
CATALOG
CATALOG
CATALOG
CATALOG
DATA
DATA
DATA

1
1
1
1
1
1
6
148
149
32

11
6
4

22
23
24
25

NAMES
OXYGEN
PERSONL
PHARM

DATA
DATA
DATA
DATA

7
31
148
6

4
7
11
3

26
27
28
29
30
31
32
33
34
35
36

POINTS
PRENAT
RESULTS
SLEEP
SPDATA
SYNDROME
TENSION
TEST2
TRAIN
VISION
WEIGHT

DATA
DATA
DATA
DATA
VIEW
DATA
DATA
DATA
DATA
DATA
DATA

6
149
10
108
.
46
4
15
7
16
83

6
6
5
6
2
8
3
5
2
3
13

Vars

Label

JAN95
Data
MAY95
Data
JUL95
Data
JAN92
Data
JUL92
Data

Multiple
Sclerosi
s Data

Sugar
Study

Californ
ia
Results

File
Size

Last Modified

62464
17408
13312
5120
5120
5120
5120

19FEB2002:14:41:15
19FEB2002:14:41:15
19FEB2002:14:41:19
19FEB2002:14:41:19
19FEB2002:14:41:19
19FEB2002:14:41:20
19FEB2002:14:41:20

5120

19FEB2002:14:41:20

5120

19FEB2002:14:41:20

5120

19FEB2002:14:41:20

5120

19FEB2002:14:41:20

17408
17408
17408
17408
17408
17408
17408
25600
17408
5120

19FEB2002:14:41:20
19FEB2002:14:41:20
19FEB2002:14:41:20
19FEB2002:14:41:20
19FEB2002:14:41:20
19FEB2002:14:41:21
19FEB2002:14:41:21
19FEB2002:14:41:21
05FEB2002:12:52:30
19FEB2002:14:41:21

5120
9216
25600
5120

19FEB2002:14:41:21
19FEB2002:14:41:21
19FEB2002:14:41:21
19FEB2002:14:41:21

5120
17408
5120
9216
5120
9216
5120
5120
5120
5120
13312

19FEB2002:14:41:21
19FEB2002:14:41:22
19FEB2002:14:41:22
19FEB2002:14:41:22
19FEB2002:14:41:29
19FEB2002:14:41:22
19FEB2002:14:41:22
19FEB2002:14:41:22
19FEB2002:14:41:22
19FEB2002:14:41:22
19FEB2002:14:41:22

The DATASETS Procedure

WGHT

DATA

Californ
ia
Results

13312

SAS Log

19FEB2002:14:41:23

6
delete tension a2(mt=catalog);
7
change a1=postdrug;
8
exchange weight=bodyfat;
NOTE: Deleting HEALTH.TENSION (memtype=DATA).
NOTE: Deleting HEALTH.A2 (memtype=CATALOG).
NOTE: Changing the name HEALTH.A1 to HEALTH.POSTDRUG (memtype=CATALOG).
NOTE: Exchanging the names HEALTH.WEIGHT and HEALTH.BODYFAT (memtype=DATA).
9
copy out=dest1 move memtype=view;
10
select spdata;
11
select etest1-etest5 / memtype=catalog;
NOTE: Moving HEALTH.SPDATA to DEST1.SPDATA (memtype=VIEW).
NOTE: Moving HEALTH.ETEST1 to DEST1.ETEST1 (memtype=CATALOG).
NOTE: Moving HEALTH.ETEST2 to DEST1.ETEST2 (memtype=CATALOG).
NOTE: Moving HEALTH.ETEST3 to DEST1.ETEST3 (memtype=CATALOG).
NOTE: Moving HEALTH.ETEST4 to DEST1.ETEST4 (memtype=CATALOG).
NOTE: Moving HEALTH.ETEST5 to DEST1.ETEST5 (memtype=CATALOG).
12
copy out=dest2;
13
exclude d: mlscl oxygen test2 vision weight;
14
quit;
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:

Copying HEALTH.ALL to DEST2.ALL (memtype=DATA).

There were 23 observations read from the data set HEALTH.ALL.
The data set DEST2.ALL has 23 observations and 17 variables.
Copying HEALTH.BODYFAT to DEST2.BODYFAT (memtype=DATA).
There were 83 observations read from the data set HEALTH.BODYFAT.
The data set DEST2.BODYFAT has 83 observations and 13 variables.
Copying HEALTH.CONFOUND to DEST2.CONFOUND (memtype=DATA).
There were 8 observations read from the data set HEALTH.CONFOUND.
The data set DEST2.CONFOUND has 8 observations and 4 variables.
Copying HEALTH.CORONARY to DEST2.CORONARY (memtype=DATA).
There were 39 observations read from the data set HEALTH.CORONARY.
The data set DEST2.CORONARY has 39 observations and 4 variables.
Copying HEALTH.ETESTS to DEST2.ETESTS (memtype=CATALOG).
Copying HEALTH.FORMATS to DEST2.FORMATS (memtype=CATALOG).
Copying HEALTH.GROUP to DEST2.GROUP (memtype=DATA).
There were 148 observations read from the data set HEALTH.GROUP.
The data set DEST2.GROUP has 148 observations and 11 variables.
Copying HEALTH.INFANT to DEST2.INFANT (memtype=DATA).
There were 149 observations read from the data set HEALTH.INFANT.
The data set DEST2.INFANT has 149 observations and 6 variables.
Copying HEALTH.NAMES to DEST2.NAMES (memtype=DATA).
There were 7 observations read from the data set HEALTH.NAMES.
The data set DEST2.NAMES has 7 observations and 4 variables.
Copying HEALTH.PERSONL to DEST2.PERSONL (memtype=DATA).
There were 148 observations read from the data set HEALTH.PERSONL.
The data set DEST2.PERSONL has 148 observations and 11 variables.
Copying HEALTH.PHARM to DEST2.PHARM (memtype=DATA).
There were 6 observations read from the data set HEALTH.PHARM.
The data set DEST2.PHARM has 6 observations and 3 variables.
Copying HEALTH.POINTS to DEST2.POINTS (memtype=DATA).
There were 6 observations read from the data set HEALTH.POINTS.
The data set DEST2.POINTS has 6 observations and 6 variables.
Copying HEALTH.POSTDRUG to DEST2.POSTDRUG (memtype=CATALOG).
Copying HEALTH.PRENAT to DEST2.PRENAT (memtype=DATA).
There were 149 observations read from the data set HEALTH.PRENAT.
The data set DEST2.PRENAT has 149 observations and 6 variables.
Copying HEALTH.RESULTS to DEST2.RESULTS (memtype=DATA).
There were 10 observations read from the data set HEALTH.RESULTS.
The data set DEST2.RESULTS has 10 observations and 5 variables.
Copying HEALTH.SLEEP to DEST2.SLEEP (memtype=DATA).
There were 108 observations read from the data set HEALTH.SLEEP.
The data set DEST2.SLEEP has 108 observations and 6 variables.
Copying HEALTH.SYNDROME to DEST2.SYNDROME (memtype=DATA).
There were 46 observations read from the data set HEALTH.SYNDROME.
The data set DEST2.SYNDROME has 46 observations and 8 variables.
Copying HEALTH.TRAIN to DEST2.TRAIN (memtype=DATA).
There were 7 observations read from the data set HEALTH.TRAIN.
The data set DEST2.TRAIN has 7 observations and 2 variables.
Copying HEALTH.WGHT to DEST2.WGHT (memtype=DATA).
There were 83 observations read from the data set HEALTH.WGHT.
The data set DEST2.WGHT has 83 observations and 13 variables.

379

380

Example 2: Saving SAS Files from Deletion

Chapter 15

Example 2: Saving SAS Files from Deletion

Procedure features:

SAVE statement option:

MEMTYPE=

This example uses the SAVE statement to save some SAS les from deletion and to
delete other SAS les.

Program

Write the programming statements to the SAS log. SAS option SOURCE writes all
programming statements to the log.
options pagesize=40 linesize=80 nodate pageno=1 source;

libname elder SAS-data-library;

Specify the procedure input library to process.

proc datasets lib=elder;

Save the data sets CHRONIC, AGING, and CLINICS, and delete all other SAS les (of
all types) in the ELDER library. MEMTYPE=DATA is necessary because the ELDER library
has a catalog named CLINICS and a data set named CLINICS.
save chronic aging clinics / memtype=data;
run;

The DATASETS Procedure

Example 3: Modifying SAS Data Sets

SAS Log
41
options pagesize=40 linesize=80 nodate pageno=1 source;
42
libname elder c:\Myfiles\elder;
NOTE: Libref ELDER was successfully assigned as follows:
Engine:
V9
Physical Name: c:\Myfiles\elder
43
proc datasets lib=elder;
Directory
Libref
Engine
Physical Name
File Name

ELDER
V9
c:\Myfiles\elder
c:\Myfiles\elder

44
45
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:
NOTE:

Name

Member
Type

File
Size

Last Modified

1 AGING
DATA
5120
06FEB2003:08:51:21
2 ALCOHOL
DATA
5120
06FEB2003:08:51:21
3 BACKPAIN DATA
5120
06FEB2003:08:51:21
4 CHRONIC
DATA
5120
06FEB2003:08:51:21
5 CLINICS
CATALOG
17408
06FEB2003:08:51:21
6 CLINICS
DATA
5120
06FEB2003:08:51:21
7 DISEASE
DATA
5120
06FEB2003:08:51:21
8 GROWTH
DATA
5120
06FEB2003:08:51:21
9 HOSPITAL CATALOG
17408
06FEB2003:08:51:21
save chronic aging clinics / memtype=data;
run;
Saving ELDER.CHRONIC (memtype=DATA).
Saving ELDER.AGING (memtype=DATA).
Saving ELDER.CLINICS (memtype=DATA).
Deleting ELDER.ALCOHOL (memtype=DATA).
Deleting ELDER.BACKPAIN (memtype=DATA).
Deleting ELDER.CLINICS (memtype=CATALOG).
Deleting ELDER.DISEASE (memtype=DATA).
Deleting ELDER.GROWTH (memtype=DATA).
Deleting ELDER.HOSPITAL (memtype=CATALOG).

Example 3: Modifying SAS Data Sets

Procedure features:

PROC DATASETS statement option:

NOLIST
FORMAT statement
INDEX CREATE statement options:
NOMISS
UNIQUE
INFORMAT statement
LABEL statement

381

382

Program

Chapter 15

MODIFY statement options:

LABEL=
READ=
SORTEDBY=
RENAME statement

This example modies two SAS data sets using the MODIFY statement and
statements subordinate to it. Example 4 on page 384 shows the modications to the
GROUP data set.
Tasks include

3 modifying SAS les

3 labeling a SAS data set
3 adding a READ password to a SAS data set
3 indicating how a SAS data set is currently sorted
3 creating an index for a SAS data set
3 assigning informats and formats to variables in a SAS data set
3 renaming variables in a SAS data set
3 labeling variables in a SAS data set.

Program

Write the programming statements to the SAS log. SAS option SOURCE writes the
programming statements to the log.
options pagesize=40 linesize=80 nodate pageno=1 source;

libname health SAS-data-library;

Specify HEALTH as the procedure input library to process. NOLIST suppresses the
directory listing for the HEALTH data library.
proc datasets library=health nolist;

Add a label to a data set, assign a READ password, and specify how to sort the data.
LABEL= adds a data set label to the data set GROUP. READ= assigns GREEN as the read
password. The password appears as Xs in the SAS log. SAS issues a warning message if you
specify a level of password protection on a SAS le that does not include alter protection.
SORTEDBY= species how the data is sorted.
modify group (label=Test Subjects read=green sortedby=lname);

The DATASETS Procedure

SAS Log

383

Create the composite index VITAL on the variables BIRTH and SALARY for the
GROUP data set. NOMISS excludes all observations that have missing values for BIRTH and
SALARY from the index. UNIQUE species that the index is created only if each observation
has a unique combination of values for BIRTH and SALARY.
index create vital=(birth salary) / nomiss unique;

Assign an informat and format, respectively, to the BIRTH variable.

informat birth date7.;
format birth date7.;

Assign a label to the variable SALARY.

label salary=current salary excluding bonus;

Rename a variable, and assign a label. Modify the data set OXYGEN by renaming the
variable OXYGEN to INTAKE and assigning a label to the variable INTAKE.
modify oxygen;
rename oxygen=intake;
label intake=Intake Measurement;
quit;

SAS Log
6
options pagesize=40 linesize=80 nodate pageno=1 source;
7
libname health c:\Myfiles\health;
NOTE: Libref HEALTH was successfully assigned as follows:
Engine:
V9
Physical Name: c:\Myfiles\health
8
proc datasets library=health nolist;
9
modify group (label=Test Subjects read=XXXXX sortedby=lname);
WARNING: The file HEALTH.GROUP.DATA is not ALTER protected. It could be
deleted or replaced without knowing the password.
10
index create vital=(birth salary) / nomiss unique;
NOTE: Composite index vital has been defined.
11
informat birth date7.;
12
format birth date7.;
13
label salary=current salary excluding bonus;
14
modify oxygen;
15
rename oxygen=intake;
NOTE: Renaming variable oxygen to intake.
16
label intake=Intake Measurement;
17
quit;
NOTE: MODIFY was successful for HEALTH.OXYGEN.DATA.
NOTE: PROCEDURE DATASETS used (Total process time):
real time
16.96 seconds
cpu time
0.73 seconds

384

Example 4: Describing a SAS Data Set

Chapter 15

Example 4: Describing a SAS Data Set

Procedure features:

CONTENTS statement option:

DATA=
Other features:

SAS data set option:

READ=

This example shows the output from the CONTENTS statement for the GROUP data
set. The output shows the modications made to the GROUP data set in Example 3 on
page 381.

Program
options pagesize=40 linesize=132 nodate pageno=1;

libname health SAS-data-library;

Specify HEALTH as the procedure input library, and suppress the directory listing.
proc datasets library=health nolist;

Create the output data set GRPOUT from the data set GROUP. Specify GROUP as the
data set to describe, give read access to the GROUP data set, and create the output data set
GRPOUT, which appears in The OUT= Data Set on page 370.
contents data=group (read=green) out=grpout;
title The Contents of the GROUP Data Set;
run;

The DATASETS Procedure

Output

385

Output

Output 15.9

The Contents of the GROUP Data Set

The DATASETS Procedure

Data Set Name
Member Type
Engine

HEALTH.GROUP
DATA
V9

Observations
Variables
Indexes

148
11
1

Created
Last Modified

Wednesday, February 05, 2003 02:20:56

Thursday, February 06, 2003 09:07:54

Observation Length
Deleted Observations

96
0

Protection
Data Set Type
Label

READ

Compressed
Sorted

NO
YES

Data Representation
Encoding

WINDOWS_32
wlatin1 Western (Windows)

Test Subjects

Engine/Host Dependent Information

Data Set Page Size
Number of Data Set Pages

8192
4

First Data Page

Max Obs per Page
Obs in First Data Page

1
84
62

Index File Page Size

Number of Index File Pages
Number of Data Set Repairs

4096
2
0

File Name
Release Created
Host Created

c:\Myfiles\health\group.sas7bdat
9.0101B0
XP_PRO

Alphabetic List of Variables and Attributes

Variable

Type

Len

Format

Informat

9
4

BIRTH
CITY

Num
Char

8
15

DATE7.
$.

FNAME
HIRED
HPHONE

Char
Num
Char

15
8
12

$.
DATE7.
$.

3
10
11

Label

386

Example 5: Concatenating Two SAS Data Sets

Chapter 15

The Contents of the GROUP Data Set

The DATASETS Procedure
Alphabetic List of Variables and Attributes
#

Variable

Type

Len

Format

Informat

IDNUM

Char

7
2

JOBCODE
LNAME

Char
Char

3
15

$.
$.

8
6
5

SALARY
SEX
STATE

Num
Char
Char

8
1
2

$.
$.

COMMA8.
$.
$.

Label

$.
$.

current salary excluding bonus

Alphabetic List of Indexes and Attributes

Unique

NoMiss

# of
Unique
Values

Index

Option

vital

YES

148

Variables
BIRTH SALARY

Sort Information
Sortedby

LNAME

Validated
Character Set

NO
ANSI

Example 5: Concatenating Two SAS Data Sets

Procedure features:

APPEND statement options:

BASE=
DATA=
FORCE=

This example appends one data set to the end of another data set.

The DATASETS Procedure

Program

387

Input Data Sets

The BASE= data set, EXP.RESULTS.

The EXP.RESULTS Data Set

TREAT

INITWT

WT3MOS

AGE

1
2
3
5
6
7
10
11
12
13

Other
Other
Other
Other
Other
Other
Other
Other
Other
Other

166.28
214.42
172.46
175.41
173.13
181.25
239.83
175.32
227.01
274.82

146.98
210.22
159.42
160.66
169.40
170.94
214.48
162.66
211.06
251.82

35
54
33
37
20
30
48
51
29
31

The data set EXP.SUR contains the variable WT6MOS, but the EXP.RESULTS data set does not.

The EXP.SUR Data Set

treat

initwt

wt3mos

wt6mos

age

14
17
18

surgery
surgery
surgery

203.60
171.52
207.46

169.78
150.33
155.22

143.88
123.18
.

38
42
41

Program
options pagesize=40 linesize=64 nodate pageno=1;

libname exp SAS-data-library;

Suppress the printing of the EXP library. LIBRARY= species EXP as the procedure input
library. NOLIST suppresses the directory listing for the EXP library.
proc datasets library=exp nolist;

388

Output

Chapter 15

Append the data set EXP.SUR to the EXP.RESULTS data set. The APPEND statement
appends the data set EXP.SUR to the data set EXP.RESULTS. FORCE causes the APPEND
statement to carry out the append operation even though EXP.SUR has a variable that
EXP.RESULTS does not. APPEND does not add the WT6MOS variable to EXP.RESULTS.
append base=exp.results data=exp.sur force;
run;

Print the data set.

proc print data=exp.results noobs;
title The EXP.RESULTS Data Set;
run;

Output

Output 15.10
The EXP.RESULTS Data Set

TREAT

INITWT

WT3MOS

AGE

1
2
3
5
6
7
10
11
12
13
14
17
18

Other
Other
Other
Other
Other
Other
Other
Other
Other
Other
surgery
surgery
surgery

166.28
214.42
172.46
175.41
173.13
181.25
239.83
175.32
227.01
274.82
203.60
171.52
207.46

146.98
210.22
159.42
160.66
169.40
170.94
214.48
162.66
211.06
251.82
169.78
150.33
155.22

35
54
33
37
20
30
48
51
29
31
38
42
41

Example 6: Aging SAS Data Sets

Procedure features:

AGE statement

This example shows how the AGE statement ages SAS les.

The DATASETS Procedure

Example 7: PROC CONTENTS ODS Output

389

Program

Write the programming statements to the SAS log. SAS option SOURCE writes the
programming statements to the log.
options pagesize=40 linesize=80 nodate pageno=1 source;

libname daily SAS-data-library;

Specify DAILY as the procedure input library and suppress the directory listing.
proc datasets library=daily nolist;

Delete the last SAS le in the list, DAY7, and then age (or rename) DAY6 to DAY7,
DAY5 to DAY6, and so on, until it ages TODAY to DAY1.
age today day1-day7;
run;

SAS Log

6 options pagesize=40 linesize=80 nodate pageno=1 source;

7
8
proc datasets library=daily nolist;
9
10
age today day1-day7;
11
run;
NOTE: Deleting DAILY.DAY7 (memtype=DATA).
NOTE: Ageing the name DAILY.DAY6 to DAILY.DAY7 (memtype=DATA).
NOTE: Ageing the name DAILY.DAY5 to DAILY.DAY6 (memtype=DATA).
NOTE: Ageing the name DAILY.DAY4 to DAILY.DAY5 (memtype=DATA).
NOTE: Ageing the name DAILY.DAY3 to DAILY.DAY4 (memtype=DATA).
NOTE: Ageing the name DAILY.DAY2 to DAILY.DAY3 (memtype=DATA).
NOTE: Ageing the name DAILY.DAY1 to DAILY.DAY2 (memtype=DATA).
NOTE: Ageing the name DAILY.TODAY to DAILY.DAY1 (memtype=DATA).

Example 7: PROC CONTENTS ODS Output

Procedures features:

CONTENTS Statement

The example shows how to get PROC CONTENTS output into an ODS output data
set for processing.

390

Program

Chapter 15

Program
title1 "PROC CONTENTS ODS Output";
options nodate nonumber nocenter formdlim=-;
data a;
x=1;
run;

Use the ODS OUTPUT statement to specify data sets to which CONTENTS data will be
directed.

ods output attributes=atr

variables=var
enginehost=eng;

Temporarily suppress output to the lst.

ods listing close;
proc contents data=a;
run;

Resume output to the lst.

ods listing;
title2 "all Attributes data";
proc print data=atr noobs;
run;
title2 "all Variables data";
proc print data=var noobs;
run;
title2 "all EngineHost data";
proc print data=eng noobs;
run;

Select specic data from ODS output.

ods output attributes=atr1(keep=member cvalue1 label1
where=(attribute in (Data Representation,Encoding))
rename=(label1=attribute cvalue1=value))
attributes=atr2(keep=member cvalue2 label2
where=(attribute in (Observations, Variables))
rename=(label2=attribute cvalue2=value));

The DATASETS Procedure

ods listing close;

proc contents data=a;
run;
ods listing;
data final;
set atr1 atr2;
run;
title2 "example of post-processing of ODS output data";
proc print data=final noobs;
run;
ods listing close;

Output 15.11

PROC CONTENTS ODS Output

all Attributes data

Member

Label1

cValue1

WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A

Data Set Name

Member Type
Engine
Created
Last Modified
Protection
Data Set Type
Label
Data Representation
Encoding

WORK.A
DATA
V9
Thursday, October 10, 2002 00:56:03
Thursday, October 10, 2002 00:56:03

nValue1
.
.
.
1349873763
1349873763
.
.
.
.
.

WINDOWS_32
wlatin1 Western (Windows)
c
Value2

Label2
Observations
Variables
Indexes
Observation Length
Deleted Observations
Compressed
Sorted

1
1
0
8
0
NO
NO

1.000000
1.000000
0
8.000000
0
.
.
0
0
0

PROC CONTENTS ODS Output

all Variables data
Member

Num

WORK.A

Variable
x

Type
Num

Len
8

nValue2

Pos
0

Program

391

392

Program

Chapter 15

PROC CONTENTS ODS Output

all EngineHost data
Member

Label1

WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A
WORK.A

Data Set Page Size

Number of Data Set Pages
First Data Page
Max Obs per Page
Obs in First Data Page
Number of Data Set Repairs
File Name
Release Created
Host Created

cValue1

nValue1

4096
4096.000000
1
1.000000
1
1.000000
501
501.000000
1
1.000000
0
0
C:\DOCUME~1\userid\LOCALS~1\Temp\SAS Temporary Files\_TD3084\a.sas7bdat
.
9.0101B0
.
XP_PRO
.

PROC CONTENTS ODS Output

example of post-processing of ODS output data
Member

attribute

value

WORK.A
WORK.A
WORK.A
WORK.A

Data Representation
Encoding
Observations
Variables

WINDOWS_32
wlatin1 Western (Windows)
1
1

For more information on the SAS Output Delivery System, see SAS Output Delivery
System: Users Guide.

393

CHAPTER

16
The DBCSTAB Procedure
Information about the DBCSTAB Procedure

393

Information about the DBCSTAB Procedure

For documentation of the DBCSTAB procedure, see SAS National Language
Support (NLS): Users Guide.

See:

394

395

CHAPTER

17
The DISPLAY Procedure
Overview: DISPLAY Procedure 395
Syntax: DISPLAY Procedure
395
PROC DISPLAY Statement 395
Example: DISPLAY Procedure
396
Example 1: Executing a SAS/AF Application

396

Overview: DISPLAY Procedure

The DISPLAY procedure executes SAS/AF applications. These applications are
composed of a variety of entries that are stored in a SAS catalog and that have been
built with the BUILD procedure in SAS/AF software. For complete documentation on
building SAS/AF applications, see SAS Guide to Applications Development.
You can use the DISPLAY procedure to execute an application that runs in NODMS
batch mode. Be aware that any SAS programming statements that you submit with the
DISPLAY procedure through the SUBMIT block in SCL are not submitted for
processing until PROC DISPLAY has executed.
If you use the SAS windowing environment, you can use the AF command to execute
an application. SUBMIT blocks execute immediately when you use the AF command.
You can use the AFA command to execute multiple applications concurrently.

Syntax: DISPLAY Procedure

PROC DISPLAY CATALOG=libref.catalog.entry.type <BATCH>;

PROC DISPLAY Statement

Featured in: Example 1 on page 396

PROC DISPLAY CATALOG=libref.catalog.entry.type <BATCH>;

396

Example: DISPLAY Procedure

Chapter 17

Required Argument
CATALOG=libref.catalog.entry.type

species a four-level name for the catalog entry.

libref
species the SAS data library where the catalog is stored.
catalog
species the name of the catalog.
entry
species the name of the entry.
type
species the entrys type, which is one of the following. For details, see the
description of catalog entry types in the BUILD procedure in online help.
CBT
FRAME
HELP
MENU
PROGRAM
SCL

Options
BATCH

runs PROGRAM and SCL entries in batch mode. If a PROGRAM entry contains a
display, then it will not run, and you will receive the following error message:
ERROR: Cannot allocate window.

Restriction: PROC DISPLAY cannot pass arguments to a PROGRAM, a FRAME,

or an SCL entry.

Example: DISPLAY Procedure

Example 1: Executing a SAS/AF Application

Procedure features:

PROC DISPLAY statement:

CATALOG = argument

Suppose that your company has developed a SAS/AF application that compiles
statistics from an invoice database. Further, suppose that this application is stored in
the SASUSER data library, as a FRAME entry in a catalog named
INVOICES.WIDGETS. You can execute this application using the following SAS code:

The DISPLAY Procedure

Program
proc display catalog=sasuser.invoices.widgets.frame;
run;

Program

397

398

399

CHAPTER

18
The DOCUMENT Procedure
Information about the DOCUMENT Procedure

399

Information about the DOCUMENT Procedure

For complete documentation of the DOCUMENT procedure, see SAS Output
Delivery System: Users Guide.

See:

400

401

CHAPTER

19
The EXPLODE Procedure
Information about the EXPLODE Procedure

401

Information about the EXPLODE Procedure

See: For documentation of the EXPLODE procedure, go to http://support.sas.com/
documentation/onlinedoc. Select Base SAS from the Product-Specic Documentation

list.

402

403

CHAPTER

20
The EXPORT Procedure
Overview: EXPORT Procedure 403
Syntax: EXPORT Procedure 404
PROC EXPORT Statement 404
Data Source Statements 408
Examples: EXPORT Procedure 411
Example 1: Exporting a Delimited External File 411
Example 2: Exporting a Subset of Observations to an Excel Spreadsheet 414
Example 3: Exporting to a Specic Spreadsheet in an Excel Workbook 415
Example 4: Exporting a Microsoft Access Table 415
Example 5: Exporting a Specic Spreadsheet in an Excel Workbook on a PC Server

416

Overview: EXPORT Procedure

The EXPORT procedure reads data from a SAS data set and writes it to an external
data source. External data sources can include Microsoft Access Database, Excel les,
Lotus spreadsheets, and delimited external les (in which columns of data values are
separated by a delimiter such as a blank, comma, or tab).
When you execute PROC EXPORT, the procedure reads the input data set and writes
the data to the external data source. PROC EXPORT exports the data by one of the
following methods:

3 generated DATA step code

3 generated SAS/ACCESS code
3 translation engines.
You control the results with options and statements that are specic to the output data
source. PROC EXPORT produces the specied output le and writes information about
the export to the SAS log. In the log, you see the DATA step or the SAS/ACCESS code
that is generated by PROC EXPORT. If a translation engine is used, then no code is
submitted.
Note: To export data, you can also use the Export Wizard, which is a windowing tool
that guides you through the steps to export a SAS data set. You can request the Export
Wizard to generate EXPORT procedure statements, which you can save to a le for
subsequent use. To invoke the Export Wizard, from the SAS windowing environment
select
File

Export Data

404

Syntax: EXPORT Procedure

Chapter 20

Syntax: EXPORT Procedure

Restriction:

PROC EXPORT is available for the following operating environments:

3 OpenVMS Alpha
3 UNIX
3 Microsoft Windows.

PROC EXPORT DATA=<libref.>SAS-data-set <(SAS-data-set-options)>

OUTFILE="lename" | OUTTABLE="tablename"
<DBMS=identier> <REPLACE>;
<data-source-statement(s);>

PROC EXPORT Statement

Featured in:

All examples

PROC EXPORT DATA=<libref.>SAS-data-set <(SAS-data-set-options)>

OUTFILE="lename" | OUTTABLE="tablename"
<DBMS=identier> <REPLACE>;

Required Arguments
DATA=<libref.>SAS-data-set

identies the input SAS data set with either a one- or two-level SAS name (library
and member name). If you specify a one-level name, by default, PROC EXPORT uses
either the USER library (if assigned) or the WORK library (if USER not assigned).
Default: If you do not specify a SAS data set, PROC EXPORT uses the most
recently created SAS data set, which SAS keeps track of with the system variable
_LAST_. However, in order to be certain that PROC EXPORT uses the correct
data set, you should identify the SAS data set.
Restriction: PROC EXPORT can export data only if the format of the data is
supported by the data source or the amount of data is within the limitations of the
data source. For example, some data sources have a maximum number of rows or
columns, and some data sources cannot support SAS user-dened formats and
informats. If the data that you want to export exceeds the limits of the data
source, PROC EXPORT may not be able to export it correctly. When incompatible
formats are encountered, the procedure formats the data to the best of its ability.
Restriction: PROC EXPORT does not support writing labels as column names.
However, SAS does support column names up to 32 characters.
Featured in: All examples
(SAS-data-set-options)

species SAS data set options. For example, if the data set that you are exporting
has an assigned password, you can use the ALTER=, PW=, READ=, or WRITE= data

The EXPORT Procedure

PROC EXPORT Statement

405

set option, or to export only data that meets a specied condition, you can use the
WHERE= data set option. For information about SAS data set options, see Data Set
Options in SAS Language Reference: Dictionary.
Restriction: You cannot specify data set options when exporting delimited,
comma-separated, or tab-delimited external les.
Featured in: Example 2 on page 414
OUTFILE="lename"

species the complete path and lename or a leref for the output PC le,
spreadsheet, or delimited external le. If you specify a leref or if the complete path
and lename does not include special characters (such as the backslash in a path),
lowercase characters, or spaces, you can omit the quotation marks. A leref is a SAS
name that is associated with the physical location of the output le. To assign a
leref, use the FILENAME statement. For more information about PC le formats,
see SAS/ACCESS for PC Files: Reference.
Featured in Example 1 on page 411, Example 2 on page 414, and Example 3 on
page 415
Restriction: PROC EXPORT does not support device types or access methods for
the FILENAME statement except for DISK. For example, PROC EXPORT does
not support the TEMP device type, which creates a temporary external le.
Restriction: For client/server applications: When running SAS/ACCESS software
on UNIX to access data that is stored on a PC server, you must specify the full
path and lename of the le that you want to import. The use of a leref is not
supported.
OUTTABLE="tablename"

species the table name of the output DBMS table. If the name does not include
special characters (such as question marks), lowercase characters, or spaces, you can
omit the quotation marks. Note that the DBMS table name may be case sensitive.
Requirement: When you export a DBMS table, you must specify the DBMS= option.
Featured in: Example 4 on page 415

Options
DBMS=identier

species the type of data to export. To export a DBMS table, you must specify
DBMS= by using a valid database identier. For example, DBMS=ACCESS species
to export a table into a Microsoft Access 2000 or 2002 database. To export PC les,
spreadsheets, and delimited external les, you do not have to specify DBMS= if the
lename that is specied in OUTFILE= contains a valid extension so that PROC
EXPORT can recognize the type of data. For example, PROC EXPORT recognizes the
lename ACCOUNTS.WK1 as a Lotus 1-2-3 Release 2 spreadsheet and the lename
MYDATA.CSV as an external le that contains comma-separated data values;
therefore, a DBMS= specication is not necessary.
The following values are valid for the DBMS= option:
Identier

Output Data Source

Extension

Host
Version
Availability of File
Created

ACCESS

Microsoft Access 2000 or 2002 table

.mdb

Microsoft
Windows *

2000

ACCESS97

Microsoft Access 97 table

.mdb

Microsoft
Windows *

406

PROC EXPORT Statement

Identier

Chapter 20

Output Data Source

Extension

Host
Version
Availability of File
Created

.csv

OpenVMS
Alpha,
UNIX,
Microsoft
Windows

DBF

dBASE 5.0, IV, III+, and III les

.dbf

UNIX,
Microsoft
Windows

DLM

delimited le (default delimiter is a

Microsoft
Windows *

EXCELCS

Excel spreadsheet

.xls

UNIX

97**

JMP

JMP table

.jmp

UNIX,
Microsoft
Windows

PCFS

Files on PC server

UNIX

TAB

delimited le (tab-delimited values)

.txt

OpenVMS
Alpha,
UNIX,
Microsoft
Windows

WK1

Lotus 1-2-3 Release 2 spreadsheet

.wk1

Microsoft
Windows

5.0

The EXPORT Procedure

PROC EXPORT Statement

Identier

Output Data Source

Extension

Host
Version
Availability of File
Created

WK3

Lotus 1-2-3 Release 3 spreadsheet

.wk3

Microsoft
Windows

WK4

Lotus 1-2-3 Release 4 and 5 spreadsheet

.wk4

407

Microsoft
Windows

* Not available for Microsoft Windows 64-Bit Edition.** Value listed here is the
default value. The real version of le loaded depends on the version of the existing
le or the value specied for VERSION= statement.
Restriction: The availability of an output data source depends on
3 the operating environment, and in some cases the platform, as specied in
the previous table.
3 whether your site has a license to the SAS/ACCESS software for PC le
formats. If you do not have a license, only delimited les are available.
Example 1 on page 411 and Example 4 on page 415
When you specify a value for DBMS=, consider the following for specic data
sources:
3 To export to an existing Microsoft Access database, PROC EXPORT can write to
Access 97, Access 2000, or Access 2002 regardless of your specication. For
example, if you specify DBMS=ACCESS2000 and the database is in Access 97
format, PROC EXPORT exports the table, and the database remains in Access
97 format. However, if you specify OUTFILE= for an Access database that does
not exist, a new database is created using the format specied in DBMS=. For
example to create a new Access database, specifying DBMS=ACCESS (which
defaults to Access 2000 or 2002 format) creates an MDB le that can be read by
Access 2000 or Access 2002, not by Access 97.
The following table lists the DBMS= specications and indicates which
version of Microsoft Access can open the resulting database:
Featured in:

Specication

Access 2002

Access 2000

Access 97

ACCESS

yes

ACCESS2002

yes

ACCESS2000

yes

ACCESS97

yes

3 To export a Microsoft Excel spreadsheet, PROC EXPORT creates an XLS le for

the version specied. The following table lists the DBMS= specications and
indicates which version of Microsoft Excel can open the resulting spreadsheet:
Specication

Excel 2002

Excel 2000

Excel 97

Excel 5.0

Excel 4.0

EXCEL

yes

EXCEL2002

yes

EXCEL2000

yes

408

Data Source Statements

Chapter 20

Specication

Excel 2002

Excel 2000

Excel 97

Excel 5.0

Excel 4.0

EXCEL97

yes

EXCEL5

yes

EXCEL4

yes

Note: Later versions of Excel can open and update les in earlier formats.

3 When exporting a SAS data set to a dBASE le (DBF), if the data set contains
missing values (for either character or numeric values), the missing values are
translated to blanks.

3 When exporting a SAS data set to a dBASE le (DBF), values for a character
variable that are longer than 255 characters are truncated in the resulting
dBASE le because of dBASE limitations.
REPLACE

overwrites an existing le. Note that for a Microsoft Access database or an Excel
workbook, REPLACE overwrites the target table or spreadsheet. If you do not
specify REPLACE, PROC EXPORT does not overwrite an existing le.
Featured in:

Example 2 on page 414 and Example 4 on page 415

Data Source Statements

PROC EXPORT provides a variety of statements that are specic to the output data
source.

Statements for PC Files, Spreadsheets, or Delimited Files

The following statement is available when you export delimited external les:
DELIMITER=char | nnx;
species the delimiter to separate columns of data in the output le. You can
specify the delimiter as a single character or as a hexadecimal value. For example,
if you want columns of data to be separated by an ampersand, specify
DELIMITER=&. If you do not specify DELIMITER=, PROC EXPORT assumes
that the delimiter is a blank. You can replace the equal sign with a blank.
Interaction: You do not have to specify DELIMITER= if you specify DBMS=CSV,

DBMS=TAB, or if the output lename has an extension of .CSV or .TXT.

Featured in:

Example 1 on page 411

SHEET=spreadsheet-name;
identies a particular spreadsheet name to load into a workbook. You use this
statement for Microsoft Excel 97, 2000, or 2002 only. If the SHEET= statement is
not specied, PROC EXPORT uses the SAS data set name as the spreadsheet
name to load the data.

The EXPORT Procedure

Data Source Statements

409

For Excel data access, a spreadsheet name is treated as a special case of a range
name with a dollar sign ($) appended. For example, if you export a table and
specify sheet=Invoice, you will see a range (table) name INVOICE and another
range (table) name INVOICES$ created. Excel appends a dollar sign ($) to a
spreadsheet name in order to distinguish it from the corresponding range name.
Note: You should not append the dollar sign ($) when you specify the
spreadsheet name. For example, SHEET= Invoice$ is not allowed. 4
You should avoid using special characters for spreadsheet names when
exporting a table to an Excel le. Special characters such as a space or a hyphen
are replaced with an underscore. For example, if you export a table and specify
sheet=Sheet Number 1, PROC EXPORT creates the range names
Sheet_Number_1 and Sheet_Number_1$.
Featured in: Example 3 on page 415

Statements for DBMS Tables

The following statements are available to establish a connection to the DBMS when
you are exporting to a DBMS table:
DATABASE="database";
species the complete path and lename of the database to contain the specied
DBMS table. If the database name does not contain lowercase characters, special
characters, or national characters ($, #, or @), you can omit the quotation marks.
You can replace the equal sign with a blank.
Note: A default may be congured in the DBMS client software; SAS does not
generate a default value. 4
Featured in: Example 4 on page 415

DBPWD="database-password";
species a password that allows access to a database. You can replace the equal
sign with a blank.
PWD="password";
species the user password used by the DBMS to validate a specic userid. If the
password does not contain lowercase characters, special characters, or national
characters, you can omit the quotation marks. You can replace the equal sign with
a blank.
Note: The DBMS client software may default to the userid and password that
was used to log in to the operating environment; SAS does not generate a default
value. 4
UID="userid";
identies the user to the DBMS. If the userid does not contain lowercase
characters, special characters, or national characters, you can omit the quotation
marks. You can replace the equal sign with a blank.
Note: The DBMS client software may default to the userid and password that
were used to log in to the operating environment; SAS does not generate a default
value. 4
WGDB="workgroup-database-name";
species the workgroup (security) database name that contains the USERID and
PWD data for the DBMS. If the workgroup database name does not contain
lowercase characters, special characters, or national characters, you can omit the
quotation marks. You can replace the equal sign with a blank.

410

Data Source Statements

Chapter 20

Note: A default workgroup database may be used by the DBMS; SAS does not
generate a default value. 4

Security Levels for Microsoft Access Tables

Microsoft Access tables have the following levels of security, for which specic
combinations of security statements must be used:
None
Do not specify DBPWD=, PWD=, UID=, or WGDB=.
Password
Specify only DBPWD=.
User-level
Specify only PWD=, UID=, and WGDB=.
Full
Specify DBPWD=, PWD=, UID=, and WGDB=.
Each statement has a default value; however, you may nd it necessary to provide a
value for each statement explicitly.

Statement for Client/Server Model

The following statements are available to establish a connection from SAS running on
UNIX to a PC server when you are exporting a table to Microsoft Access database or
Excel workbook:
SERVER="PC-server-name";
species the name of the PC server. You must bring up the listener on the PC
server before you can establish a connection to it. You can congure the service
name, port number, maximum number of connections allowed, and use of data
encryption on your PC server. This is a required statement. Refer to your PC
server administrator for the information that is needed.
Alias: SERVER_NAME=
SERVICE="service-name";
species the service name that is dened on your service le for your client and
server machines. This statement and the PORT= statement should not be used in
the same procedure. Note that this service name needs to be dened on both your
UNIX machine and your PC server.
Alias: SERVER_NAME=, SERVICE_NAME=
PORT=port-number;
species the number of the port that is listening on the PC server. The valid value
is between 1 and 32767. This statement and the SERVICE= statement should not
be used in the same procedure.
Alias: PORT_NUMBER=
VERSION="le-version";
species the version of the le that you want to create with if the le does not
exist on your PC server yet. The default version is data-source specic. For
Microsoft Access database, the valid values are 2002, 2000 and 97, and its
default value is 2000. For Microsoft Excel workbook, the valid values are 2002,
2000, 97, 95 and 5, and its default value is 97.
Note: Always quote the version value.
Note:
ignored.

If the le already exists in the PC Server, then this value can be

The EXPORT Procedure

Example 1: Exporting a Delimited External File

Examples: EXPORT Procedure

Example 1: Exporting a Delimited External File

Procedure features:

PROC EXPORT statement arguments:

DATA=
DBMS=
OUTFILE=
Data source statement:
DELIMITER=

This example exports the following SAS data set named SASHELP.CLASS and
creates a delimited external le:

Output 20.1

PROC PRINT of SASHELP.CLASS

The SAS System

Obs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

Name
Alfred
Alice
Barbara
Carol
Henry
James
Jane
Janet
Jeffrey
John
Joyce
Judy
Louise
Mary
Philip
Robert
Ronald
Thomas
William

Sex
M
F
F
F
M
M
F
F
M
M
F
F
F
F
M
M
M
M
M

Age

Height

Weight

14
13
13
14
14
12
12
15
13
12
11
14
12
15
16
12
15
11
15

69
56.5
65.3
62.8
63.5
57.3
59.8
62.5
62.5
59
51.3
64.3
56.3
66.5
72
64.8
67
57.5
66.5

112.5
84
98
102.5
102.5
83
84.5
112.5
84
99.5
50.5
90
77
112
150
128
133
85
112

411

412

Program

Chapter 20

Program

Identify the input SAS data set, specify the output lename, and specify the type of
le. Note that the lename does not contain an extension. DBMS=DLM species that the
output le is a delimited external le.
proc export data=sashelp.class
outfile=c:\myfiles\class
dbms=dlm;

Specify the delimiter. The DELIMITER= option species that an & (ampersand) will delimit
data elds in the output le. The delimiter separates the columns of data in the output le.
delimiter=&;
run;

SAS Log
The SAS log displays the following information about the successful export. Notice
the generated SAS DATA step.

The EXPORT Procedure

SAS Log

47
/**********************************************************************
48
*
PRODUCT:
SAS
49
*
VERSION:
9.00
50
*
CREATOR:
External File Interface
51
*
DATE:
07FEB02
52
*
DESC:
Generated SAS Datastep Code
53
*
TEMPLATE SOURCE: (None Specified.)
54
***********************************************************************/
55
data _null_;
56
set SASHELP.CLASS
end=EFIEOD;
57
%let _EFIERR_ = 0; /* set the ERROR detection macro variable */
58
%let _EFIREC_ = 0;
/* clear export record count macro variable */
59
file c:\myfiles\class delimiter=& DSD DROPOVER
59 ! lrecl=32767;
60
format Name $8. ;
61
format Sex $1. ;
62
format Age best12. ;
63
format Height best12. ;
64
format Weight best12. ;
65
if _n_ = 1 then
/* write column names */
66
do;
67
put
68
Name
69
&
70
Sex
71
&
72
Age
73
&
74
Height
75
&
76
Weight
77
;
78
end;
79
do;
80
EFIOUT + 1;
81
put Name $ @;
82
put Sex $ @;
83
put Age @;
84
put Height @;
85
put Weight ;
86
;
87
end;
88
if _ERROR_ then call symput(_EFIERR_,1); /* set ERROR detection
88 ! macro variable */
89
If EFIEOD then
90
call symput(_EFIREC_,EFIOUT);
91
run;
NOTE: Numeric values have been converted to character
values at the places given by: (Line):(Column).
88:44
90:31
NOTE: The file c:\myfiles\class is:
File Name=c:\myfiles\class,
RECFM=V,LRECL=32767
NOTE: 20 records were written to the file c:\myfiles\class.
The minimum record length was 17.
The maximum record length was 26.
NOTE: There were 19 observations read from the data set SASHELP.CLASS.
NOTE: DATA statement used (Total process time):
real time
0.13 seconds
cpu time
0.05 seconds

19 records created in c:\myfiles\class from SASHELP.CLASS

NOTE: c:\myfiles\class was successfully created.

413

414

Output

Chapter 20

Output
The external le produced by PROC EXPORT follows.
Name&Sex&Age&Height&Weight
Alfred&M&14&69&112.5
Alice&F&13&56.5&84
Barbara&F&13&65.3&98
Carol&F&14&62.8&102.5
Henry&M&14&63.5&102.5
James&M&12&57.3&83
Jane&F&12&59.8&84.5
Janet&F&15&62.5&112.5
Jeffrey&M&13&62.5&84
John&M&12&59&99.5
Joyce&F&11&51.3&50.5
Judy&F&14&64.3&90
Louise&F&12&56.3&77
Mary&F&15&66.5&112
Philip&M&16&72&150
Robert&M&12&64.8&128
Ronald&M&15&67&133
Thomas&M&11&57.5&85
William&M&15&66.5&112

Example 2: Exporting a Subset of Observations to an Excel Spreadsheet

Procedure features:

PROC EXPORT statement arguments:

DATA=
DBMS=
OUTFILE=
REPLACE

This example exports the SAS data set SASHELP.CLASS, shown in Output 20.1.
PROC EXPORT creates an Excel le named Femalelist.xsl, and by default, creates a
spreadsheet named Class. Since the SHEET= data source statement is not specied,
PROC EXPORT uses the name of the SAS data set as the spreadsheet name. The
WHERE= SAS data set option is specied in order to export a subset of the
observations, which results in the spreadsheet containing only the female students.

Program

Identify the input SAS data set, request a subset of the observations, specify the
output data source, specify the output le, and overwrite the target spreadsheet if it
exists. The output le is an Excel 2000 spreadsheet.
proc export data=sashelp.class (where=(sex=F))
outfile=c:\myfiles\Femalelist.xls

The EXPORT Procedure

Example 4: Exporting a Microsoft Access Table

415

dbms=excel
replace;
run;

Example 3: Exporting to a Specic Spreadsheet in an Excel Workbook

Procedure features:

PROC EXPORT statement arguments:

DATA=
DBMS=
OUTFILE=
Data Source Statement:
SHEET=

This example exports a SAS data set named MYFILES.GRADES1 and creates an
Excel 2000 workbook named Grades.xsl. MYFILES.GRADES1 becomes one spreadsheet
in the workbook named Grades1.

Program
Identify the input SAS data set, specify the output data source, and specify the output
le.
proc export data=myfiles.grades1
dbms=excel2000
outfile=c:\Myfiles\Grades.xls;

Identify a particular spreadsheet to write to in a workbook.

sheet=Grades1;
run;

Example 4: Exporting a Microsoft Access Table

Procedure features:

PROC EXPORT statement arguments:

DATA=
DBMS=
OUTTABLE=
REPLACE
Data Source Statement:
DATABASE=

This example exports a SAS data set named SASUSER.CUST, the rst ve
observations of which follow, and creates a Microsoft Access 97 table. The security level

416

Program

Chapter 20

for this Access table is none, so it is not necessary to specify any of the database
security statements.
Obs

Name

Zipcode

David Taylor
Theo Barnes
Lydia Stirog
Anton Niroles
Cheryl Gaspar

1
2
3
4
5

Street
124 Oxbow Street
2412 McAllen Avenue
12550 Overton Place
486 Gypsum Street
36 E. Broadway

72511
72513
72516
72511
72515

Program

Identify the input SAS data set, specify the output DBMS table name and the output
data source, and overwrite the output le if it exists. The output le is a Microsoft Access
97 table. The option REPLACE overwrites an existing le. If you do not specify REPLACE,
PROC EXPORT does not overwrite an existing le.
proc export data=sasuser.cust
outtable="customers"
dbms=access97
replace;

Specify the path and lename of the database to contain the table.
database="c:\myfiles\mydatabase.mdb";
run;

Example 5: Exporting a Specic Spreadsheet in an Excel Workbook on a PC

Server
Procedure features:

PROC EXPORT statement arguments:

DATA=
DBMS=
OUTFILE=
Data Source Statement:
SHEET=
SERVER=
PORT=
VERSION=

This example exports a SAS data set named SASHELP.CLASS and creates an Excel
2000 workbook named demo.xls. SASHELP.CLASS becomes one spreadsheet named
Class in the workbook named demo.xls.

The EXPORT Procedure

Program
proc export data=sashelp.class
dbms=excelcs
outfile=c:\Myfiles\demo.xls;
sheet=Class;
server=sales;
port= 4632;
version=2000;
run;

Program

417

418

419

CHAPTER

21
The FONTREG Procedure
Overview: FONTREG Procedure 419
Syntax: FONTREG Procedure 419
PROC FONTREG Statement 420
FONTFILE Statement 421
FONTPATH Statement 421
TRUETYPE Statement 422
TYPE1 Statement (Experimental) 422
Concepts: FONTREG Procedure 423
Supported Font Types and Font Naming Conventions 423
Removing Fonts from the SAS Registry 424
Modifying SAS/GRAPH Device Drivers to Use System Fonts 425
Examples: FONTREG Procedure 425
Example 1: Adding a Single Font File 425
Example 2: Adding All Font Files from Multiple Directories 426
Example 3: Replacing Existing TrueType Font Files from a Directory

427

Overview: FONTREG Procedure

The FONTREG procedure enables you to update the SAS registry to include system
fonts, which can then be used in SAS output. PROC FONTREG uses FreeType
technology to recognize and incorporate various types of font denitions. Fonts of any
type that can be incorporated and used by SAS are known collectively in this
documentation as FreeType fonts.
Note: Including a system font in the SAS registry means that SAS knows where to
nd the font le. The font le is not actually used until the font is called for in a SAS
program. Therefore, do not move or delete font les after you have included the fonts in
the SAS registry. 4

Syntax: FONTREG Procedure

Interaction: If no statements are specied, then PROC FONTREG searches for TrueType
font les in the directory that is indicated in the FONTSLOC= SAS system option.
Tip: If more than one statement is specied, then the statements are executed in the
order in which they appear. You can use the same statement more than once in a single
PROC FONTREG step.
See

FONTREG Procedure in SAS Companion for z/OS

420

PROC FONTREG Statement

Chapter 21

PROC FONTREG < option(s)>;

FONTFILE le <le>;
FONTPATH directory <directory>;
TRUETYPE directory <directory>;
TYPE1 directory <directory>;

Operating Environment Information: For z/OS sites that do not use the hierarchical
le system (HFS), only the FONTFILE statement is supported. See FONTREG
Procedure in SAS Companion for z/OS for details. 4

PROC FONTREG Statement

PROC FONTREG <option(s)>;

Options
MODE=ADD | REPLACE | ALL

species how to handle new and existing fonts in the SAS registry:
ADD
add fonts that do not already exist in the SAS registry. Do not modify existing
fonts.
REPLACE
replace fonts that already exist in the SAS registry. Do not add new fonts.
ALL
add new fonts that do not already exist in the SAS registry and replace fonts that
already exist in the SAS registry.
Default: ADD
Featured in:

Example 3 on page 427

MSGLEVEL=VERBOSE | NORMAL | TERSE | NONE

species the level of detail to include in the SAS log:

VERBOSE
SAS log messages include which fonts were added, which fonts were not added,
and which fonts were not understood, as well as a summary that indicates the
number of fonts that were added, not added, and not understood.
NORMAL
SAS log messages include which fonts were added, and a summary that indicates
the number of fonts that were added, not added, and not understood.
TERSE
SAS log messages include only the summary that indicates the number of fonts
that were added, not added, and not understood.
NONE
No messages are written to the SAS log, except for errors (if encountered).

The FONTREG Procedure

FONTPATH Statement

421

Default: TERSE
Featured in:

Example 2 on page 426

NOUPDATE

species that the procedure should run without actually updating the SAS registry.
This option enables you to test the procedure on the specied fonts before modifying
the SAS registry.
USESASHELP

species that the SAS registry in the SASHELP library should be updated. You must
have write access to the SASHELP library in order to use this option. If the
USESASHELP option is not specied, then the SAS registry in the SASUSER library
is updated.

FONTFILE Statement
Species one or more font les to be processed.
Featured in: Example 1 on page 425

FONTFILE le <le>;

Argument
le

is the complete pathname to a font le. If the le is recognized as a valid font le,
then the le is processed. Each pathname must be enclosed in quotation marks. If
you specify more than one pathname, then you must separate the pathnames with a
space.

FONTPATH Statement
Species one or more directories to be searched for valid font les to process.
Featured in: Example 2 on page 426

FONTPATH directory <directory>;

422

TRUETYPE Statement

Chapter 21

Argument
directory

species a directory to search. All les that are recognized as valid font les are
processed. Each directory must be enclosed in quotation marks. If you specify more
than one directory, then you must separate the directories with a space.

TRUETYPE Statement
Species one or more directories to be searched for TrueType font les.
Featured in:

Example 3 on page 427

TRUETYPE directory <directory>;

Argument
directory

species a directory to search. Only les that are recognized as valid TrueType font
les are processed. Each directory must be enclosed in quotation marks. If you
specify more than one directory, then you must separate the directories with a space.

TYPE1 Statement (Experimental)

Species one or more directories to be searched for valid Type 1 font les.
TYPE1 directory <directory>;

CAUTION:

TYPE1 is an experimental statement that is available in SAS 9.1. Do not use this
statement in production jobs. 4

Argument
directory

species a directory to search. Only les that are recognized as valid Type 1 font les
are processed. Each directory must be enclosed in quotation marks. If you specify
more than one directory, then you must separate the directories with a space.

The FONTREG Procedure

Supported Font Types and Font Naming Conventions

423

Concepts: FONTREG Procedure

Supported Font Types and Font Naming Conventions

When a font is added to the SAS registry, the font name is prexed with a
three-character tag, enclosed in angle brackets (< >), that indicates the font type. For
example, if you add the TrueType font Arial to the SAS registry, then the name in the
registry is <ttf> Arial. This naming convention enables you to add and distinguish
between fonts that have the same name but are of different types. When you specify a
font in a SAS program (for example, in the TEMPLATE procedure or in the STYLE=
option in the REPORT procedure), use the tag to distinguish between fonts that have
the same name:
proc report data=grocery nowd
style(header)=[font_face=<ttf> Palatino Linotype];
run;

If you do not include a tag in your font specication, then SAS searches the registry
for fonts with that name. If more than one font with that name is encountered, then
SAS uses the one that has the highest rank in the following table.
Table 21.1

Supported Font Types

Rank

Type

Tag

File extension(s)

TrueType

<ttf>

.ttf

Type1

<at1>

.pfa
.pfb

PFR

<pfr>

.pfr

CAUTION:

Support for the Type1 and PFR font types is experimental in SAS 9.1. Do not use these
fonts in production jobs. 4
Note: SAS does not support nonscalable FreeType fonts of any type. Even if they
are recognized as valid FreeType fonts, they will not be added to the SAS registry. 4
Font les that are not produced by major vendors can be unreliable, and in some
cases SAS might not be able to use them.
The following SAS output methods and device drivers can use FreeType fonts:

3
3
3
3
3
3
3
3
3

SAS/GRAPH GIF, GIF733, GIFANIM

SAS/GRAPH JPEG
SAS/GRAPH PNG
SAS/GRAPH SASEMF
SAS/GRAPH SASWMF
SAS/GRAPH TIFFP, TIFFB
Universal Printing GIF
Universal Printing PCL
Universal Printing PDF.

424

Removing Fonts from the SAS Registry

Chapter 21

Removing Fonts from the SAS Registry

There are two ways to remove a font from the SAS registry:
3 by using the SAS Registry Editor
3 by using the REGISTRY procedure.
To remove a font by using the SAS Registry Editor, select
Solutions

Accessories

Registry Editor

(Alternatively, you can type regedit in the command window or Command ===>
prompt.)
Display 21.1

SAS Registry Editor

In the left pane of the Registry Editor window, navigate to the

[CORE\PRINTING\FREETYPE\FONTS] key. Select the font that you want to delete,
and use one of these methods to delete it:
3 Right-click the font name and select Delete.

3 Select the Delete button

3 Select
Edit

Delete

Key

To delete a font by using PROC REGISTRY, submit a program similar to the

following example. This example removes the <ttf> Arial font.
/* Write the key name for the font to an external file */
proc registry export=external-filename
startat=core\printing\freetype\fonts\<ttf> Arial;
run;
/* Remove the "<ttf> Arial" font from the SAS registry */
proc registry uninstall=external-filename fullstatus;
run;

For more information about PROC REGISTRY, see Chapter 41, The REGISTRY
Procedure, on page 831.

The FONTREG Procedure

Example 1: Adding a Single Font File

425

Modifying SAS/GRAPH Device Drivers to Use System Fonts

To access FreeType fonts with the SAS/GRAPH device drivers, the CHARREC eld of
the device driver entry must be modied from its default value of DMS Font to any
FreeType font. It is recommended that you use the <ttf> SAS Monospace font for this
purpose, because it is shipped with SAS and is always available in the SAS registry.
Changing the CHARREC value in this way enables you to use any FreeType font in
your SAS/GRAPH programs.
Here is an example that shows how to modify the CHARREC eld:
/* Assign a location for the personal devices catalog */
libname gdevice0 .;
/* Create a new GIF device driver, FTGIF, */
/* that will recognize FreeType fonts
*/
proc gdevice nofs c=gdevice0.devices;
copy GIF from=sashelp.devices newname=FTGIF;
mod FTGIF charrec=(0, 1, 1, <ttf> SAS Monospace, Y);
end;

The following device drivers can be modied to recognize FreeType fonts:

3
3
3
3

GIF, GIF733, GIFANIM

JPEG
PNG
TIFFP, TIFFB.

The SASWMF and SASEMF device drivers do not require this change.
For more information about SAS/GRAPH device drivers and the GDEVICE
procedure, see SAS/GRAPH Reference, Volumes 1 and 2.

Examples: FONTREG Procedure

Example 1: Adding a Single Font File

Procedure features:

FONTFILE statement

This example shows how to add a single font le to the SAS registry.

426

Program

Chapter 21

Program

Specify a font le to add. The FONTFILE statement species the complete path to a single
font le.
proc fontreg;
fontfile your-font-file;
run;

Log
NOTE: PROCEDURE PRINTTO used (Total process time):
real time
0.03 seconds
cpu time
0.00 seconds
20
proc fontreg;
21
fontfile your-font-file;
22
run;
SUMMARY:
Files processed: 1
Unusable files: 0
Files identified as fonts: 1
Fonts that were processed: 1
Fonts replaced in the SAS registry: 0
Fonts added to the SAS registry: 1
Fonts that could not be used: 0
NOTE: PROCEDURE FONTREG used (Total process time):
real time
0.17 seconds
cpu time
0.03 seconds

Example 2: Adding All Font Files from Multiple Directories

Procedure features:

MSGLEVEL= option
FONTPATH statement

This example shows how to add all valid font les from two different directories and
how to write detailed information to the SAS log.

Program

Write complete details to the SAS log. The MSGLEVEL=VERBOSE option writes complete
details about what fonts were added, what fonts were not added, and what font les were not
understood.
proc fontreg msglevel=verbose;

The FONTREG Procedure

Example 3: Replacing Existing TrueType Font Files from a Directory

427

Specify the directories to search for valid fonts. You can specify more than one directory in
the FONTPATH statement. Each directory must be enclosed in quotation marks. If you specify
more than one directory, then you must separate the directories with a space.
fontpath your-font-directory-1 your-font-directory-2;
run;

Log (Partial)
NOTE: PROCEDURE PRINTTO used (Total process time):
real time
0.03 seconds
cpu time
0.00 seconds
34
35
36
37

proc fontreg msglevel=verbose;

fontpath your-font-directory-1
your-font-directory-2;
run;

ERROR: FreeType base module FT_New_Face -- unknown file format.

WARNING: A problem was encountered with file
your-font-directory-2\SCRIPT.FON.
NOTE: The font Albertus Extra Bold (Style: Regular, Weight: Bold) has been
added to the SAS Registry at [CORE\PRINTING\FREETYPE\FONTS\<ttf>
Albertus Extra Bold]. Since it is a TRUETYPE font, it must be
referenced as <ttf> Albertus Extra Bold in SAS. The font resides in
file
your-font-directory-1\albr85w.ttf.
. . . more log entries . . .
SUMMARY:
Files processed: 138
Unusable files: 4
Files identified as fonts: 134
Fonts that were processed: 134
Fonts replaced in the SAS registry: 0
Fonts added to the SAS registry: 127
Fonts that could not be used: 7
NOTE: PROCEDURE FONTREG used (Total process time):
real time
7.11 seconds
cpu time
3.80 seconds

Example 3: Replacing Existing TrueType Font Files from a Directory

Procedure features:

MODE= option
TRUETYPE statement

This example reads all the TrueType Fonts in the specied directory and replaces
those that already exist in the SAS registry.

428

Program

Chapter 21

Program

Replace existing fonts only. The MODE=REPLACE option limits the action of the procedure
to replacing fonts that are already dened in the SAS registry. New fonts will not be added.
proc fontreg mode=replace;

Specify a directory that contains TrueType font les. Files in the directory that are not
recognized as being TrueType font les are ignored.
truetype your-font-directory;
run;

Log
53
proc fontreg mode=replace;
54
truetype your-font-directory;
55
run;
SUMMARY:
Files processed: 49
Unusable files: 4
Files identified as fonts: 45
Fonts that were processed: 39
Fonts replaced in the SAS registry: 39
Fonts added to the SAS registry: 0
Fonts that could not be used: 0
NOTE: PROCEDURE FONTREG used (Total process time):
real time
1.39 seconds
cpu time
0.63 seconds

See Also
3 The GDEVICE procedure in SAS/GRAPH Reference, Volumes 1 and 2
3 The FONTSLOC and SYSPRINTFONT SAS system options in SAS Language
Reference: Dictionary

3 http://www.freetype.org for more information about the FreeType project.

429

CHAPTER

22
The FORMAT Procedure
Overview: FORMAT Procedure 430
What Does the FORMAT Procedure Do? 430
What Are Formats and Informats? 430
How Are Formats and Informats Associated with a Variable? 430
Syntax: FORMAT Procedure 431
PROC FORMAT Statement 432
EXCLUDE Statement 434
INVALUE Statement 435
PICTURE Statement 438
SELECT Statement 447
VALUE Statement 448
Informat and Format Options 451
Specifying Values or Ranges 453
Concepts: FORMAT Procedure 455
Associating Informats and Formats with Variables 455
Methods of Associating Informats and Formats with Variables 455
Tips 455
See Also 456
Storing Informats and Formats 456
Format Catalogs 456
Temporary Informats and Formats 456
Permanent Informats and Formats 456
Accessing Permanent Informats and Formats 457
Missing Formats and Informats 457
Results: FORMAT Procedure 458
Output Control Data Set 458
Input Control Data Set 460
Procedure Output 461
Examples: FORMAT Procedure 463
Example 1: Creating a Picture Format 464
Example 2: Creating a Format for Character Values 466
Example 3: Writing a Format for Dates Using a Standard SAS Format 468
Example 4: Converting Raw Character Data to Numeric Values 470
Example 5: Creating a Format from a Data Set 472
Example 6: Printing the Description of Informats and Formats 477
Example 7: Retrieving a Permanent Format 478
Example 8: Writing Ranges for Character Strings 480
Example 9: Filling a Picture Format 483

430

Overview: FORMAT Procedure

Chapter 22

Overview: FORMAT Procedure

What Does the FORMAT Procedure Do?
The FORMAT procedure enables you to dene your own informats and formats for
variables. In addition, you can print the parts of a catalog that contain informats or
formats, store descriptions of informats or formats in a SAS data set, and use a SAS
data set to create informats or formats.

What Are Formats and Informats?

Informats determine how raw data values are read and stored. Formats determine
how variable values are printed. For simplicity, this section uses the terminology the
informat converts and the format prints.
Informats and formats tell the SAS System the datas type (character or numeric)
and form (such as how many bytes it occupies; decimal placement for numbers; how to
handle leading, trailing, or embedded blanks and zeros; and so forth). The SAS System
provides informats and formats for reading and writing variables. For a thorough
description of informats and formats that SAS provides, see the sections on formats and
informats in SAS Language Reference: Dictionary.
With informats, you can
3 convert a number to a character string (for example, convert 1 to YES)
3 convert a character string to a different character string (for example, convert
YES to OUI)
3 convert a character string to a number (for example, convert YES to 1)

3 convert a number to another number (for example, convert 0 through 9 to 1, 10

through 100 to 2, and so forth.
With formats, you can
3 print numeric values as character values (for example, print 1 as MALE and 2 as
FEMALE)
3 print one character string as a different character string (for example, print YES as
OUI)

3 print numeric values using a template (for example, print 9458763450 as

945-876-3450).

How Are Formats and Informats Associated with a Variable?

The following gure summarizes what occurs when you associate an informat and
format with a variable. The COMMAw.d informat and the DOLLARw.d format are
provided by SAS.

The FORMAT Procedure

Display 22.1

Syntax: FORMAT Procedure

431

Associating an Informat and a Format with a Variable

raw data value

$1,544.32

read with
COMMA9.2
informat

converted value

1544.32

printed using
DOLLAR9.2
format

printed value

$1,544.32

In the gure, SAS reads the raw data value that contains the dollar sign and comma.
The COMMA9.2 informat ignores the dollar sign and comma and converts the value to
1544.32. The DOLLAR9.2 format prints the value, adding the dollar sign and comma.
For more information about associating informats and formats with variables, see
Associating Informats and Formats with Variables on page 455.

Syntax: FORMAT Procedure

Restriction: You cannot use a SELECT statement and an EXCLUDE statement within
the same PROC FORMAT step.
Reminder: You can also use appropriate global statements with this procedure. See

Global Statements on page 18 for a list.

See:

FORMAT Procedure in the documentation for your operating environment.

PROC FORMAT < option(s)>;

EXCLUDE entry(s);
INVALUE <$>name < (informat-option(s))>
value-range-set(s);
PICTURE name <(format-option(s))>
value-range-set-1 <(picture-1-option(s) )>
<value-range-set-n <(picture-n-option(s))>>;
SELECT entry(s);
VALUE <$>name < (format-option(s))>
value-range-set(s);

432

PROC FORMAT Statement

Chapter 22

To do this

Use this statement

Exclude catalog entries from processing by the FMTLIB and

CNTLOUT= options

EXCLUDE

Create an informat for reading and converting raw data values

INVALUE

Create a template for printing numbers

PICTURE

Select catalog entries from processing by the FMTLIB and

CNTLOUT= options

SELECT

Create a format that species character strings to use to print

variable values

VALUE

PROC FORMAT Statement

You can use data set options with the CNTLIN= and CNTLOUT= data set
options. See Section 2, "Fundamental Concepts for Using Base SAS Procedures," for a
list.

Reminder:

PROC FORMAT <option(s)>;

To do this

Use this option

Specify a SAS data set from which PROC FORMAT builds

informats or formats

CNTLIN=

Create a SAS data set that stores information about informats or

formats

CNTLOUT=

Print information about informats or formats

FMTLIB

Specify a SAS library or catalog that will contain the informats or

formats that you are creating in the PROC FORMAT step

LIBRARY=

Specify the number of characters of the informatted or formatted

value that appear in PROC FORMAT output

MAXLABLEN=

Specify the number of characters of the start and end values that
appear in the PROC FORMAT output

MAXSELEN=

Prevent a new informat or format from replacing an existing one

of the same name

NOREPLACE

Print information about each format and informat on a separate

page1

PAGE

1 Used in conjunction with FMTLIB. If PAGE is specied, FMTLIB is invoked (or assumed).

The FORMAT Procedure

PROC FORMAT Statement

433

Options
CNTLIN=input-control-SAS-data-set

species a SAS data set from which PROC FORMAT builds informats and formats.
CNTLIN= builds formats and informats without using a VALUE, PICTURE, or
INVALUE statement. If you specify a one-level name, then the procedure searches
only the default data library (either the WORK data library or USER data library)
for the data set, regardless of whether you specify the LIBRARY= option.
Note: LIBRARY= can point to either a data library or a catalog. If only a libref is
specied, a catalog name of FORMATS is assumed. 4
A common source for an input control data set is the output from the
CNTLOUT= option of another PROC FORMAT step.
See also: Input Control Data Set on page 460
Tip:

Featured in:

Example 5 on page 472

CNTLOUT=output-control-SAS-data-set

creates a SAS data set that stores information about informats and formats that are
contained in the catalog specied in the LIBRARY= option.
Note: LIBRARY= can point to either a data library or a catalog. If only a libref is
specied, then a catalog name of FORMATS is assumed. 4
If you are creating an informat or format in the same step that the CNTLOUT=
option appears, then the informat or format that you are creating is included in the
CNTLOUT= data set.
If you specify a one-level name, then the procedure stores the data set in the
default data library (either the WORK data library or the USER data library),
regardless of whether you specify the LIBRARY= option.
Tip: You can use an output control data set as an input control data set in
subsequent PROC FORMAT steps.
See also: Output Control Data Set on page 458
FMTLIB

prints information about all the informats and formats in the catalog that is specied
in the LIBRARY= option. To get information only about specic informats or formats,
subset the catalog using the SELECT or EXCLUDE statement.
Interaction: The PAGE option invokes FMTLIB.

If your output from FMTLIB is not formatted correctly, then try increasing the
value of the LINESIZE= system option.
Tip: If you use the SELECT or EXCLUDE statement and omit the FMTLIB and
CNTLOUT= options, then the procedure invokes the FMTLIB option and you
receive FMTLIB option output.
Tip:

Featured in:

Example 6 on page 477

LIBRARY=libref<.catalog>

species a catalog to contain informats or formats that you are creating in the current
PROC FORMAT step. The procedure stores these informats and formats in the
catalog that you specify so that you can use them in subsequent SAS sessions or jobs.
Note: LIBRARY= can point to either a data library or a catalog. If only a libref is
specied, then a catalog name of FORMATS is assumed. 4
Alias:

LIB=

Default: If you omit the LIBRARY= option, then formats and informats are stored

in the WORK.FORMATS catalog. If you specify the LIBRARY= option but do not

434

EXCLUDE Statement

Chapter 22

specify a name for catalog, then formats and informats are stored in the
libref.FORMATS catalog.
SAS automatically searches LIBRARY.FORMATS. You might want to use the
LIBRARY libref for your format catalog. You can control the order in which SAS
searches for format catalogs with the FMTSEARCH= system option. For further
information about FMTSEARCH=, see the section on SAS system options in SAS
Language Reference: Dictionary.

Tip:

Example 1 on page 464

MAXLABLEN=number-of-characters

species the number of characters in the informatted or formatted value that you
want to appear in the CNTLOUT= data set or in the output of the FMTLIB option.
The FMTLIB option prints a maximum of 40 characters for the informatted or
formatted value.
MAXSELEN=number-of-characters

species the number of characters in the start and end values that you want to
appear in the CNTLOUT= data set or in the output of the FMTLIB option. The
FMTLIB option prints a maximum of 16 characters for start and end values.
NOREPLACE

prevents a new informat or format that you are creating from replacing an existing
informat or format of the same name. If you omit NOREPLACE, then the procedure
warns you that the informat or format already exists and replaces it.
Note:

You can have a format and an informat of the same name.

PAGE

prints information about each format and informat (that is, each entry) in the catalog
on a separate page.
Tip:

The PAGE option activates the FMTLIB option.

EXCLUDE Statement
Excludes entries from processing by the FMTLIB and CNTLOUT= options.
Restriction:

Only one EXCLUDE statement can appear in a PROC FORMAT step.

Restriction: You cannot use a SELECT statement and an EXCLUDE statement within
the same PROC FORMAT step.

EXCLUDE entry(s);

Required Arguments
entry(s)

species one or more catalog entries to exclude from processing. Catalog entry names
are the same as the name of the informat or format that they store. Because
informats and formats can have the same name, and because character and numeric
informats or formats can have the same name, you must use certain prexes when

The FORMAT Procedure

INVALUE Statement

435

specifying informats and formats in the EXCLUDE statement. Follow these rules
when specifying entries in the EXCLUDE statement:
3 Precede names of entries that contain character formats with a dollar sign ($).
3 Precede names of entries that contain character informats with an at sign and a
dollar sign (for example, @$entry-name).
3 Precede names of entries that contain numeric informats with an at sign (@).
3 Specify names of entries that contain numeric formats without a prex.

Shortcuts to Specifying Names

You can use the colon (:) and hyphen (-) wildcard characters to exclude entries. For
example, the following EXCLUDE statement excludes all formats or informats that
begin with the letter a.
exclude a:;

In addition, the following EXCLUDE statement excludes all formats or informats

that occur alphabetically between apple and pear, inclusive:
exclude apple-pear;

FMTLIB Output
If you use the EXCLUDE statement without either FMTLIB or CNTLOUT= in the
PROC FORMAT statement, then the procedure invokes FMTLIB.

INVALUE Statement
Creates an informat for reading and converting raw data values.
Featured in: Example 4 on page 470.
See also: The section on informats in SAS Language Reference: Dictionary for
documentation on informats supplied by SAS.

INVALUE <$>name <(informat-option(s))>

<value-range-set(s)>;

To do this

Use this option

Specify the default length of the informat

DEFAULT=

Specify a fuzz factor for matching values to a range

FUZZ=

Specify a maximum length for the informat

MAX=

Specify a minimum length for the informat

MIN=

Store values or ranges in the order that you dene them

NOTSORTED

Left-justify all input strings before they are compared to ranges

JUST

Uppercase all input strings before they are compared to ranges

UPCASE

436

INVALUE Statement

Chapter 22

Required Arguments
name

names the informat that you are creating.

The name must be a valid SAS name. A numeric informat name can
be up to 31 characters in length; a character informat name can be up to 30
characters in length and cannot end in a number. If you are creating a character
informat, then use a dollar sign ($) as the rst character; this is why a character
informat is limited to 30 characters.

Requirement:

Restriction: A user-dened informat name cannot be the same as an informat

name that is supplied by SAS.

Interaction: The maximum length of an informat name is controlled by the

VALIDFMTNAME= SAS system option. See SAS Language Reference: Dictionary

for details on VALIDFMTNAME=.
Refer to the informat later by using the name followed by a period. However,
do not use a period after the informat name in the INVALUE statement.

Tip:

When SAS prints messages that refer to a user-written informat, the name is
prexed by an at sign (@). When the informat is stored, the at sign is prexed to
the name that you specify for the informat; this is why the name is limited to 31
or 30 characters. You need to use the at sign only when you are using the name in
an EXCLUDE or SELECT statement; do not prex the name with an at sign when
you are associating the informat with a variable.

Tip:

Options
The following options are common to the INVALUE, PICTURE, and VALUE
statements and are described in Informat and Format Options on page 451:
DEFAULT=length
FUZZ= fuzz-factor
MAX=length
MIN=length
NOTSORTED
In addition, you can use the following options:
JUST

left-justies all input strings before they are compared to the ranges.
UPCASE

converts all raw data values to uppercase before they are compared to the possible
ranges. If you use UPCASE, then make sure the values or ranges you specify are in
uppercase.
value-range-set(s)

species raw data and values that the raw data will become. The value-range-set(s)
can be one or more of the following:
value-or-range-1 <, value-or-range-n>=informatted-value|[existing-informat]
The informat converts the raw data to the values of informatted-value on the right
side of the equal sign.
informatted-value
is the value you want the raw data in value-or-range to become. Use one of the
following forms for informatted-value:

The FORMAT Procedure

INVALUE Statement

437

character-string
is a character string up to 32,767 characters long. Typically, character-string
becomes the value of a character variable when you use the informat to convert
raw data. Use character-string for informatted-value only when you are creating
a character informat. If you omit the single or double quotation marks around
character-string, then the INVALUE statement assumes that the quotation
marks are there.
For hexadecimal literals, you can use up to 32,767 typed characters, or up to
16,382 represented characters at 2 hexadecimal characters per represented
character.
number
is a number that becomes the informatted value. Typically, number becomes the
value of a numeric variable when you use the informat to convert raw data. Use
number for informatted-value when you are creating a numeric informat. The
maximum for number depends on the host operating environment.
_ERROR_
treats data values in the designated range as invalid data. SAS assigns a
missing value to the variable, prints the data line in the SAS log, and issues a
warning message.
_SAME_
prevents the informat from converting the raw data as any other value. For
example, the following GROUP. informat converts values 01 through 20 and
assigns the numbers 1 through 20 as the result. All other values are assigned a
missing value.
invalue group 01-20= _same_
other= .;

existing-informat
is an informat that is supplied by SAS or a user-dened informat. The informat
you are creating uses the existing informat to convert the raw data that match
value-or-range on the left side of the equals sign. If you use an existing informat,
then enclose the informat name in square brackets (for example, [date9.]) or with
parentheses and vertical bars, for example, (|date9.|). Do not enclose the name of
the existing informat in single quotation marks.
value-or-range
See Specifying Values or Ranges on page 453.
Consider the following examples:
3 The $GENDER. character informat converts the raw data values F and M to
character values 1 and 2:
invalue $gender F=1
M=2;

The dollar sign prex indicates that the informat converts character data.
3 When you are creating numeric informats, you can specify character strings or
numbers for value-or-range. For example, the TRIAL. informat converts any
character string that sorts between A and M to the number 1 and any character
string that sorts between N and Z to the number 2. The informat treats the
unquoted range 13000 as a numeric range, which includes all numeric values
between 1 and 3000:
invalue trial A-M=1
N-Z=2
1-3000=3;

438

PICTURE Statement

Chapter 22

If you use a numeric informat to convert character strings that do not

correspond to any values or ranges, then you receive an error message.

3 The CHECK. informat uses _ERROR_ and _SAME_ to convert values of 1

through 4 and 99. All other values are invalid:
invalue check 1-4=_same_
99=.
other=_error_;

PICTURE Statement
Creates a template for printing numbers.
Featured in:

Example 1 on page 464 and Example 9 on page 483

The section on formats in SAS Language Reference: Dictionary for

documentation on formats supplied by SAS.

See also:

PICTURE name <(format-option(s))>

<value-range-set-1 <(picture-1-option(s) )>
<value-range-set-n <(picture-n-option(s))>>>;

To do this

Use this option

Control the attributes of the format

Specify that you can use directives in the picture as a template
to format date, time, or datetime values

DATATYPE=

Specify the default length of the format

DEFAULT=

Specify the separator character for the fractional part of a

number

DECSEP=

Specify the three-digit separator character for a number

DIG3SEP=

Specify a fuzz factor for matching values to a range

FUZZ=

Specify a maximum length for the format

MAX=

Specify a minimum length for the format

MIN=

Specify multiple pictures for a given value or range and for

overlapping ranges

MULTILABEL

Store values or ranges in the order that you dene them

NOTSORTED

Round the value to the nearest integer before formatting

ROUND

Control the attributes of each picture in the format

Specify a character that completes the formatted value

FILL=

Specify a number to multiply the variables value by before it

is formatted

MULTIPLIER=

The FORMAT Procedure

To do this

PICTURE Statement

439

Use this option

Specify that numbers are message characters rather than digit

selectors

NOEDIT

Specify a character prex for the formatted value

PREFIX=

Required Arguments
name

names the format you are creating.

The name must be a valid SAS name. A numeric format name can
be up to 32 characters in length; a character format name can be up to 31
characters in length, not ending in a number. If you are creating a character
format, then use a dollar sign ($) as the rst character, which is why a character
informat is limited to 30 characters.

Requirement:

Restriction: A user-dened format cannot be the name of a format supplied by SAS.

Interaction: The maximum length of a format name is controlled by the

VALIDFMTNAME= SAS system option. See SAS Language Reference: Dictionary

for details on VALIDFMTNAME=.
Refer to the format later by using the name followed by a period. However, do
not put a period after the format name in the VALUE statement.

Tip:

Options
The following options are common to the INVALUE, PICTURE, and VALUE
statements and are described in Informat and Format Options on page 451:
DEFAULT= length
FUZZ=fuzz-factor
MAX=length
MIN=length
NOTSORTED
In addition, you can use the following arguments:
DATATYPE=DATE | TIME | DATETIME

species that you can use directives in the picture as a template to format date, time,
or datetime values. See the denition and list of directives on page 442.
If you format a numeric missing value, then the resulting label will be ERROR.
Adding a clause to your program that checks for missing values can eliminate the
ERROR label.

Tip:

DECSEP=character

species the separator character for the fractional part of a number.

Default: . (a decimal point)
DIG3SEP=character

species the three-digit separator character for a number.

Default: , (a comma)

440

PICTURE Statement

Chapter 22

FILL=character

species a character that completes the formatted value. If the number of signicant
digits is less than the length of the format, then the format must complete, or ll, the
formatted value:
3 The format uses character to ll the formatted value if you specify zeros as digit
selectors.
3 The format uses zeros to ll the formatted value if you specify nonzero digit
selectors. The FILL= option has no effect.
If the picture includes other characters, such as a comma, which appear to the left
of the digit selector that maps to the last signicant digit placed, then the characters
are replaced by the ll character or leading zeros.
Default: (a blank)
Interaction: If you use the FILL= and PREFIX= options in the same picture, then
the format places the prex and then the ll characters.
Featured in: Example 9 on page 483
MULTILABEL

allows the assignment of multiple labels or external values to internal values. The
following PICTURE statements show the two uses of the MULTILABEL option. In
each case, number formats are assigned as labels. The rst PICTURE statement
assigns multiple labels to a single internal value. Multiple labels may also be
assigned to a single range of internal values. The second PICTURE statement
assigns labels to overlapping ranges of internal values. The MULTILABEL option
allows the assignment of multiple labels to the overlapped internal values.
picture abc (multilabel)
1000=9,999
1000=9999;
picture overlap (multilabel)
/* without decimals */
0-999=999
1000-9999=9,999
/* with decimals */
0-9=9.999
10-99=99.99
100-999=999.9;

Only multilabel-enabled procedures such as PROC MEANS, PROC SUMMARY, and

PROC TABULATE can use multiple labels. All other procedures and the DATA step
recognize only the primary label. The primary label for a given entry is the external
value that is assigned to the rst internal value or range of internal values that
matches or contains the entry when all internal values are ordered sequentially. For
example, in the rst PICTURE statement, the primary label for 1000 is 1,000
because the format 9,999 is the rst external value that is assigned to 1000. The
secondary label for 1000 is 1000, based on the 9999 format.
In the second PICTURE statement, the primary label for 5 is 5.000 based on the
9.999 format that is assigned to the range 09 because 09 is sequentially the rst
range of internal values containing 5. The secondary label for 5 is 005 because the
range 0999 occurs in sequence after the range 09. Consider carefully when you
assign multiple labels to an internal value. Unless you use the NOTSORTED option
when you assign variables, the SAS System stores the variables in sorted order. This
may produce unexpected results when variables with the MULTILABEL format are
processed. For example, in the second PICTURE statement, the primary label for 15

The FORMAT Procedure

PICTURE Statement

441

is 015, and the secondary label for 15 is 15.00 because the range 0999 occurs in
sequence before the range 1099. If you want the primary label for 15 to use the
99.99 format, then you might want to change the range 1099 to 099 in the
PICTURE statement. The range 099 occurs in sequence before the range 0999 and
will produce the desired result.
MULTIPLIER=n

species a number that the variables value is to be multiplied by before it is

formatted. For example, the following PICTURE statement creates the MILLION.
format, which formats the variable value 1600000 as $1.6M:
picture million low-high=00.0M
(prefix=$ mult=.00001);

MULT=

Alias:

Default: 10 , where n is the number of digits after the rst decimal point in the

picture. For example, suppose your data contains a value 123.456 and you want to
3
print it using a picture of 999.999. The format multiplies 123.456 by 10 to obtain
a value of 123456, which results in a formatted value of 123.456.
Example: Example 1 on page 464
NOEDIT

species that numbers are message characters rather than digit selectors; that is, the
format prints the numbers as they appear in the picture. For example, the following
PICTURE statement creates the MILES. format, which formats any variable value
greater than 1000 as >1000 miles:
picture miles 1-1000=0000
1000<-high=>1000 miles(noedit);

PREFIX=prex

species a character prex to place in front of the values rst signicant digit. You
must use zero digit selectors or the prex will not be used.
The picture must be wide enough to contain both the value and the prex. If the
picture is not wide enough to contain both the value and the prex, then the format
truncates or omits the prex. Typical uses for PREFIX= are printing leading
currency symbols and minus signs. For example, the PAY. format prints the variable
value 25500 as $25,500.00:
picture pay low-high=000,009.99
(prefix=$);

Default: no prex
Interaction: If you use the FILL= and PREFIX= options in the same picture, then

the format places the prex and then the ll characters.

Featured in:

Example 1 on page 464 and Example 9 on page 483

ROUND

rounds the value to the nearest integer before formatting. Without the ROUND
option, the format multiplies the variable value by the multiplier, truncates the
decimal portion (if any), and prints the result according to the template that you
dene. With the ROUND option, the format multiplies the variable value by the
multiplier, rounds that result to the nearest integer, and then formats the value
according to the template. Note that if the FUZZ= option is also specied, the
rounding takes place after SAS has used the fuzz factor to determine which range
the value belongs to.
Tip:

Note that the ROUND option rounds a value of .5 to the next highest integer.

442

PICTURE Statement

Chapter 22

value-range-set

species one or more variable values and a template for printing those values. The
value-range-set is the following:
value-or-range-1 <, value-or-range-n>=picture
picture
species a template for formatting values of numeric variables. The picture is a
sequence of characters in single quotation marks. The maximum length for a
picture is 40 characters. Pictures are specied with three types of characters: digit
selectors, message characters, and directives. You can have a maximum of 16 digit
selectors in a picture.
Digit selectors are numeric characters (0 through 9) that dene positions for
numeric values. A picture format with nonzero digit selectors prints any leading
zeros in variable values; picture digit selectors of 0 do not print leading zeros in
variable values. If the picture format contains digit selectors, then a digit selector
must be the rst character in the picture.
Note: This chapter uses 9s as nonzero digit selectors. 4
Message characters are nonnumeric characters that print as specied in the
picture. The following PICTURE statement contains both digit selectors (99) and
message characters (illegal day value). Because the DAYS. format has nonzero
digit selectors, values are printed with leading zeros. The special range OTHER
prints the message characters for any values that do not fall into the specied
range (1 through 31).
picture days 01-31=99
other=99-illegal day value;

For example, the values 02 and 67 print as

02
67-illegal day value

Directives are special characters that you can use in the picture to format date,
time, or datetime values.
Restriction: You can only use directives when you specify the DATATYPE= option

in the PICTURE statement.

The permitted directives are
%a

Locales abbreviated weekday name

Locales full weekday name

Locales abbreviated month name

Locales full month name

Day of the month as a decimal number (131), with no leading

zero

Hour (24-hour clock) as a decimal number (023), with no

leading zero

Hour (12-hour clock) as a decimal number (112), with no

leading zero

Day of the year as a decimal number (1366), with no leading

zero

Month as a decimal number (112), with no leading zero

Minute as a decimal number (059), with no leading zero

The FORMAT Procedure

PICTURE Statement

443

Locales equivalent of either AM or PM

Second as a decimal number (059), with no leading zero

Week number of the year (Sunday as the rst day of the week)
as a decimal number (0,53), with no leading zero

Weekday as a decimal number (1= Sunday, 7=Saturday)

Year without century as a decimal number (099), with no

leading zero

Year with century as a decimal number

%%
%
Any directive that generates numbers can produce a leading zero, if desired, by
adding a 0 before the directive. This applies to %d, %H, %I, %j, %m, %M, %S, %U,
and %y. For example, if you specify %y in the picture, then 2001 would be
formatted as 1, but if you specify %0y, then 2001 would be formatted as 01.
Tip: Add code to your program to direct how you want missing values to be
displayed.
value-or-range
See Specifying Values or Ranges on page 453.

Building a Picture Format: Step by Step

This section shows how to write a picture format for formatting numbers with
leading zeros. In the SAMPLE data set, the default printing of the variable Amount
has leading zeros on numbers between 1 and 1:
options nodate pageno=1 linesize=64 pagesize=60;
data sample;
input Amount;
datalines;
-2.051
-.05
-.017
0
.093
.54
.556
6.6
14.63
;

proc print data=sample;

title Default Printing of the Variable Amount;
run;

444

PICTURE Statement

Chapter 22

Default Printing of the Variable Amount

Obs

Amount

1
2
3
4
5
6
7
8
9

-2.051
-0.050
-0.017
0.000
0.093
0.540
0.556
6.600
14.630

The following PROC FORMAT step uses the ROUND format option and creates the
NOZEROS. format, which eliminates leading zeros in the formatted values:
libname library SAS-data-library;

proc format library=library;

picture nozeros (round)
low - -1 = 00.00
(prefix=-)
-1 <-< 0 =
99
(prefix=-. mult=100)
0 -< 1
=
99
(prefix=. mult=100)
1 - high = 00.00;
run;

The following table explains how one value from each range is formatted. Figure 22.1
on page 446 provides an illustration of each step. The circled numbers in the gure
correspond to the step numbers in the table.

Table 22.1

Building a Picture Format

Step

Rule

In this example

Determine into which range the value falls and

use that picture.

In the second range, the exclusion operator <

appears on both sides of the hyphen and excludes 1
and 0 from the range.

Take the absolute value of the numeric value.

Because the absolute value is used, you need a

separate range and picture for the negative
numbers in order to prex the minus sign.

The FORMAT Procedure

PICTURE Statement

Step

Rule

In this example

Multiply the number by the MULT= value. If

you do not specify the MULT= option, then the
PICTURE statement uses the default. The
n
default is 10 , where n is the number of digit
selectors to the right of the decimal1 in the
picture. (Step 6 discusses digit selectors further.)

Specifying a MULT= value is necessary for numbers

between 0 and 1 and numbers between 0 and 1
because no decimal appears in the pictures for those
ranges. Because MULT= defaults to 1, truncation of
the signicant digits results without a MULT=
value specied. (Truncation is explained in the next
step.) For the two ranges that do not have MULT=
values specied, the MULT= value defaults to 100
because the corresponding picture has two digit
selectors to the right of the decimal. After the
MULT= value is applied, all signicant digits are
moved to the left of the decimal.

Truncate the number after the decimal. If the

ROUND option is in effect, then the format
rounds the number after the decimal to the next
highest integer if the number after the decimal
is greater than or equal to .5.

Because the example uses MULT= values that

ensured that all of the signicant digits were moved
to the left of the decimal, no signicant digits are
lost. The zeros are truncated.

Turn the number into a character string. If the

number is shorter than the picture, then the
length of the character string is equal to the
number of digit selectors in the picture. Pad the
character string with leading zeros. (The results
are equivalent to using the Zw. format. Zw. is
explained in the section on SAS formats in SAS
Language Reference: Dictionary.

The numbers 205, 5, and 660 become the character

strings 0205, 05, and 0660, respectively. Because
each picture is longer than the numbers, the format
adds a leading zero to each value. The format does
not add leading zeros to the number 55 because the
corresponding picture only has two digit selectors.

445

446

PICTURE Statement

Chapter 22

Step

Rule

In this example

Apply the character string to the picture. The

format only maps the rightmost n characters in
the character string, where n is the number of
digit selectors in the picture. Thus, it is
important to make sure that the picture has
enough digit selectors to accommodate the
characters in the string. After the format takes
the rightmost n characters, it then maps those
characters to the picture from left to right.
Choosing a zero or nonzero digit selector is
important if the character string contains
leading zeros. If one of the leading zeros in the
character string maps to a nonzero digit
selector, then it and all subsequent leading zeros
become part of the formatted value. If all of the
leading zeros map to zero digit selectors, then
none of the leading zeros become part of the
formatted value; the format replaces the leading
zeros in the character string with blanks.2

The leading zero is dropped from each of the

character strings 0205 and 0660 because the
leading zero maps to a zero digit selector in the
picture.

Prex any characters that are specied in the

PREFIX= option. You need the PREFIX= option
because when a picture contains any digit
selectors, the picture must begin with a digit
selector. Thus, you cannot begin your picture
with a decimal point, minus sign, or any other
character that is not a digit selector.

The PREFIX= option reclaims the decimal point and

the negative sign, as shown with the formatted
values -.05 and .55.

1 A decimal in a PREFIX= option is not part of the picture.

2 You can use the FILL= option to specify a character other than a blank to become part of the formatted value.

Figure 22.1

Formatting One Value in Each Range

The FORMAT Procedure

SELECT Statement

447

The following PROC PRINT step associates the NOZEROS. format with the
AMOUNT variable in SAMPLE. The output shows the result of rounding.
proc print data=sample noobs;
format amount nozeros.;
title Formatting the Variable Amount;
title2 with the NOZEROS. Format;
run;

Formatting the Variable Amount

with the NOZEROS. Format

Amount
-2.05
-.05
-.02
.00
.09
.54
.56
6.60
14.63

CAUTION:

The picture must be wide enough for the prex and the numbers. In this example, if the
value 45.00 were formatted with NOZEROS. then the result would be 45.00 because
it falls into the rst range, low - 1, and the picture for that range is not wide
enough to accommodate the prexed minus sign and the number. 4

Specifying No Picture
This PICTURE statement creates a picture-name format that has no picture:
picture picture-name;

Using this format has the effect of applying the default SAS format to the values.

SELECT Statement
Selects entries from processing by the FMTLIB and CNTLOUT= options.
Restriction:

Only one SELECT statement can appear in a PROC FORMAT step.

Restriction: You cannot use a SELECT statement and an EXCLUDE statement within
the same PROC FORMAT step.
Featured in: Example 6 on page 477.

SELECT entry(s);

448

VALUE Statement

Chapter 22

Required Arguments
entry(s)

species one or more catalog entries for processing. Catalog entry names are the
same as the name of the informat or format that they store. Because informats and
formats can have the same name, and because character and numeric informats or
formats can have the same name, you must use certain prexes when specifying
informats and formats in the SELECT statement. Follow these rules when specifying
entries in the SELECT statement:
3 Precede names of entries that contain character formats with a dollar sign ($).
3 Precede names of entries that contain character informats with an at sign and a
dollar sign, for example, @$entry-name.
3 Precede names of entries that contain numeric informats with an at sign (@).
3 Specify names of entries that contain numeric formats without a prex.

Shortcuts to Specifying Names

You can use the colon (:) and hyphen (-) wildcard characters to select entries. For
example, the following SELECT statement selects all formats or informats that begin
with the letter a.
select a:;

In addition, the following SELECT statement selects all formats or informats that
occur alphabetically between apple and pear, inclusive:
select apple-pear;

FMTLIB Output
If you use the SELECT statement without either FMTLIB or CNTLOUT= in the
PROC FORMAT statement, then the procedure invokes FMTLIB.

VALUE Statement
Creates a format that species character strings to use to print variable values.
Featured in:

Example 2 on page 466.

See also: The chapter about formats in SAS Language Reference: Dictionary for

documentation about SAS formats.

VALUE < $>name <(format-option(s))>

<value-range-set(s)>;

To do this

Use this option

Specify the default length of the format

DEFAULT=

Specify a fuzz factor for matching values to a range

FUZZ=

Specify a maximum length for the format

MAX=

The FORMAT Procedure

VALUE Statement

To do this

Use this option

Specify a minimum length for the format

MIN=

Specify multiple values for a given range, or for overlapping ranges

MULTILABEL

Store values or ranges in the order that you dene them

449

NOTSORTED

Required Arguments
name

names the format that you are creating.

Requirement: The name must be a valid SAS name. A numeric format name can
be up to 32 characters in length. A character format name can be up to 31
characters in length and cannot end in a number. If you are creating a character
format, then use a dollar sign ($) as the rst character.
Restriction: The name of a user-dened format cannot be the same as the name of
a format that is supplied by SAS.
Interaction: The maximum length of a format name is controlled by the
VALIDFMTNAME= SAS system option. See SAS Language Reference: Dictionary
for details about VALIDFMTNAME=.
Tip: Refer to the format later by using the name followed by a period. However, do
not use a period after the format name in the VALUE statement.

allows the assignment of multiple labels or external values to internal values. The
following VALUE statements show the two uses of the MULTILABEL option. The
rst VALUE statement assigns multiple labels to a single internal value. Multiple
labels may also be assigned to a single range of internal values. The second VALUE
statement assigns labels to overlapping ranges of internal values. The MULTILABEL
option allows the assignment of multiple labels to the overlapped internal values.
value one (multilabel)
1=ONE
1=UNO
1=UN
value agefmt (multilabel)
15-29=below 30 years

450

VALUE Statement

Chapter 22

30-50=between 30 and 50
51-high=over 50 years
15-19=15 to 19
20-25=20 to 25
25-39=25 to 39
40-55=40 to 55
56-high=56 and above;

Only multilabel-enabled procedures such as PROC MEANS, PROC SUMMARY, and

PROC TABULATE can use multiple labels. All other procedures and the data step
recognize only the primary label. The primary label for a given entry is the external
value that is assigned to the rst internal value or range of internal values that
matches or contains the entry when all internal values are ordered sequentially. For
example, in the rst VALUE statement, the primary label for 1 is ONE because ONE
is the rst external value that is assigned to 1. The secondary labels for 1 are UNO
and UN. In the second VALUE statement, the primary label for 33 is 25 to 39
because the range 2539 is sequentially the rst range of internal values that
contains 33. The secondary label for 33 is between 30 and 50 because the range
3050 occurs in sequence after the range 2539.
value-range-set(s)

species one or more variable values and a character string or an existing format.
The value-range-set(s) can be one or more of the following:
value-or-range-1 <, value-or-range-n>=formatted-value|[existing-format]
The variable values on the left side of the equals sign print as the character string
on the right side of the equals sign.
formatted-value
species a character string that becomes the printed value of the variable value
that appears on the left side of the equals sign. Formatted values are always
character strings, regardless of whether you are creating a character or numeric
format.
Formatted values can be up to 32,767 characters. For hexadecimal literals, you
can use up to 32,767 typed characters, or up to 16,382 represented characters at 2
hexadecimal characters per represented character. Some procedures, however, use
only the rst 8 or 16 characters of a formatted value.
Requirement: You must enclose a formatted value in single or double quotation
marks. The following example shows a formatted value that is enclosed in
double quotation marks.
value $ score
M=Male "(pass)"
F=Female "(pass)";

Requirement: If a formatted value contains a single quotation mark, then enclose

the value in double quotation marks:

value sect 1="Smiths class"
2="Leungs class";

Tip: Formatting numeric variables does not preclude the use of those variables in

arithmetic operations. SAS uses stored values for arithmetic operations.

existing-format
species a format supplied by SAS or an existing user-dened format. The format
you are creating uses the existing format to convert the raw data that match
value-or-range on the left side of the equals sign.
If you use an existing format, then enclose the format name in square brackets
(for example, [date9.]) or with parentheses and vertical bars, for example,

The FORMAT Procedure

Informat and Format Options

451

(|date9.|). Do not enclose the name of the existing format in single quotation
marks.
Using an existing format can be thought of as nesting formats. A nested level of
one means that if you are creating the format A with the format B as a formatted
value, then the procedure has to use only one existing format to create A.
Tip: Avoid nesting formats more than one level. The resource requirements can

increase dramatically with each additional level.

value-or-range
For details on how to specify value-or-range, see Specifying Values or Ranges on
page 453.
Consider the following examples:

3 The $STATE. character format prints the postal code for selected states:
value $state Delaware=DE
Florida=FL
Ohio=OH;

The variable value Delaware prints as DE, the variable value Florida prints
as FL, and the variable value Ohio prints as OH. Note that the $STATE. format
begins with a dollar sign.
Note: Range specications are case sensitive. In the $STATE. format above,
the value OHIO would not match any of the specied ranges. If you are not
certain what case the data values are in, then one solution is to use the
UPCASE function on the data values and specify all uppercase characters for
the ranges. 4

3 The numeric format ANSWER.writes the values 1 and 2 as yes and no:
value answer 1=yes
2=no;

Specifying No Ranges
This VALUE statement creates a format-name format that has no ranges:
value format-name;

Using this format has the effect of applying the default SAS format to the values.

Informat and Format Options

This section discusses options that are valid in the INVALUE, PICTURE, and
VALUE statements. These options appear in parentheses after the informat or format
name. They affect the entire informat or format that you are creating.
DEFAULT=length
species the default length of the informat or format. The value for DEFAULT=
becomes the length of the informat or format if you do not give a specic length
when you associate the informat or format with a variable.
The default length of a format is the length of the longest formatted value.
The default length of an informat depends on whether the informat is character
or numeric. The default length of character informats is the length of the longest
informatted value. The default of a numeric informat is 12 if you have numeric
data to the left of the equals sign. If you have a quoted string to the left of the
equals sign, then the default length is the length of the longest string.

452

Informat and Format Options

Chapter 22

FUZZ=fuzz-factor
species a fuzz factor for matching values to a range. If a number does not match
or fall in a range exactly but comes within fuzz-factor, then the format considers it
a match. For example, the following VALUE statement creates the LEVELS.
format, which uses a fuzz factor of .2:
value levels (fuzz=.2) 1=A
2=B
3=C;

FUZZ=.2 means that if a variable value falls within .2 of a value on either end
of the range, then the format uses the corresponding formatted value to print the
variable value. So the LEVELS. format formats the value 2.1 as B.
If a variable value matches one value or range without the fuzz factor, and also
matches another value or range with the fuzz factor, then the format assigns the
variable value to the value or range that it matched without the fuzz factor.
Default: 1E12 for numeric formats and 0 for character formats.
Tip: Specify FUZZ=0 to save storage space when you use the VALUE statement
to create numeric formats.
Tip: A value that is excluded from a range using the < operator does not receive
the formatted value, even if it falls into the range when you use the fuzz factor.
MAX=length
species a maximum length for the informat or format. When you associate the
format with a variable, you cannot specify a width greater than the MAX= value.
Default: 40
Range: 140
MIN=length
species a minimum length for the informat or format.
Default: 1
Range: 140
NOTSORTED
stores values or ranges for informats or formats in the order in which you dene
them. If you do not specify NOTSORTED, then values or ranges are stored in
sorted order by default, and SAS uses a binary searching algorithm to locate the
range that a particular value falls into. If you specify NOTSORTED, then SAS
searches each range in the order in which you dene them until a match is found.
Use NOTSORTED if
3 you know the likelihood of certain ranges occurring, and you want your
informat or format to search those ranges rst to save processing time.
3 you want to preserve the order that you dene ranges when you print a
description of the informat or format using the FMTLIB option.
3 you want to preserve the order that you dene ranges when you use the
ORDER=DATA option and the PRELOADFMT option to analyze class
variables in PROC MEANS, PROC SUMMARY, or PROC TABULATE.
Do not use NOTSORTED if the distribution of values is uniform or unknown, or
if the number of values is relatively small. The binary searching algorithm that
SAS uses when NOTSORTED is not specied optimizes the performance of the
search under these conditions.
Note: SAS automatically sets the NOTSORTED option when you use the
CPORT and the CIMPORT procedures to transport informats or formats between

The FORMAT Procedure

Specifying Values or Ranges

453

operating environments with different standard collating sequences. This

automatic setting of NOTSORTED can occur when you transport informats or
formats between ASCII and EBCDIC operating environments. If this situation is
undesirable, then do the following:
1 Use the CNTLOUT= option in the PROC FORMAT statement to create an

output control data set.

2 Use the CPORT procedure to create a transport le for the control data set.
3 Use the CIMPORT procedure in the target operating environment to import

the transport le.

4 In the target operating environment, use PROC FORMAT with the CNTLIN=

option to build the formats and informats from the imported control data set.

Specifying Values or Ranges

As the syntax of the INVALUE, PICTURE, and VALUE statements indicates, you
must specify values as value-range-sets. On the left side of the equals sign you specify
the values that you want to convert to other values. On the right side of the equals
sign, you specify the values that you want the values on the left side to become. This
section discusses the different forms that you can use for value-or-range, which
represents the values on the left side of the equals sign. For details about how to
specify values for the right side of the equals sign, see the Required Arguments
section for the appropriate statement.
The INVALUE, PICTURE, and VALUE statements accept numeric values on the left
side of the equals sign. INVALUE and VALUE also accept character strings on the left
side of the equals sign.
As the syntax shows, you can have multiple occurrences of value-or-range in each
value-range-set, with commas separating the occurrences. Each occurrence of
value-or-range is either one of the following:
value
a single value, such as 12 or CA. For character formats and informats, enclose
the character values in single quotation marks. If you omit the quotation marks
around value, then PROC FORMAT assumes the quotation marks to be there.
You can use the keyword OTHER as a single value. OTHER matches all values
that do not match any other value or range.
range
a list of values, for example, 1268 or A-Z. For ranges with character strings,
be sure to enclose each string in single quotation marks. For example, if you want
a range that includes character strings from A to Z, then specify the range as
A-Z, with single quotation marks around the A and around the Z.
If you specify A-Z, then the procedure interprets it as a three-character string
with A as the rst character, a hyphen (-) as the second character, and a Z as the
third character.
If you omit the quotation marks, then the procedure assumes quotation marks
around each string. For example, if you specify the range abc-zzz, then the
procedure interprets it as abc-zzz.

454

Specifying Values or Ranges

Chapter 22

You can use LOW or HIGH as one value in a range, and you can use the range
LOW-HIGH to encompass all values. For example, these are valid ranges:
low-ZZ
35-high
low-high

You can use the less than (<) symbol to exclude values from ranges. If you are
excluding the rst value in a range, then put the < after the value. If you are
excluding the last value in a range, then put the < before the value. For example,
the following range does not include 0:
0<-100

Likewise, the following range does not include 100:

0-<100

If a value at the high end of one range also appears at the low end of another
range, and you do not use the < noninclusion notation, then PROC FORMAT
assigns the value to the rst range. For example, in the following ranges, the
value AJ is part of the rst range:
AA-AJ=1 AJ-AZ=2

In this example, to include the value AJ in the second range, use the noninclusive
notation on the rst range:
AA-<AJ=1 AJ-AZ=2

If you overlap values in ranges, then PROC FORMAT returns an error message
unless, for the VALUE statement, the MULTILABEL option is specied. For
example, the following ranges will cause an error:
AA-AK=1 AJ-AZ=2

Each value-or-range can be up to 32,767 characters. If value-or-range has more than

32,767 characters, then the procedure truncates the value after it processes the rst
32,767 characters.
Note: You do not have to account for every value on the left side of the equals sign.
Those values are converted using the default informat or format. For example, the
following VALUE statement creates the TEMP. format, which prints all occurrences of
98.6 as NORMAL:
value temp 98.6=NORMAL;

If the value were 96.9, then the printed result would be 96.9.

The FORMAT Procedure

Associating Informats and Formats with Variables

455

Concepts: FORMAT Procedure

Associating Informats and Formats with Variables

Methods of Associating Informats and Formats with Variables

Table 22.2 on page 455 summarizes the different methods for associating informats
and formats with variables.
Table 22.2

Associating Informats and Formats with Variables

Step

Informats

Formats

In a DATA step

Use the ATTRIB or INFORMAT statement

to permanently associate an informat with
a variable. Use the INPUT function or
INPUT statement to associate the informat
with the variable only for the duration of
the DATA step.

Use the ATTRIB or FORMAT statement to

permanently associate a format with a
variable. Use the PUT function or PUT
statement to associate the format with the
variable only for the duration of the DATA
step.

In a PROC step

The ATTRIB and INFORMAT statements

are valid in base SAS procedures. However,
in base SAS software, typically you do not
assign informats in PROC steps because the
data has already been read into SAS
variables.

Use the ATTRIB statement or the FORMAT

statement to associate formats with
variables. If you use either statement in a
procedure that produces an output data set,
then the format is permanently associated
with the variable in the output data set. If
you use either statement in a procedure that
does not produce an output data set or
modify an existing data set, the statement
associates the format with the variable only
for the duration of the PROC step.

Tips
3 Do not confuse the FORMAT statement with the FORMAT procedure. The
FORMAT and INFORMAT statements associate an existing format or informat
(either standard SAS or user-dened) with one or more variables. PROC FORMAT
creates user-dened formats or informats. Assigning your own format or informat
to a variable is a two-step process: creating the format or informat with the
FORMAT procedure, and then assigning the format or informat with the
FORMAT, INFORMAT, or ATTRIB statement.

3 It is often useful to assign informats in the FSEDIT procedure in SAS/FSP

software and in the BUILD procedure in SAS/AF software.

456

Storing Informats and Formats

Chapter 22

See Also
3 For complete documentation on the ATTRIB, INFORMAT, and FORMAT
statements, see the section on statements in SAS Language Reference: Dictionary.
3 For complete documentation on the INPUT and PUT functions, see the section on
functions in SAS Language Reference: Dictionary.
3 See Formatted Values on page 25 for more information and examples of using
formats in base SAS procedures.

Storing Informats and Formats

Format Catalogs
PROC FORMAT stores user-dened informats and formats as entries in SAS
catalogs.* You use the LIBRARY= option in the PROC FORMAT statement to specify
the catalog. If you omit the LIBRARY= option, then formats and informats are stored
in the WORK.FORMATS catalog. If you specify LIBRARY=libref but do not specify a
catalog name, then formats and informats are stored in the libref.FORMATS catalog.
Note that this use of a one-level name differs from the use of a one-level name
elsewhere in SAS. With the LIBRARY= option, a one-level name indicates a library;
elsewhere in SAS, a one-level name indicates a le in the WORK library.
The name of the catalog entry is the name of the format or informat. The entry types
are
3 FORMAT for numeric formats
3 FORMATC for character formats
3 INFMT for numeric informats
3 INFMTC for character informats.

Temporary Informats and Formats

Informats and formats are temporary when they are stored in a catalog in the
WORK library. If you omit the LIBRARY= option, then PROC FORMAT stores the
informats and formats in the temporary catalog WORK.FORMATS. You can retrieve
temporary informats and formats only in the same SAS session or job in which they are
created. To retrieve a temporary format or informat, simply include the name of the
format or informat in the appropriate SAS statement. SAS automatically looks for the
format or informat in the WORK.FORMATS catalog.

Permanent Informats and Formats

If you want to use a format or informat that is created in one SAS job or session in a
subsequent job or session, then you must permanently store the format or informat in a
SAS catalog.
You permanently store informats and formats by using the LIBRARY= option in the
PROC FORMAT statement. See the discussion of the LIBRARY= option in PROC
FORMAT Statement on page 432.
* Catalogs are a type of SAS le and reside in a SAS data library. If you are unfamiliar with the types of SAS les or the SAS
data library structure, then see the section on SAS les in SAS Language Reference: Concepts.

The FORMAT Procedure

Storing Informats and Formats

457

Accessing Permanent Informats and Formats

After you have permanently stored an informat or format, you can use it in later SAS
sessions or jobs. If you associate permanent informats or formats with variables in a
later SAS session or job, then SAS must be able to access the informats and formats.
Thus, you must use a LIBNAME statement to assign a libref to the library that stores
the catalog that stores the informats or formats.
SAS uses one of two methods when searching for user-dened formats and informats:
3 By default, SAS always searches a library that is referenced by the LIBRARY
libref for a FORMATS catalog. If you have only one format catalog, then you
should do the following:
1 Assign the LIBRARY libref to a SAS data library in the SAS session in which
you are running the PROC FORMAT step.
2 Specify LIBRARY=LIBRARY in the PROC FORMAT statement. PROC
FORMAT will store the informats and formats that are dened in that step
in the LIBRARY.FORMATS catalog.
3 In the SAS program that uses your user-dened formats and informats,
include a LIBNAME statement to assign the LIBRARY libref to the library
that contains the permanent format catalog.

3 If you have more than one format catalog, or if the format catalog is named
something other than FORMATS, then you should do the following:
1 Assign a libref to a SAS data library in the SAS session in which you are
running the PROC FORMAT step.
2 Specify LIBRARY=libref or LIBRARY=libref.catalog in the PROC FORMAT
step, where libref is the libref that you assigned in step 1.
3 In the SAS program that uses your user-dened formats and informats, use
the FMTSEARCH= option in an OPTIONS statement, and include libref or
libref.catalog in the list of format catalogs.
The syntax for specifying a list of format catalogs to search is
OPTIONS FMTSEARCH=(catalog-specication-1< catalog-specication-n>);
where each catalog-specication can be libref or libref.catalog. If only libref is specied,
then SAS assumes that the catalog name is FORMATS.
When searching for a format or informat, SAS always searches in WORK.FORMATS
rst, and then LIBRARY.FORMATS, unless one of them appears in the FMTSEARCH=
list. SAS searches the catalogs in the FMTSEARCH= list in the order that they are
listed until the format or informat is found.
For further information on FMTSEARCH=, see the section on SAS system options in
SAS Language Reference: Dictionary. For an example that uses the LIBRARY= and
FMTSEARCH= options together, see Example 8 on page 480.

Missing Formats and Informats

If you reference an informat or format that SAS cannot nd, then you receive an
error message and processing stops unless the SAS system option NOFMTERR is in
effect. When NOFMTERR is in effect, SAS uses the w. or $w. default format to print
values for variables with formats that it cannot nd. For example, to use NOFMTERR,
use this OPTIONS statement:
options nofmterr;

Refer to the section on SAS system options in SAS Language Reference: Dictionary
for more information on NOFMTERR.

458

Results: FORMAT Procedure

Chapter 22

Results: FORMAT Procedure

Output Control Data Set

The output control data set contains information that describes informats or formats.
Output control data sets have a number of uses. For example, an output control data
set can be edited with a DATA step to programmatically change value ranges or can be
subset with a DATA step to create new formats and informats. Additionally, you can
move formats and informats from one operating environment to another by creating an
output control data set, using the CPORT procedure to create a transfer le of the data
set, and then using the CIMPORT and FORMAT procedures in the target operating
environment to create the formats and informats there.
You create an output control data set with the CNTLOUT= option in the PROC
FORMAT statement. You use output control data sets, or a set of observations from an
output control data set, as an input control data set in a subsequent PROC FORMAT
step with the CNTLIN= option.
Output control data sets contain an observation for every value or range in each of
the informats or formats in the LIBRARY= catalog. The data set consists of variables
that give either global information about each format and informat created in the
PROC FORMAT step or specic information about each range and value.
The variables in the output control data set are
DEFAULT
a numeric variable that indicates the default length for format or informat
END
a character variable that gives the ranges ending value
EEXCL
a character variable that indicates whether the ranges ending value is excluded.
Values are
Y

the ranges ending value is excluded

the ranges ending value is not excluded

FILL
for picture formats, a numeric variable whose value is the value of the FILL=
option
FMTNAME
a character variable whose value is the format or informat name
FUZZ
a numeric variable whose value is the value of the FUZZ= option
HLO
a character variable that contains range information about the format or informat
in the form of eight different letters that can appear in any combination. Values
are
F

standard SAS format or informat used for formatted value or

informatted value

ranges ending value is HIGH

numeric informat range (informat dened with unquoted

numeric range)

The FORMAT Procedure

Output Control Data Set

459

ranges starting value is LOW

format or informat has no ranges, including no OTHER= range

range is OTHER

MULTILABEL option is in effect

ROUND option is in effect

NOTSORTED option is in effect

LABEL
a character variable whose value is the informatted or formatted value or the
name of an existing informat or format
LENGTH
a numeric variable whose value is the value of the LENGTH= option
MAX
a numeric variable whose value is the value of the MAX= option
MIN
a numeric variable whose value is the value of the MIN= option
MULT
a numeric variable whose value is the value of the MULT= option
NOEDIT
for picture formats, a numeric variable whose value indicates whether the
NOEDIT option is in effect. Values are
1

NOEDIT option is in effect

NOEDIT option is not in effect

PREFIX
for picture formats, a character variable whose value is the value of the PREFIX=
option
SEXCL
a character variable that indicates whether the ranges starting value is excluded.
Values are
Y

the ranges starting value is excluded

the ranges starting value is not excluded

START
a character variable that gives the ranges starting value
TYPE
a character variable that indicates the type of format. Possible values are
C

character format

numeric informat

character informat

numeric format (excluding pictures)

picture format

Output 22.1 shows an output control data set that contains information on all the
informats and formats created in Examples: FORMAT Procedure on page 463.

460

Input Control Data Set

Output 22.1

Chapter 22

Output Control Data Set for PROC FORMAT Examples

An Output Control Data Set

1
D L
D A A

M
T
N

S
T

O A
b M

A
R

E
N

B
E

M
I

s E

1 BENEFIT LOW

7304 WORDDATE20.

2 BENEFIT
3 NOZEROS LOW
4 NOZEROS

7305 HIGH
-1

5 NOZEROS
6 NOZEROS
7 PTSFRMT

0
1 HIGH
0

8 PTSFRMT
9 PTSFRMT
10 PTSFRMT

4
7
9

11 PTSFRMT
12 USCURR LOW

E
F
A

L
E
N

M
A

U
L

G
T

L
A

1 99
00.00
3 0%

N
O
S E
M F E T E E

D I T N
E G A G
C 3 T U

U
Z

F
I

U I D Y X X H
L L I P C C L

S S Y A
E E P G

T L T E L L O

P P E E

1E-12

0.00

0 N N N LF

1E-12
0.00
1E-12 - 100.00
1E-12 -. 100.00

0 N N N H
0 P N N L
0 P Y Y
0 P N Y
0 P N N H
0 N N N

100.00
100.00
0.00

. ,
. ,

1 40
1 40
1 40

5
5
3

1E-12 .
1E-12
1E-12

1 40
1 40
1 40

3
3
3

1E-12
1E-12
1E-12

0.00
0.00
0.00

0 N N N
0 N N N
0 N N N

10%
000,000

1 40
1 40

3
7

1E-12
1E-12 $

0.00
1.61

0 N N N H
0 P N N LH . ,

6 3%
8 6%
10 8%

11 HIGH
HIGH

1 40 20 20

** Not Eligible ** 1 40 20 20
-1 00.00
1 40 5 5
0 99
1 40 5 5

P
R
E

13 CITY
14 CITY
15 CITY

BR1
BR2
BR3

Birmingham UK
Plymouth UK
York UK

1 40 14 14
1 40 14 14
1 40 14 14

0
0
0

0.00
0.00
0.00

0 C N N
0 C N N
0 C N N

16 CITY
17 CITY
18 CITY

US1
US2
**OTHER**

Denver USA
Miami USA
INCORRECT CODE

1 40 14 14
1 40 14 14
1 40 14 14

0
0
0

0.00
0.00
0.00

0 C N N
0 C N N
0 C N N O

19 EVAL
20 EVAL
21 EVAL

C
E
N

1 1 40
2 1 40
0 1 40

1
1
1

0
0
0

0.00
0.00
0.00

0 I N N
0 I N N
0 I N N

22 EVAL
23 EVAL

O
S

4 1 40
3 1 40

1
1

0
0

0.00
0.00

. ,
. ,

0 I N N
0 I N N

You can use the SELECT or EXCLUDE statement to control which formats and
informats are represented in the output control data set. For details, see SELECT
Statement on page 447 and EXCLUDE Statement on page 434.

Input Control Data Set

You specify an input control data set with the CNTLIN= option in the PROC
FORMAT statement. The FORMAT procedure uses the data in the input control data
set to construct informats and formats. Thus, you can create informats and formats
without writing INVALUE, PICTURE, or VALUE statements.
The input control data set must have these characteristics:

3 For both numeric and character formats, the data set must contain the variables
FMTNAME, START, and LABEL, which are described in Output Control Data
Set on page 458. The remaining variables are not always required.

3 If you are creating a character format or informat, then you must either begin the
format or informat name with a dollar sign ($) or specify a TYPE variable with the
value C.

3 If you are creating a PICTURE statement format, then you must specify a TYPE
variable with the value P.

The FORMAT Procedure

Procedure Output

461

3 If you are creating a format with ranges of input values, then you must specify the
END variable. If range values are to be noninclusive, then the variables SEXCL
and EEXCL must each have a value of Y. Inclusion is the default.
You can create more than one format from an input control data set if the
observations for each format are grouped together.
You can use a VALUE, INVALUE, or PICTURE statement in the same PROC
FORMAT step with the CNTLIN= option. If the VALUE, INVALUE, or PICTURE
statement is creating the same informat or format that the CNTLIN= option is
creating, then the VALUE, INVALUE, or PICTURE statement creates the informat or
format and the CNTLIN= data set is not used. You can, however, create an informat or
format with VALUE, INVALUE, or PICTURE and create a different informat or format
with CNTLIN= in the same PROC FORMAT step.
For an example featuring an input control data set, see Example 5 on page 472.

Procedure Output
The FORMAT procedure prints output only when you specify the FMTLIB option or
the PAGE option in the PROC FORMAT statement. The printed output is a table for
each format or informat entry in the catalog that is specied in the LIBRARY= option.
The output also contains global information and the specics of each value or range
that is dened for the format or informat. You can use the SELECT or EXCLUDE
statement to control which formats and informats are represented in the FMTLIB
output. For details, see SELECT Statement on page 447 and EXCLUDE Statement
on page 434. For an example, see Example 6 on page 477.
The FMTLIB output shown in Output 22.2 contains a description of the NOZEROS.
format, which is created in Building a Picture Format: Step by Step on page 443, and
the EVAL. informat, which is created in Example 4 on page 470.

462

Procedure Output

Chapter 22

Output 22.2

Output from PROC FORMAT with the FMTLIB Option

FMTLIB Output for the NOZEROS. Format and the
EVAL. Informat

---------------------------------------------------------------------------|
FORMAT NAME: NOZEROS LENGTH:
5
NUMBER OF VALUES:
4
|
|
MIN LENGTH:
1 MAX LENGTH: 40 DEFAULT LENGTH
5 FUZZ: STD
|
|--------------------------------------------------------------------------|
|START
|END
|LABEL (VER. 7.00
29MAY98:10:00:24) |
|----------------+----------------+----------------------------------------|
|LOW
|
-1|00.00
P- F M100
|
|
-1<
0<99
P-. F M100
|
|
0|
1<99
P. F M100
|
|
1|HIGH
|00.00
P
F M100
|
----------------------------------------------------------------------------

---------------------------------------------------------------------------|
INFORMAT NAME: @EVAL
LENGTH:
1
NUMBER OF VALUES:
5
|
|
MIN LENGTH:
1 MAX LENGTH: 40 DEFAULT LENGTH
1 FUZZ:
0
|
|--------------------------------------------------------------------------|
|START
|END
|INVALUE(VER. 7.00
29MAY98:10:00:25) |
|----------------+----------------+----------------------------------------|
|C
|C
|
1|
|E
|E
|
2|
|N
|N
|
0|
|O
|O
|
4|
|S
|S
|
3|
----------------------------------------------------------------------------

The elds are described below in the order they appear in the output, from left to right:
INFORMAT NAME
FORMAT NAME

the name of the informat or format. Informat names begin with an at-sign (@).
LENGTH

the length of the informat or format. PROC FORMAT determines the length in the
following ways:

3 For character informats, the value for LENGTH is the length of the longest
raw data value on the left side of the equals sign.

3 For numeric informats

3 LENGTH is 12 if all values on the left side of the equals sign are
numeric.

3 LENGTH is the same as the longest raw data value on the left side of
the equal sign.

3 For formats, the value for LENGTH is the length of the longest value on the
right side of the equal sign.
In the output for @EVAL., the length is 1 because 1 is the length of the longest
raw data value on the left side of the equals sign.
In the output for NOZEROS., the LENGTH is 5 because the longest picture is 5
characters.
NUMBER OF VALUES

the number of values or ranges associated with the informat or format.

NOZEROS. has 4 ranges, EVAL. has 5.

The FORMAT Procedure

Examples: FORMAT Procedure

463

MIN LENGTH

the minimum length of the informat or format. The value for MIN LENGTH is 1
unless you specify a different minimum length with the MIN= option.
MAX LENGTH

the maximum length of the informat or format. The value for MAX LENGTH is 40
unless you specify a different maximum length with the MAX= option.
DEFAULT LENGTH

the length of the longest value in the INVALUE or LABEL eld, or the value of
the DEFAULT= option.
FUZZ

the fuzz factor. For informats, FUZZ always is 0. For formats, the value for this
eld is STD if you do not use the FUZZ= option. STD signies the default fuzz
value.
START

the beginning value of a range. FMTLIB prints only the rst 16 characters of a
value in the START and END columns.
END

the ending value of a range. The exclusion sign (<) appears after the values in
START and END, if the value is excluded from the range.
INVALUE
LABEL

INVALUE appears only for informats and contains the informatted values.
LABEL appears only for formats and contains either the formatted value or
picture. The SAS release number and the date on which the format or informat
was created are in parentheses after INVALUE or LABEL.
For picture formats, such as NOZEROS., the LABEL section contains the
PREFIX=, FILL=, and MULT= values. To note these values, FMTLIB prints the
letters P, F, and M to represent each option, followed by the value. For example, in
the LABEL section, P-. indicates that the prex value is a dash followed by a
period.
FMTLIB prints only 40 characters in the LABEL column.

Examples: FORMAT Procedure

Several examples in this section use the PROCLIB.STAFF data set. In addition,
many of the informats and formats that are created in these examples are stored in
LIBRARY.FORMATS. The output data set shown in Output Control Data Set on page
458 contains a description of these informats and the formats.
libname proclib SAS-data-library;

Create the data set PROCLIB.STAFF. The INPUT statement assigns the names Name,
IdNumber, Salary, Site, and HireDate to the variables that appear after the DATALINES
statement. The FORMAT statement assigns the standard SAS format DATE7. to the variable
HireDate.

data proclib.staff;
input Name & $16. IdNumber $ Salary

464

Example 1: Creating a Picture Format

Chapter 22

30JAN79
18JUN76
20MAR84
18SEP74
10JAN93
16FEB83
02FEB90
15APR86
19JUN93
18DEC91

The variables are about a small subset of employees who work for a corporation that
has sites in the U.S. and Britain. The data contain the name, identication number,
salary (in British pounds), location, and date of hire for each employee.

Example 1: Creating a Picture Format

Procedure features:

PROC FORMAT statement options:

LIBRARY=
PICTURE statement options:
MULT=
PREFIX=
LIBRARY libref
LOW and HIGH keywords
Data set:

PROCLIB.STAFF on page 463.

This example uses a PICTURE statement to create a format that prints the values
for the variable Salary in the data set PROCLIB.STAFF in U.S. dollars.

Program

Assign two SAS library references (PROCLIB and LIBRARY). Assigning a library
reference LIBRARY is useful in this case because if you use PROC FORMAT, then SAS
automatically searches for informats and formats in any library that is referenced with the
LIBRARY libref.
libname proclib SAS-data-library-1 ;
libname library SAS-data-library-2;

The FORMAT Procedure

Program

465

Specify that user-dened formats will be stored in the catalog LIBRARY.FORMATS.

The LIBRARY= option species a SAS catalog that will contain the formats or informats that
you create with PROC FORMAT. When you create the library named LIBRARY, SAS
automatically creates a catalog named FORMATS inside LIBRARY.
proc format library=library;

Dene the USCurrency. picture format. The PICTURE statement creates a template for
printing numbers. LOW-HIGH ensures that all values are included in the range. The MULT=
statement option species that each value is multiplied by 1.61. The PREFIX= statement adds a
US dollar sign to any number that you format. The picture contains six digit selectors, ve for
the salary and one for the dollar sign prex.
picture uscurrency low-high=000,000 (mult=1.61 prefix=$);
run;

Print the PROCLIB.STAFF data set. The NOOBS option suppresses the printing of
observation numbers. The LABEL option uses variable labels instead of variable names for
column headings.
proc print data=proclib.staff noobs label;

Specify a label and format for the Salary variable. The LABEL statement substitutes the
specic label for the variable in the report. In this case, Salary in US Dollars is substituted for
the variable Salary for this print job only. The FORMAT statement associates the USCurrency.
format with the variable name Salary for the duration of this procedure step.
label salary=Salary in U.S. Dollars;
format salary uscurrency.;

Specify the title.

title PROCLIB.STAFF with a Format for the Variable Salary;
run;

466

Output

Chapter 22

Output
PROCLIB.STAFF with a Format for the Variable Salary

Name

Id
Number

Salary in
U.S.
Dollars

2355
5889
3878
4409
3985
0740
2398
5162
4421
7385

$34,072
$33,771
$31,509
$55,256
$78,980
$80,648
$56,643
$64,561
$62,403
$37,010

Capalleti, Jimmy
Chen, Len
Davis, Brad
Leung, Brenda
Martinez, Maria
Orfali, Philip
Patel, Mary
Smith, Robert
Sorrell, Joseph
Zook, Carla

Site
BR1
BR1
BR2
BR2
US2
US2
BR3
BR5
US1
BR3

Hire
Date
30JAN79
18JUN76
20MAR84
18SEP74
10JAN93
16FEB83
02FEB90
15APR86
19JUN93
18DEC91

Example 2: Creating a Format for Character Values

Procedure features:

VALUE statement
OTHER keyword
Data set:

PROCLIB.STAFF on page 463.

Format: USCurrency. on page 465

This example uses a VALUE statement to create a character format that prints a
value of a character variable as a different character string.

Program

The FORMAT Procedure

Program

467

Create the catalog named LIBRARY.FORMATS, where the user-dened formats will be
stored. The LIBRARY= option species a permanent storage location for the formats that you
create. It also creates a catalog named FORMAT in the specied library. If you do not use
LIBRARY=, then SAS temporarily stores formats and informats that you create in a catalog
named WORK.FORMATS.
proc format library=library;

Dene the $CITY. format. The special codes BR1, BR2, and so on, are converted to the names
of the corresponding cities. The keyword OTHER species that values in the data set that do
not match any of the listed city code values are converted to the value INCORRECT CODE.
value

$city BR1=Birmingham UK
BR2=Plymouth UK
BR3=York UK
US1=Denver USA
US2=Miami USA
other=INCORRECT CODE;

run;

Specify a label for the Salary variable. The LABEL statement substitutes the label Salary
in U.S. Dollars for the name SALARY.
label salary=Salary in U.S. Dollars;

Specify formats for Salary and Site. The FORMAT statement temporarily associates the
USCurrency. format (created in Example 1 on page 464) with the variable SALARY and also
temporarily associates the format $CITY. with the variable SITE.
format salary uscurrency. site $city.;

Specify the titles.

title PROCLIB.STAFF with a Format for the Variables;
title2 Salary and Site;
run;

468

Output

Chapter 22

Output
PROCLIB.STAFF with a Format for the Variables
Salary and Site

Name
Capalleti, Jimmy
Chen, Len
Davis, Brad
Leung, Brenda
Martinez, Maria
Orfali, Philip
Patel, Mary
Smith, Robert
Sorrell, Joseph
Zook, Carla

Id
Number

Salary in
U.S.
Dollars

2355
5889
3878
4409
3985
0740
2398
5162
4421
7385

$34,072
$33,771
$31,509
$55,256
$78,980
$80,648
$56,643
$64,561
$62,403
$37,010

Site
Birmingham UK
Birmingham UK
Plymouth UK
Plymouth UK
Miami USA
Miami USA
York UK
INCORRECT CODE
Denver USA
York UK

Hire
Date
30JAN79
18JUN76
20MAR84
18SEP74
10JAN93
16FEB83
02FEB90
15APR86
19JUN93
18DEC91

Example 3: Writing a Format for Dates Using a Standard SAS Format

Procedure features:

VALUE statement:
HIGH keyword
Data set:

PROCLIB.STAFF on page 463.

Formats:

USCurrency. on page 465 and $CITY. on page 467.

This example uses an existing format that is supplied by SAS as a formatted value.
Tasks include

3
3
3
3

creating a numeric format

nesting formats
writing a format using a standard SAS format
formatting dates.

Program
This program denes a format called BENEFIT, which differentiates between
employees hired on or before 31DEC1979. The purpose of this program is to indicate
any employees who are eligible to receive a benet, based on a hire date on or prior to
December 31, 1979. All other employees with a later hire date are listed as ineligible
for the benet.

The FORMAT Procedure

Program

469

Store the BENEFIT. format in the catalog LIBRARY.FORMATS. The LIBRARY= option
species the permanent storage location LIBRARY for the formats that you create. If you do not
use LIBRARY=, then SAS temporarily stores formats and informats that you create in a catalog
named WORK.FORMATS.
proc format library=library;

Dene the rst range in the BENEFIT. format. This rst range differentiates between the
employees who were hired on or before 31DEC1979 and those who were hired after that date.
The keyword LOW and the SAS date constant 31DEC1979D create the rst range, which
includes all date values that occur on or before December 31, 1979. For values that fall into this
range, SAS applies the WORDDATEw. format.*
value benefit

low-31DEC1979d=[worddate20.]

Dene the second range in the BENEFIT. format. The second range consists of all dates on
or after January 1, 1980. The SAS date constant 01JAN1980D and the keyword HIGH specify
the range. Values that fall into this range receive ** Not Eligible ** as a formatted value.
01JAN1980d-high=

Not Eligible ;

run;

Print the data set PROCLIB.STAFF. The NOOBS option suppresses the printing of
observation numbers. The LABEL option uses variable labels instead of variable names for
column headings.
proc print data=proclib.staff noobs label;

* For more information about SAS date constants, see the section on dates, times, and intervals in SAS Language Reference:
Concepts. For complete documentation on WORDDATEw., see the section on formats in SAS Language Reference: Dictionary.

470

Output

Chapter 22

Specify a label for the Salary variable. The LABEL statement substitutes the label Salary
in U.S. Dollars for the name SALARY.
label salary=Salary in U.S. Dollars;

Specify formats for Salary, Site, and Hiredate. The FORMAT statement associates the
USCurrency. format (created in Example 1 on page 464) with SALARY, the $CITY. format
(created in Example 2 on page 466) with SITE, and the BENEFIT. format with HIREDATE.
format salary uscurrency. site $city. hiredate benefit.;

Specify the titles.

title PROCLIB.STAFF with a Format for the Variables;
title2 Salary, Site, and HireDate;
run;

Output
PROCLIB.STAFF with a Format for the Variables
Salary, Site, and HireDate

Name
Capalleti, Jimmy
Chen, Len
Davis, Brad
Leung, Brenda
Martinez, Maria
Orfali, Philip
Patel, Mary
Smith, Robert
Sorrell, Joseph
Zook, Carla

Id
Number

Salary in
U.S.
Dollars

2355
5889
3878
4409
3985
0740
2398
5162
4421
7385

$34,072
$33,771
$31,509
$55,256
$78,980
$80,648
$56,643
$64,561
$62,403
$37,010

Site
Birmingham UK
Birmingham UK
Plymouth UK
Plymouth UK
Miami USA
Miami USA
York UK
INCORRECT CODE
Denver USA
York UK

HireDate
January 30, 1979
June 18, 1976
** Not Eligible **
September 18, 1974
** Not Eligible **
** Not Eligible **
** Not Eligible **
** Not Eligible **
** Not Eligible **
** Not Eligible **

Example 4: Converting Raw Character Data to Numeric Values

Procedure feature:

INVALUE statement

This example uses an INVALUE statement to create a numeric informat that

converts numeric and character raw data to numeric data.

Program
This program converts quarterly employee evaluation grades, which are alphabetic,
into numeric values so that reports can be generated that sum the grades up as points.

The FORMAT Procedure

Program

471

Set up two SAS library references, one named PROCLIB and the other named
LIBRARY.
libname proclib SAS-data-library-1;
libname library SAS-data-library-2;

Store the Evaluation. informat in the catalog LIBRARY.FORMATS.

proc format library=library;

Create the numeric informat Evaluation. The INVALUE statement converts the specied
values. The letters O (Outstanding), S (Superior), E (Excellent), C (Commendable), and N (None)
correspond to the numbers 4, 3, 2, 1, and 0, respectively.
invalue evaluation O=4
S=3
E=2
C=1
N=0;
run;

Create the PROCLIB.POINTS data set. The instream data, which immediately follows the
DATALINES statement, contains a unique identication number (EmployeeId) and bonus
evaluations for each employee for each quarter of the year (Q1Q4). Some of the bonus
evaluation values that are listed in the data lines are numbers; others are character values.
Where character values are listed in the data lines, the Evaluation. informat converts the value
O to 4, the value S to 3, and so on. The raw data values 0 through 4 are read as themselves
because they are not referenced in the denition of the informat. Converting the letter values to
numbers makes it possible to calculate the total number of bonus points for each employee for
the year. TotalPoints is the total number of bonus points.
data proclib.points;
input EmployeeId $ (Q1-Q4) (evaluation.,+1);
TotalPoints=sum(of q1-q4);
datalines;
2355 S O O S
5889 2 2 2 2
3878 C E E E
4409 0 1 1 1
3985 3 3 3 2
0740 S E E S
2398 E E C C

472

Output

Chapter 22

5162 C C C E
4421 3 2 2 2
7385 C C C N
;

Print the PROCLIB.POINTS data set. The NOOBS option suppresses the printing of
observation numbers.
proc print data=proclib.points noobs;

Specify the title.

title The PROCLIB.POINTS Data Set;
run;

Output
The PROCLIB.POINTS Data Set

Employee
Id

Total
Points

2355
5889
3878
4409
3985
0740
2398
5162
4421
7385

3
2
1
0
3
3
2
1
3
1

4
2
2
1
3
2
2
1
2
1

4
2
2
1
3
2
1
1
2
1

3
2
2
1
2
3
1
2
2
0

14
8
7
3
11
10
6
5
9
3

Example 5: Creating a Format from a Data Set

Procedure features:

PROC FORMAT statement option:

CNTLIN=
Input control data set
Data set:

WORK.POINTS, created from data lines in the sample code.

The FORMAT Procedure

Program

473

This example shows how to create a format from a SAS data set.
Tasks include

3 creating a format from an input control data set

3 creating an input control data set from an existing SAS data set.

Program

Create a temporary data set named scale. The rst two variables in the data lines, called
BEGIN and END, will be used to specify a range in the format. The third variable in the data
lines, called AMOUNT, contains a percentage that will be used as the formatted value in the
format. Note that all three variables are character variables as required for PROC FORMAT
input control data sets.
data scale;
input begin $ 1-2 end $ 5-8 amount $ 10-12;
datalines;
0
3
0%
4
6
3%
7
8
6%
9
10
8%
11 16
10%
;

Create the input control data set CTRL and set the length of the LABEL variable. The
LENGTH statement ensures that the LABEL variable is long enough to accommodate the label
***ERROR***.
data ctrl;
length label $ 11;

Rename variables and create an end-of-le ag. The data set CTRL is derived from
WORK.SCALE. RENAME= renames BEGIN and AMOUNT as START and LABEL,
respectively. The END= option creates the variable LAST, whose value is set to 1 when the last
observation is processed.
set scale(rename=(begin=start amount=label)) end=last;

474

Program

Chapter 22

Create the variables FMTNAME and TYPE with xed values. The RETAIN statement is
more efcient than an assignment statement in this case. RETAIN retains the value of
FMTNAME and TYPE in the program data vector and eliminates the need for the value to be
written on every iteration of the DATA step. FMTNAME species the name PercentageFormat,
which is the format that the input control data set creates. The TYPE variable species that the
input control data set will create a numeric format.
retain fmtname PercentageFormat type n;

Write the observation to the output data set.

output;

Create an other category. Because the only valid values for this application are 016, any
other value (such as missing) should be indicated as an error to the user. The IF statement
executes only after the DATA step has processed the last observation from the input data set.
When IF executes, HLO receives a value of O to indicate that the range is OTHER, and LABEL
receives a value of ***ERROR***. The OUTPUT statement writes these values as the last
observation in the data set. HLO has missing values for all other observations.
if last then do;
hlo=O;
label=***ERROR***;
output;
end;
run;

Print the control data set, CTRL. The NOOBS option suppresses the printing of observation
numbers.
proc print data=ctrl noobs;

Specify the title.

title The CTRL Data Set;
run;

The FORMAT Procedure

Program

475

Note that although the last observation contains values for START and END, these values are ignored because
of the O value in the HLO variable.

The CTRL Data Set

label

start

end

fmtname

1
type

PercentageFormat

3%
6%
8%

4
7
9

6
8
10

PercentageFormat
PercentageFormat
PercentageFormat

n
n
n

10%
***ERROR***

11
11

16
16

PercentageFormat
PercentageFormat

n
n

hlo

Store the created format in the catalog WORK.FORMATS and specify the source for
the format. The CNTLIN= option species that the data set CTRL is the source for the format
PTSFRMT.
proc format library=work cntlin=ctrl;
run;

Create the WORK.POINTS data set. The instream data, which immediately follows the
DATALINES statement, contains a unique identication number (EmployeeId) and bonus
evaluations for each employee for each quarter of the year (Q1Q4). Some of the bonus
evaluation values that are listed in the data lines are numbers; others are character values.
Where character values are listed in the data lines, the Evaluation. informat converts the value
O to 4, the value S to 3, and so on. The raw data values 0 through 4 are read as themselves
because they are not referenced in the denition of the informat. Converting the letter values to
numbers makes it possible to calculate the total number of bonus points for each employee for
the year. TotalPoints is the total number of bonus points. The addition operator is used instead
of the SUM function so that any missing value will result in a missing value for TotalPoints.
data points;
input EmployeeId $ (Q1-Q4) (evaluation.,+1);
TotalPoints=q1+q2+q3+q4;
datalines;
2355 S O O S
5889 2 . 2 2

476

Output

Chapter 22

3878
4409
3985
0740
2398
5162
4421
7385
;

C
0
3
S
E
C
3
C

E
1
3
E
E
C
2
C

E
1
3
E

E
1
2
S
C
C E
2 2
C N

Generate a report for WORK.POINTS and associate the PTSFRMT. format with the
TotalPoints variable. The DEFINE statement performs the association. The column that
contains the formatted values of TotalPoints is using the alias Pctage. Using an alias enables
you to print a variable twice, once with a format and once with the default format. See Chapter
42, The REPORT Procedure, on page 845 for more information about PROC REPORT.
proc report data=work.points nowd headskip split=#;
column employeeid totalpoints totalpoints=Pctage;
define employeeid / right;
define totalpoints / Total#Points right;
define pctage / format=PercentageFormat12. Percentage left;
title The Percentage of Salary for Calculating Bonus;
run;

Output

Output 22.3
The Percentage of Salary for Calculating Bonus
Employee

Total

Points

2355

5889
3878
4409

.
7
3

3985
0740

11
10

2398
5162
4421

.
5
9

***ERROR***
3%
8%

7385

Percentage
10%
***ERROR***
6%
0%
10%
8%

The FORMAT Procedure

Program

477

Example 6: Printing the Description of Informats and Formats

Procedure features:

PROC FORMAT statement option:

FMTLIB
SELECT statement
Format:

NOZEROS on page 444.

Informat:

Evaluation. on page 471

This example illustrates how to print a description of an informat and a format. The
description shows the values that are input and output.

Program

Set up a SAS library reference named LIBRARY.

libname library SAS-data-library;

Print a description of Evaluation. and NOZEROS. The FMTLIB option prints information
about the formats and informats in the catalog that the LIBRARY= option species.
LIBRARY=LIBRARY points to the LIBRARY.FORMATS catalog.
proc format library=library fmtlib;

Select an informat and a format. The SELECT statement selects EVAL and NOZEROS,
which were created in previous examples. The at sign (@) in front of EVAL indicates that EVAL.
is an informat.
select @evaluation nozeros;

Specify the titles.

title FMTLIB Output for the NOZEROS. Format and the;
title2 Evaluation. Informat;
run;

478

Output

Chapter 22

Output

The output is described in Procedure Output on page 461.

FMTLIB Output for the NOZEROS. Format and the

Evaluation. Informat

---------------------------------------------------------------------------|
FORMAT NAME: NOZEROS LENGTH:
5
NUMBER OF VALUES:
4
|
|
MIN LENGTH:
1 MAX LENGTH: 40 DEFAULT LENGTH
5 FUZZ: STD
|
|--------------------------------------------------------------------------|
|START
|END
|LABEL (VER. V7|V8
10APR2002:18:55:08)|
|----------------+----------------+----------------------------------------|
|LOW
|
-1|00.00
P- F M100
|
|
-1<
0<99
P-. F M100
|
|
0|
1<99
P. F M100
|
|
1|HIGH
|00.00
P
F M100
|
----------------------------------------------------------------------------

---------------------------------------------------------------------------|
INFORMAT NAME: @EVALUATION LENGTH: 1
|
|
MIN LENGTH:
1 MAX LENGTH: 40 DEFAULT LENGTH
1 FUZZ:
0
|
|--------------------------------------------------------------------------|
|START
|END
|INVALUE(VER. 9.00
10APR2002:18:55:11)|
|----------------+----------------+----------------------------------------|
|C
|C
|
1|
|E
|E
|
2|
|N
|N
|
0|
|O
|O
|
4|
|S
|S
|
3|
----------------------------------------------------------------------------

Example 7: Retrieving a Permanent Format

Procedure features:

PROC FORMAT statement options:

LIBRARY=
Other features:

FMTSEARCH= system option

Data sets:

SAMPLE on page 443.

This example uses the LIBRARY= option and the FMTSEARCH= system option to
store and retrieve a format stored in a catalog other than WORK.FORMATS or
LIBRARY.FORMATS.

The FORMAT Procedure

Program

479

Program

Set up a SAS library reference named PROCLIB.

libname proclib SAS-data-library;

Store the NOZEROS. format in the PROCLIB.FORMATS catalog.

proc format library=proclib;

Create the NOZEROS. format. The PICTURE statement denes the picture format
NOZEROS. See Building a Picture Format: Step by Step on page 443.
picture nozeros
low
-1 <-<
0
-<
1
run;

-1
0
1
high

=
=
=
=

00.00 (prefix=-
)
99 (prefix=-. mult=100)
99 (prefix=. mult=100)
00.00;

Add the PROCLIB.FORMATS catalog to the search path that SAS uses to nd
user-dened formats. The FMTSEARCH= system option denes the search path. The
FMTSEARCH= system option requires only a libref. FMTSEARCH= assumes that the catalog
name is FORMATS if no catalog name appears. Without the FMTSEARCH= option, SAS would
not nd the NOZEROS. format.*
options

fmtsearch=(proclib);

Print the SAMPLE data set. The FORMAT statement associates the NOZEROS. format with
the Amount variable.
proc print data=sample;
format amount nozeros.;

* For complete documentation on the FMTSEARCH= system option, see the section on SAS system options in SAS Language
Reference: Dictionary.

480

Output

Chapter 22

Specify the titles.

title1 Retrieving the NOZEROS. Format from PROCLIB.FORMATS;
title2 The SAMPLE Data Set;
run;

Output
Retrieving the NOZEROS. Format from PROCLIB.FORMATS
The SAMPLE Data Set
Obs

Amount

1
2
3
4
5
6
7
8
9

-2.05
-.05
-.01
.00
.09
.54
.55
6.60
14.63

Example 8: Writing Ranges for Character Strings

Data sets:

PROCLIB.STAFF on page 463.

This example creates a format and shows how to use ranges with character strings.

Program
libname proclib SAS-data-library;

The FORMAT Procedure

Program

481

Create the TRAIN data set from the PROCLIB.STAFF data set. PROCLIB.STAFF was
created in Examples: FORMAT Procedure on page 463.
data train;
set proclib.staff(keep=name idnumber);
run;

Print the data set TRAIN without a format. The NOOBS option suppresses the printing of
observation numbers.
proc print data=train noobs;

Specify the title.

title The TRAIN Data Set without a Format;
run;
The TRAIN Data Set without a Format

Name
Capalleti, Jimmy
Chen, Len
Davis, Brad
Leung, Brenda
Martinez, Maria
Orfali, Philip
Patel, Mary
Smith, Robert
Sorrell, Joseph
Zook, Carla

Id
Number
2355
5889
3878
4409
3985
0740
2398
5162
4421
7385

Store the format in WORK.FORMATS. Because the LIBRARY= option does not appear, the
format is stored in WORK.FORMATS and is available only for the current SAS session.
proc format;

Create the $SkillTest. format. The $SKILL. format prints each employees identication
number and the skills test that they have been assigned. Employees must take either TEST A,
TEST B, or TEST C, depending on their last name. The exclusion operator (<) excludes the last
value in the range. Thus, the rst range includes employees whose last name begins with any
letter from A through D, and the second range includes employees whose last name begins with
any letter from E through M. The tilde (~) in the last range is necessary to include an entire
string that begins with the letter Z.
value $skilltest a-<e,A-<E=Test A
e-<m,E-<M=Test B
m-z~,M-Z~=Test C;
run;

482

Output

Chapter 22

Generate a report of the TRAIN data set. The FORMAT= option in the DEFINE statement
associates $SkillTest. with the NAME variable. The column that contains the formatted values
of NAME is using the alias Test. Using an alias enables you to print a variable twice, once with
a format and once with the default format. See Chapter 42, The REPORT Procedure, on page
845for more information about PROC REPORT.
proc report data=train nowd headskip;
column name name=test idnumber;
define test / display format=$skilltest. Test;
define idnumber / center;
title Test Assignment for Each Employee;
run;

Output
Test Assignment for Each Employee
Name

Test

Capalleti, Jimmy
Chen, Len
Davis, Brad
Leung, Brenda
Martinez, Maria
Orfali, Philip
Patel, Mary
Smith, Robert
Sorrell, Joseph
Zook, Carla

Test
Test
Test
Test
Test
Test
Test
Test
Test
Test

IdNumber
A
A
A
B
C
C
C
C
C
C

2355
5889
3878
4409
3985
0740
2398
5162
4421
7385

The FORMAT Procedure

Program

483

Example 9: Filling a Picture Format

Procedure features:

PICTURE statement options:

FILL=
PREFIX=

This example

3 prexes the formatted value with a specied character

3 lls the leading blanks with a specied character
3 shows the interaction between the FILL= and PREFIX= options.

Program

Create the PAY data set. The PAY data set contains the monthly salary for each employee.
data pay;
input Name $ MonthlySalary;
datalines;
Liu
1259.45
Lars 1289.33
Kim
1439.02
Wendy 1675.21
Alex 1623.73
;

Dene the SALARY. picture format and specify how the picture will be lled. When
FILL= and PREFIX= PICTURE statement options appear in the same picture, the format
places the prex and then the ll characters. The SALARY. format lls the picture with the ll
character because the picture has zeros as digit selectors. The leftmost comma in the picture is
replaced by the ll character.
proc format;
picture salary low-high=00,000,000.00 (fill=* prefix=$);
run;

484

Output

Chapter 22

Print the PAY data set. The NOOBS option suppresses the printing of observation numbers.
The FORMAT statement temporarily associates the SALARY. format with the variable
MonthlySalary.
proc print data=pay noobs;
format monthlysalary salary.;

Specify the title.

title Printing Salaries for a Check;
run;

Output
Printing Salaries for a Check
Name

MonthlySalary

Liu
Lars
Kim
Wendy
Alex

****$1,259.45
****$1,289.33
****$1,439.02
****$1,675.21
****$1,623.73

See Also
FMTSEARCH= System option
VALIDFMTNAME= System option
FORMAT Statement

485

CHAPTER

23
The FORMS Procedure
Information about the FORMS Procedure

485

Information about the FORMS Procedure

For documentation of the FORMS procedure, go to http://support.sas.com/
documentation/onlinedoc. Select Base SAS from the Product-Specic Documentation

See:

list.

486

487

CHAPTER

24
The FREQ Procedure
Information about the FREQ Procedure

487

Information about the FREQ Procedure

See:

The documentation for the FREQ procedure has moved to Volume 3 of this book.

488

489

CHAPTER

25
The FSLIST Procedure
Overview: FSLIST Procedure 489
Syntax: FSLIST Procedure 489
Statement Descriptions 490
PROC FSLIST Statement 490
FSLIST Command 492
Using the FSLIST Window 494
General Information about the FSLIST Window
FSLIST Window Commands 494
Global Commands 494
Scrolling Commands 494
Searching Commands 496
Display Commands 498
Other Commands 499

494

Overview: FSLIST Procedure

The FSLIST procedure enables you to browse external les that are not SAS data
sets within a SAS session. Because the les are displayed in an interactive window, the
procedure provides a highly convenient mechanism for examining le contents. In
addition, you can copy text from the FSLIST window into any window that uses the
SAS Text Editor.

Syntax: FSLIST Procedure

PROC FSLIST
FILEREF=le-specication|UNIT=nn <option(s)>;

3 You must specify either the FILEREF= or the UNIT= argument with the PROC
FSLIST statement.

3 Option(s) can be one or more of the following:

490

Statement Descriptions

Chapter 25

NUM|NONUM
OVP|NOOVP

Statement Descriptions
The only statement that the FSLIST procedure supports is the PROC FSLIST
statement, which starts the procedure.

Requirements
You must specify an external le for PROC FSLIST to browse.

FSLIST Command
The FSLIST procedure can also be initiated by entering the following command on
the command line of any SAS window:
FSLIST <*|?| le-specication <carriage-control-option <overprinting-option>>>
where carriage-control-option can be CC, FORTCC, or NOCC and overprinting-option
can be OVP or NOOVP.
Note:

OVP is ignored if NOCC is in effect.

PROC FSLIST Statement

The PROC FSLIST statement initiates the FSLIST procedure and species the
external le to browse. Statement options enable you to modify the default behavior of
the procedure.

PROC FSLIST Statement Requirements

The PROC FSLIST statement must include one of the following arguments that
species the external le to browse.
FILEREF=le-specication
DDNAME=le-specication
DD=le-specication

species the external le to browse. le-specication can be one of the following:

external-le
is the complete operating environment le specication (called the fully qualied
pathname under some operating environments) for the external le. You must
enclose the name in quotation marks.
leref
is a leref that has been previously assigned to the external le. You can use the
FILENAME statement to associate a leref with an actual lename. For
information about the FILENAME statement, see the section on statements in
SAS Language Reference: Dictionary.

The FSLIST Procedure

PROC FSLIST Statement

491

UNIT=nn

denes the FORTRAN-style logical unit number of the external le to browse. This
option is useful when the le to browse has a leref of the form FTnnF001, where nn
is the logical unit number that is specied in the UNIT= argument. For example, you
can specify
proc fslist unit=20;

instead of
proc fslist fileref=ft20f001;

PROC FSLIST Statement Options

The following options can be used with the PROC FSLIST statement:
CAPS | NOCAPS

controls how search strings for the FIND command are treated:
CAPS

converts search strings into uppercase unless they are enclosed in

quotation marks. For example, with this option in effect, the
command
find nc

locates occurrences of NC, but not nc. To locate lowercase

characters, enclose the search string in quotation marks:
find nc

NOCAPS

does not perform a translation; the FIND command locates only

those text strings that exactly match the search string.

The default is NOCAPS. You can use the CAPS command in the FSLIST window to
change the behavior of the procedure while you are browsing a le.
CC | FORTCC | NOCC

indicates whether carriage-control characters are used to format the display. You can
specify one of the following values for this option:
CC

uses the native carriage-control characters of the operating

environment.

FORTCC

uses FORTRAN-style carriage control. The rst column of each

line in the external le is not displayed; the character in this
column is interpreted as a carriage-control code. The FSLIST
procedure recognizes the following carriage-control characters:
+
blank

skip one line and print (single space)

skip two lines and print (double space)

skip three lines and print (triple space)

1
NOCC

skip zero lines and print (overprint)

go to new page and print.

treats carriage-control characters as regular text.

If the FSLIST procedure can determine from the les attributes that the le
contains carriage-control information, then that carriage-control information is used
to format the displayed text (the CC option is the default). Otherwise, the entire
contents of the le are treated as text (the NOCC option the default).

492

FSLIST Command

Chapter 25

Note: Under some operating environments, FORTRAN-style carriage control is

the native carriage control. For these environments, the FORTCC and CC options
produce the same behavior. 4
HSCROLL=n|HALF|PAGE

indicates the default horizontal scroll amount for the LEFT and RIGHT commands.
The following values are valid:
n

sets the default scroll amount to n columns.

HALF

sets the default scroll amount to half the window width.

PAGE
sets the default scroll amount to the full window width.
The default is HSCROLL=HALF. You can use the HSCROLL command in the
FSLIST window to change the default scroll amount.
NOBORDER

suppresses the sides and bottom of the FSLIST windows border. When this option is
used, text can appear in the columns and row that are normally occupied by the
border.
NUM | NONUM

controls the display of line sequence numbers in les that have a record length of 80
and contain sequence numbers in columns 73 through 80. NUM displays the line
sequence numbers; NONUM suppresses them. The default is NONUM.
OVP| NOOVP

indicates whether the carriage-control code for overprinting is honored:

OVP

causes the procedure to honor the overprint code and print the
current line over the previous line when the code is encountered.

NOOVP

causes the procedure to ignore the overprint code and print each
line from the le on a separate line of the display.

The default is NOOVP. The OVP option is ignored if the NOCC option is in effect.

FSLIST Command
The FSLIST command provides a handy way to initiate an FSLIST session from any
SAS window. The command enables you to use either a leref or a lename to specify
the le to browse. It also enables you to specify how carriage-control information is
interpreted.

FSLIST Command Syntax

The general form of the FSLIST command is
FSLIST <*|?| le-specication <carriage-control-option <overprinting-option>>>
where carriage-control-option can be CC, FORTCC, or NOCC and overprinting-option
can be OVP or NOOVP.
Note:

OVP is ignored if NOCC is in effect.

The FSLIST Procedure

FSLIST Command

493

FSLIST Command Arguments

You can specify one of the following arguments with the FSLIST command:
*

opens a dialog window in which you can specify the name of the le to browse, along
with various FSLIST procedure options. In the dialog window, you can specify either
a physical lename, a leref, or a directory name. If you specify a directory name,
then a selection list of the les in the directory appears, from which you can choose
the desired le.
?

opens a selection window from which you can choose the external le to browse. The
selection list in the window includes all external les that are identied in the
current SAS session (all les with dened lerefs).
Note: Only lerefs that are dened within the current SAS session appear in the
selection list. Under some operating environments, it is possible to allocate lerefs
outside of SAS. Such lerefs do not appear in the selection list that is displayed by
the FSLIST command. 4
To select a le, position the cursor on the corresponding leref and press ENTER.
Note: The selection window is not opened if no lerefs have been dened in the
current SAS session. Instead, an error message is printed, instructing you to enter a
lename with the FSLIST command. 4
le-specication

identies the external le to browse. le-specication can be one of the following:

external-le
the complete operating environment le specication (called the fully qualied
pathname under some operating environments) for the external le. You must
enclose the name in quotation marks.
If the specied le is not found, then a selection window opens that shows all
available lerefs.
leref
a leref that is currently assigned to an external le. If you specify a leref that is
not currently dened, then a selection window opens that shows all available
lerefs. An error message in the selection window indicates that the specied
leref is not dened.
If you do not specify any of these three arguments, then a selection window opens
that enables you to select an external lename.

FSLIST Command Options

If you use a le-specication with the FSLIST command, then you can also use the
following options. These options are not valid with the ? argument, or when no
argument is used:
CC | FORTCC | NOCC

indicates whether carriage-control characters are used to format the display. You can
specify one of the following values for this option:
CC

uses the native carriage-control characters of the operating

environment.

FORTCC

uses FORTRAN-style carriage control. See the discussion of the

PROC FSLIST statements FORTCC option on page 491 for
details.

494

Using the FSLIST Window

Chapter 25

NOCC
treats carriage-control characters as regular text.
If the FSLIST procedure can determine from the les attributes that the le
contains carriage-control information, then that carriage-control information is used
to format the displayed text (the CC option is the default). Otherwise, the entire
contents of the le are treated as text (the NOCC option is the default).
OVP | NOOVP

indicates whether the carriage-control code for overprinting is honored. OVP causes
the overprint code to be honored; NOOVP causes it to be ignored. The default is
NOOVP. The OVP option is ignored if NOCC is in effect.

Using the FSLIST Window

General Information about the FSLIST Window
The FSLIST window displays les for browsing only. You cannot edit les in the
FSLIST window. However, you can copy text from the FSLIST window into a paste
buffer by doing one of the following, depending on your operating environment:
3 use a mouse to select text, and select Copy from the Edit menu
3 use the global MARK and STORE commands.
Depending on your operating environment, this text can then be pasted into any SAS
window that uses the SAS text editor, including the FSLETTER window in SAS/FSP
software, or into any other application that allows pasting of text.
You can use commands in the command window or command line to control the
FSLIST window.

The FSLIST Procedure

FSLIST Window Commands

497

The CAPS option on the PROC FSLIST statement and the CAPS ON command
cause search strings to be converted to uppercase for the purposes of the search,
unless the strings are enclosed in quotation marks. See the discussion of the FIND
command for details.
By default, the BFIND command locates any occurrence of the specied string,
even where the string is embedded in other strings. You can use any one of the
following options to alter the commands behavior:
PREFIX
causes the search string to match the text string only when the text string
occurs at the beginning of a word.
SUFFIX
causes the search string to match the text string only when the text string
occurs at the end of a word.
WORD
causes the search string to match the text string only when the text string is
a distinct word.
You can use the RFIND command to repeat the most recent BFIND command.
CAPS <ON|OFF>
controls how the FIND, BFIND, and RFIND commands locate matches for a
search string. By default, the FIND, BFIND, and RFIND commands locate only
those text strings that exactly match the search string as it was entered. When
you issue the CAPS command, the FIND, BFIND, and RFIND commands convert
search strings into uppercase for the purposes of searching (displayed text is not
affected), unless the strings are enclosed in quotation marks. Strings in quotation
marks are not affected.
For example, after you issue a CAPS ON command, both of the following
commands locate occurrences of NC but not occurrences of nc:
find NC
find nc

If you omit the ON or OFF argument, then the CAPS command acts as a toggle,
turning the attribute on if it was off or off if it was on.
FIND search-string <NEXT|FIRST|LAST|PREV|ALL>
<PREFIX|SUFFIX|WORD>
locates an occurrence of the specied search-string in the le. The search-string
must be enclosed in quotation marks if it contains embedded blanks.
The text in the search-string must match the text in the le in terms of both
characters and case. For example, the command
find raleigh

will locate not the text Raleigh in the le. You must instead use
find Raleigh

498

FSLIST Window Commands

Chapter 25

When the CAPS option is used with the PROC FSLIST statement or when a
CAPS ON command is issued in the window, the search string is converted to
uppercase for the purposes of the search, unless the string is enclosed in quotation
marks. In that case, the command
find raleigh

will locate only the text RALEIGH in the le. You must instead use the command
find Raleigh

to locate the text Raleigh.

You can modify the behavior of the FIND command by adding any one of the
following options:
ALL
reports the total number of occurrences of the string in the le in the
windows message line and moves the cursor to the rst occurrence.
FIRST
moves the cursor to the rst occurrence of the string in the le.
LAST
moves the cursor to the last occurrence of the string in the le.
NEXT
moves the cursor to the next occurrence of the string in the le.
PREV
moves the cursor to the previous occurrence of the string in the le.
The default option is NEXT.
By default, the FIND command locates any occurrence of the specied string,
even where the string is embedded in other strings. You can use any one of the
following options to alter the commands behavior:
PREFIX
causes the search string to match the text string only when the text string
occurs at the beginning of a word.
SUFFIX
causes the search string to match the text string only when the text string
occurs at the end of a word.
WORD
causes the search string to match the text string only when the text string is
a distinct word.
After you issue a FIND command, you can use the RFIND command to repeat the
search for the next occurrence of the string, or you can use the BFIND command
to repeat the search for the previous occurrence.
RFIND
repeats the most recent FIND command, starting at the current cursor position
and proceeding forward toward the end of the le.

Display Commands
COLUMN <ON|OFF>
displays a column ruler below the message line in the FSLIST window. The ruler
is helpful when you need to determine the column in which a particular character

The FSLIST Procedure

FSLIST Window Commands

499

is located. If you omit the ON or OFF specication, then the COLUMN command
acts as a toggle, turning the ruler on if it was off and off if it was on.
HEX <ON|OFF>
controls the special hexadecimal display format of the FSLIST window. When the
hexadecimal format is turned on, each line of characters from the le occupies
three lines of the display. The rst is the line displayed as characters; the next two
lines of the display show the hexadecimal value of the operating environments
character codes for the characters in the line of text. The hexadecimal values are
displayed vertically, with the most signicant byte on top. If you omit the ON or
OFF specication, then the HEX command acts as a toggle, turning the
hexadecimal format on if it was off and off if it was on.
NUMS <ON|OFF>
controls whether line numbers are shown at the left side of the window. By
default, line numbers are not displayed. If line numbers are turned on, then they
remain at the left side of the display when text in the window is scrolled right and
left. If you omit the ON or OFF argument, then the NUMS command acts as a
toggle, turning line numbering on if it was off or off if it was on.

Other Commands
BROWSE leref|actual-lename <CC|FORTCC|NOCC <OVP|NOOVP>>
closes the current le and displays the specied le in the FSVIEW window. You
can specify either a leref previously associated with a le or an actual lename
enclosed in quotation marks. The BROWSE command also accepts the same
carriage-control options as the FSLIST command. See FSLIST Command
Options on page 493 for details.
END
closes the FSLIST window and ends the FSLIST session.
HELP <command>
opens a Help window that provides information about the FSLIST procedure and
about the commands available in the FSLIST window. To get information about a
specic FSLIST window command, follow the HELP command with the name of
the desired command.
KEYS
opens the KEYS window for browsing and editing function key denitions for the
FSLIST window. The default key denitions for the FSLIST window are stored in
the FSLIST.KEYS entry in the SASHELP.FSP catalog.
If you change any key denitions in the KEYS window, then a new
FSLIST.KEYS entry is created in your personal PROFILE catalog
(SASUSER.PROFILE, or WORK.PROFILE if the SASUSER library is not
allocated).
When the FSLIST procedure is initiated, it looks for function key denitions
rst in the FSLIST.KEYS entry in your personal PROFILE catalog. If that entry
does not exist, then the default entry in the SASHELP.FSP catalog is used.

500

501

CHAPTER

26
The IMPORT Procedure
Overview: IMPORT Procedure 501
Syntax: IMPORT Procedure 502
PROC IMPORT Statement 502
Data Source Statements 506
Examples: IMPORT Procedure 514
Example 1: Importing a Delimited External File 514
Example 2: Importing a Specic Spreadsheet from an Excel Workbook 517
Example 3: Importing a Subset of Records from an Excel Spreadsheet 518
Example 4: Importing a Microsoft Access Table 519
Example 5: Importing a Specic Spreadsheet from an Excel Workbook on a PC Server

521

Overview: IMPORT Procedure

The IMPORT procedure reads data from an external data source and writes it to a
SAS data set. External data sources can include Microsoft Access Database, Excel les,
Lotus spreadsheets, and delimited external les (in which columns of data values are
separated by a delimiter such as a blank, comma, or tab).
When you execute PROC IMPORT, the procedure reads the input le and writes the
data to a SAS data set. The SAS variable denitions are based on the input records.
PROC IMPORT imports the data by one of the following methods:

3 generated DATA step code

3 generated SAS/ACCESS code
3 translation engines.
You control the results with statements and options that are specic to the input data
source. PROC IMPORT generates the specied output SAS data set and writes
information regarding the import to the SAS log. In the log, you see the DATA step or
the SAS/ACCESS code that is generated by PROC IMPORT. If a translation engine is
used, then no code is submitted.
Note: To import data, you can also use the Import Wizard, which is a windowing
tool that guides you through the steps to import an external data source. You can
request the Import Wizard to generate IMPORT procedure statements, which you can
save to a le for subsequent use. To invoke the Import Wizard, from the SAS
windowing environment select
File

Import Data

502

Syntax: IMPORT Procedure

Chapter 26

Syntax: IMPORT Procedure

PROC IMPORT is available for the following operating environments:
3 OpenVMS Alpha
3 UNIX

Restriction:

3 Microsoft Windows.

PROC IMPORT
DATAFILE="lename" | TABLE="tablename"
OUT=<libref.>SAS-data-set < (SAS-data-set-options)>
<DBMS=identier>< REPLACE> ;
<data-source-statement(s);>

PROC IMPORT Statement

Featured in:

All examples

PROC IMPORT
DATAFILE="lename" | TABLE="tablename"
OUT=<libref.>SAS-data-set < (SAS-data-set-options)>
<DBMS=identier>< REPLACE> ;

Required Arguments
DATAFILE="lename"

species the complete path and lename or a leref for the input PC le,
spreadsheet, or delimited external le. If you specify a leref or if the complete path
and lename does not include special characters (such as the backslash in a path),
lowercase characters, or spaces, you can omit the quotation marks. A leref is a SAS
name that is associated with the physical location of the output le. To assign a
leref, use the FILENAME statement. For more information about PC le formats,
see SAS/ACCESS for PC Files: Reference.
Featured in:

Example 1 on page 514, Example 2 on page 517, and Example 3 on

page 518
Restriction: PROC IMPORT does not support device types or access methods for

the FILENAME statement except for DISK. For example, PROC IMPORT does not
support the TEMP device type, which creates a temporary external le.
Restriction: For client/server applications: When running SAS/ACCESS software

on UNIX to access data that is stored on a PC server, you must specify the full
path and lename of the le that you want to import. The use of a leref is not
supported.
Interaction: For some input data sources like a Microsoft Excel spreadsheet, in

order to determine the data type (numeric or character) for a column, the rst

The IMPORT Procedure

PROC IMPORT Statement

503

eight rows of data are scanned and the most prevalent type of data is used. If
most of the data in the rst eight rows is missing, SAS defaults to the character
data type; any subsequent numeric data for that column becomes missing as well.
Mixed data can also create missing values. For example, if the rst eight rows
contain mostly character data, SAS assigns the column as a character data type;
any subsequent numeric data for that column becomes missing.
Restriction: PROC IMPORT can import data only if the data type is supported by
SAS. SAS supports numeric and character types of data but not, for example,
binary objects. If the data that you want to import is a type not supported by SAS,
PROC IMPORT may not be able to import it correctly. In many cases, the
procedure attempts to convert the data to the best of its ability; however, for some
types, this is not possible.
Tip: For information about how SAS converts data types, see the specic
information for the data source that you are importing in SAS/ACCESS for PC
Files: Reference. For example, see the chapter Understanding XLS Essentials for
a table that lists XLS data types and the resulting SAS variable data type and
formats.
Tip: For a DBF le, if the le was created by Microsoft Visual FoxPro, the le must
be exported by Visual FoxPro into an appropriate dBASE format in order to import
the le to SAS.
TABLE="tablename"

species the table name of the input DBMS table. If the name does not include
special characters (such as question marks), lowercase characters, or spaces, you can
omit the quotation marks. Note that the DBMS table name may be case sensitive.
Requirement: When you import a DBMS table, you must specify the DBMS=
option.
Featured in: Example 4 on page 519
OUT=<libref.>SAS-data-set

identies the output SAS data set with either a one- or two-level SAS name (library
and member name). If the specied SAS data set does not exist, PROC IMPORT
creates it. If you specify a one-level name, by default PROC IMPORT uses either the
USER library (if assigned) or the WORK library (if USER not assigned).
Featured in: All examples
(SAS-data-set-options)

species SAS data set options. For example, to assign a password to the resulting
SAS data set, you can use the ALTER=, PW=, READ=, or WRITE= data set option,
or to import only data that meets a specied condition, you can use the WHERE=
data set option. For information about all SAS data set options, see Data Set
Options in SAS Language Reference: Dictionary.
Restriction: You cannot specify data set options when importing delimited,
comma-separated, or tab-delimited external les.
Featured in: Example 3 on page 518

Options
DBMS=identier

species the type of data to import. To import a DBMS table, you must specify
DBMS= using a valid database identier. For example, DBMS=ACCESS species to
import a Microsoft Access 2000 or 2002 table. To import PC les, spreadsheets, and
delimited external les, you do not have to specify DBMS= if the lename that is

504

PROC IMPORT Statement

Chapter 26

.dbf

UNIX,
Microsoft
Windows

DLM

delimited le (default delimiter is a blank)

UNIX

JMP

JMP table

.jmp

UNIX,
Microsoft
Windows

PCFS

Files on PC server

UNIX

The IMPORT Procedure

3 the operating environment, and in some cases the platform, as specied in

the previous table.

3 whether your site has a license to the SAS/ACCESS software for PC le

formats. If you do not have a license, only delimited les are supported.
Featured in:

Example 1 on page 514 and Example 4 on page 519

When you specify a value for DBMS=, consider the following:

3 To import a Microsoft Access table, PROC IMPORT can distinguish whether the
table is in Access 97, 2000, or 2002 format regardless of your specication. For
example, if you specify DBMS=ACCESS and the table is an Access 97 table,
PROC IMPORT will import the le.

3 To import a Microsoft Excel spreadsheet, PROC IMPORT can distinguish some

versions regardless of your specication. For example, if you specify
DBMS=EXCEL and the spreadsheet is an Excel 97 spreadsheet, PROC
IMPORT can import the le. However, if you specify DBMS=EXCEL4 and the
spreadsheet is an Excel 2000 spreadsheet, PROC IMPORT cannot import the
le. The following table lists the spreadsheets and whether PROC IMPORT can
distinguish them based on the DBMS= specication:
Specication Excel 2002

Excel 2000

Excel 97

Excel 5.0

Excel 4.0

EXCEL

yes

EXCEL2002

yes

EXCEL2000

yes

EXCEL97

yes

EXCEL5

yes

EXCEL4

yes

Note: Although Excel 4.0 and Excel 5.0 spreadsheets are often
interchangeable, it is recommended that you specify the exact version.

506

Data Source Statements

Chapter 26

REPLACE

overwrites an existing SAS data set. If you do not specify REPLACE, PROC
IMPORT does not overwrite an existing data set.
Featured in: Example 1 on page 514

Data Source Statements

Featured in:

All examples

PROC IMPORT provides a variety of statements that are specic to the input data
source.

Statements for PC Files, Spreadsheets, or Delimited External Files

The following table lists the statements that are available to import PC les,
spreadsheets, and delimited external les, and it denotes which statements are valid for
a specic data source. For example, Excel spreadsheets have optional statements to
indicate whether column names are in the rst row of data or which sheet and range of
data to import, while a dBASE le (DBF) does not. For more information about PC le
formats, see SAS/ACCESS for PC Files: Reference.
Data Source

Supported Syntax

Valid Values

Default Value

CSV/TAB

GETNAMES=

YES | NO

YES

DATAROW=

1 to 32767

GUESSING ROWS=

1 to 32767

none

GETNAMES=

YES | NO

YES

DATAROW=

1 to 32767

GUESSINGROWS=

1 to 32767

none

DBF

GETDELETED=

YES | NO

WK1 / WK3 / WK4

GETNAMES=

YES | NO

YES

RANGE=

Range Name or
Absolute Range Value,
such as A1...C4

DLM

JMP

SHEET=

Sheet Name
EXCEL4 / EXCEL5

GETNAMES=

YES | NO

RANGE=

Range Name or
Absolute Range Value,
such as A1...C4

SHEET=

Sheet Name

YES

The IMPORT Procedure

Data Source Statements

Data Source

Supported Syntax

Valid Values

Default Value

EXCEL

GETNAMES=

YES | NO

YES

EXCEL97

RANGE=

EXCEL2000

SHEET=

EXCEL2002

MIXED=

Range Name or
Absolute Range Value,
such as A1...C4

507

SCANTEXT=
SCANTIME=
USEDATE=
TEXTSIZE=
DBSASLABEL=

NO
YES

Sheet Name

YES

YES | NO

YES

YES | NO

1024

YES | NO

COMPAT

YES | NO
1 to 32767
COMPAT | NONE

EXCELCS

VERSION=
SERVER=

5 | 95 | 97 | 2000
| 2002

SERVICE=

Server Name

PORT=

Service Name

RANGE=

1 to 32767

SHEET=

Range Name or
Absolute Range Value,
such as A1...C4

SCANTEXT=
SCANTIME=
USEDATE=
TEXTSIZE=
DBSASLABEL=

YES
YES
YES
1024
COMPAT

Sheet Name
YES | NO
YES | NO
YES | NO
1 to 32767
COMPAT | NONE

DATAROW=n;
starts reading data from row number n in the external le.
Default:

when GETNAMES=NO

when GETNAMES=YES (default for GETNAMES=)

Interaction: When GETNAMES=YES, DATAROW= must be equal to or greater

than 2. When GETNAMES=NO, DATAROW must be equal to or greater than 1.

DBSASLABEL=COMPAT | NONE;
When DBSASLABEL=COMPAT, the data sources column names are saved as the
corresponding SAS label names. This is the default value.
WHEN DBSASLABEL=NONE, the data sources column names are not saved
as SAS label names. SAS label names are left as nulls.
Featured in: Example 1 on page 514

DELIMITER=char | nnx;
for a delimited external le, species the delimiter that separates columns of data
in the input le. You can specify the delimiter as a single character or as a
hexadecimal value. For example, if columns of data are separated by an

508

Data Source Statements

Chapter 26

ampersand, specify DELIMITER=&. If you do not specify DELIMITER=, PROC

IMPORT assumes that the delimiter is the blank. You can replace the equal sign
with a blank.
GETDELETED=YES | NO;
for a dBASE le (DBF), indicates whether to write records to the SAS data set
that are marked for deletion but have not been purged. You can replace the equal
sign with a blank.
GETNAMES=YES | NO;
for spreadsheets and delimited external les, determines whether to generate SAS
variable names from the column names in the input les rst row of data. You
can replace the equal sign with a blank.
If you specify GETNAMES=NO or if the column names are not valid SAS
names, PROC IMPORT uses default variable names. For example, for a delimited
le, PROC IMPORT uses VAR1, VAR2, VAR3, and so on.
Note that if a column name contains special characters that are not valid in a
SAS name, such as a blank, SAS converts the character to an underscore. For
example, the column name Occupancy Code would become the variable name
Occupancy_Code.
GUESSING ROWS=1 to 3276;
scans data for its data type from row 1 to the row number that is specied.
Note: This number should be greater than the value that is specied for
DATAROW=. 4
MIXED=YES | NO;
converts numeric data values into character data values for a column that contains
mixed data types. This option is valid only while importing data from Excel. The
default is NO, which means that numeric data will be imported as missing values
in a character column. If MIXED=YES, then the engine will assign a SAS
character type for the column and convert all numeric data values to character
data values. This option is valid only while reading (importing) data into SAS.
PORT=1 to 3276;
scans data for its data type from row 1 to the row number that is specied.
Note: This number should be greater than the value that is specied for
DATAROW=. 4
TEXTSIZE=1 to 32767
species the eld length that is allowed for importing Microsoft Excel 97, 2000, or
2002 Memo elds.
RANGE="range-name | absolute-range";
subsets a spreadsheet by identifying the rectangular set of cells to import from the
specied spreadsheet. The syntax for range-name and absolute-range is native to
the le being read. You can replace the equal sign with a blank.
range-name

is a name that has been assigned to represent a range, such as

a range of cells within the spreadsheet.
Limitation: SAS supports range names up to 32 characters. If

a range name exceeds 32 characters, SAS will notify you

that the name is invalid.
Tip: For Microsoft Excel, range names do not contain special

characters such as spaces or hyphens.

The IMPORT Procedure

absolute-range

Data Source Statements

509

identies the top left cell that begins the range and the bottom
right cell that ends the range. For Excel 4.0, 5.0, and 7.0 (95),
the beginning and ending cells are separated by two periods;
that is, C9..F12 species a cell range that begins at cell C9,
ends at cell F12, and includes all the cells in between. For
Excel 97, 2000, and 2002, the beginning and ending cells are
separated by a colon that is, C9:F12.
Tip: For Excel 97, 2000, and 2002, you can include the

spreadsheet name with an absolute range, such as

range="North B$a1:d3". If you do not include the

spreadsheet name, PROC IMPORT uses the rst sheet in the

workbook or the spreadsheet name specied with SHEET=.
Default: The entire spreadsheet is selected.
Interaction: For Excel 97, 2000, and 2002 spreadsheets, when RANGE= is

specied, a spreadsheet name specied with SHEET= is ignored when the

conict occurs.
SCANTEXT=YES | NO;
scans the length of text data for a data source column and uses the length of the
longest string data that it nds as the SAS column width. However, if the
maximum length that it nds is greater than what is specied in the TEXTSIZE=
option, then the smaller value that is specied in TEXTSIZE= will be applied as
the SAS variable width.
SCANTIME=YES | NO;
scans all row values for a DATETIME data type eld and automatically
determines the TIME data type if only time values (that is, no date or datetime
values) exist in the column.
SERVER="PC-server-name";
species the name of the PC server. You must bring up the listener on the PC
server before you can establish a connection to it. You can congure the service
name, port number, maximum number of connections allowed, and use of data
encryption on your PC server. This is a required statement. Refer to your PC
server administrator for the information that is needed. Alias: SERVER_NAME=.
SERVICE="service-name";
species the service name that is dened on your service le for your client and
server machines. This statement and the PORT= statement should not be used in
the same procedure. Note that this service name must be dened on both your
UNIX machine and your PC server. Alias: SERVER_NAME=, SERVICE_NAME=.
SHEET=spreadsheet-name;
identies a particular spreadsheet in a group of spreadsheets. Use this statement
with spreadsheets that support multiple spreadsheets within a single le. The
naming convention for the spreadsheet name is native to the le being read.
Featured in: Example 2 on page 517
Default: The default depends on the type of spreadsheet. For Excel 4.0 and 5.0,

PROC IMPORT reads the rst spreadsheet in the le. For Excel 97 and later,
PROC IMPORT reads the rst spreadsheet from an ascending sort of the
spreadsheet names. To be certain that PROC IMPORT reads the desired
spreadsheet, you should identify the spreadsheet by specifying SHEET=.
Limitation: SAS supports spreadsheet names up to 31 characters. With the $

appended, the maximum length of a spreadsheet name is 32 characters.

510

Data Source Statements

Chapter 26

USEDATE=YES | NO;
If USEDATE=YES, then DATE. format is used for date/time columns in the data
source table while importing data from Excel workbook. If USEDATE=NO, then
DATETIME. format is used for date/time.
VERSION="le-version";
species the version of le that you want to create with if the le does not exist on
your PC server yet. The default version is data-source specic. For Microsoft Excel
workbook, the valid values are 2002, 2000, 97, 95 and 5, and its default value
is 97.
Note: Always quote the version value.

Note: If the le already exists in the PC server, then this value can be
ignored. 4

The IMPORT Procedure

Data Source Statements

511

Statements for DBMS Tables

The following data source statements are available to establish a connection to the
DBMS when you import a DBMS table.
Data Source

Supported Syntax

Valid Values

Default Value

ACCESS

DATABASE=

ACCESS97

DBPWD=

YES

ACCESS2000

UID=

The complete path and

lename for the MS
ACCESS database le.

ACCESS2002

PWD=
WGDB=
SCANMEMO=

Database password
User ID
User password

MEMOSIZE=

NO
1024
COMPAT

The complete path and

lename for the
Workgroup
Administration le.

DBSASLABEL=

YES

YES | NO

SCANTIME=
USEDATE=

YES | NO
YES | NO
1 to 32767
COMPAT | NONE
ACCESSCS

VERSION=

97 | 2000 | 2002

SERVER=

Server Name

SERVICE=

Service Name

PORT=

1 to 32767

DATABASE=

The complete path and

lename for the MS
ACCESS database le.

DBPWD=
UID=
PWD=
WGDB=
SCANMEMO=

Database password

2000

YES
YES
YES
1024
COMPAT

User ID
User password

MEMOSIZE=

The complete path and

le name for the
Workgroup
Administration le.

DBSASLABEL=

YES | NO

SCANTIME=
USEDATE=

YES | NO
YES | NO
1 to 32767
COMPAT | NONE

DATABASE="database";
species the complete path and lename of the database that contains the
specied DBMS table. If the database name does not contain lowercase characters,
special characters, or national characters ($, #, or @), you can omit the quotation
marks. You can replace the equal sign with a blank.

512

Data Source Statements

Chapter 26

Note: A default may be congured in the DBMS client software; however, SAS
does not generate a default value. 4
DBPWD="database password";
species a password that allows access to a database. You can replace the equal
sign with a blank.
DBSASLABEL=COMPAT | NONE;
When DBSASLABEL=COMPAT, the data sources column names are saved as the
corresponding SAS label names. This is the default value.
WHEN DBSASLABEL=NONE, the data sources column names are not saved
as SAS label names. SAS label names are left as nulls.
Featured in:

Example 1 on page 514

MEMOSIZE="eld-length";
species the eld length for importing Microsoft Access Memo elds.
Range:
Default:
Tip:
1 to 32,767
1024
To prevent Memo
elds from being
imported, you can
specify
MEMOSIZE=0
Range: 1 - 32,767 Default: 1024 Tip: To prevent Memo elds from being
imported, you can specify MEMOSIZE=0.
PORT=1 to 3276;
scans data for its data type from row 1 to the row number that is specied.
Note: This number should be greater than the value that is specied for
DATAROW=. 4
PWD="password";
species the user password used by the DBMS to validate a specic userid. If the
password does not contain lowercase characters, special characters, or national
characters, you can omit the quotation marks. You can replace the equal sign with
a blank.
Note: The DBMS client software may default to the userid and password that
were used to log in to the operating environment; SAS does not generate a default
value. 4
SCANMEMO=YES | NO;
scans the length of data for memo elds and uses the length of the longest string
data that it nds as the SAS column width. However, if the maximum length that
it nds is greater than what is specied in the MEMOSIZE= option, then the
smaller value that is specied in MEMOSIZE= will be applied as the SAS variable
width.
SCANTIME=YES | NO;
scans all row values for a DATETIME data type eld and automatically
determines the TIME data type if only time values (that is, no date or datetime
values) exist in the column.

The IMPORT Procedure

Data Source Statements

513

SERVER="PC-server-name";
species the name of the PC server. You must bring up the listener on the PC
server before you can establish a connection to it. You can congure the service
name, port number, maximum number of connections allowed, and use of data
encryption on your PC server. This is a required statement. Refer to your PC
server administrator for the information that is needed. Alias: SERVER_NAME=.
SERVICE="service-name";
species the service name that is dened on your service le for your client and
server machines. This statement and the PORT= statement should not be used in
the same procedure. Note that this service name must be dened on both your
UNIX machine and your PC server. Alias: SERVER_NAME=.
UID= "user-id";
identies the user to the DBMS. If the userid does not contain lowercase
characters, special characters, or national characters, you can omit the quotation
marks. You can replace the equal sign with a blank.
Note: The DBMS client software may default to the userid and password that
were used to log in to the operating environment; SAS does not generate a default
value. 4
WGDB= "workgroup-database-name" ;
species the workgroup (security) database name that contains the USERID and
PWD data for the DBMS. If the workgroup database name does not contain
lowercase characters, special characters, or national characters, you can omit the
quotation marks. You can replace the equal sign with a blank.
Note: A default workgroup database may be used by the DBMS; SAS does not
generate a default value. 4
USEDATE=YES | NO;
If USEDATE=YES, then DATE. format is used for date/time columns in the data
source table while importing data from Excel workbook. If USEDATE=NO, then
DATETIME. format is used for date/time.
VERSION="le-version";
species the version of le that you want to create with if the le does not exist on
your PC server yet. The default version is data-source specic. For Microsoft Excel
workbook, the valid values are 2002, 2000, 97, 95 and 5, and its default value
is 97.
Note: Always quote the version value.

Note: If the le already exists in the PC Server, this value can be ignored.

Security Levels for Microsoft Access Tables

514

Examples: IMPORT Procedure

Chapter 26

Full
Specify DBPWD=, PWD=, UID=, and WGDB=.
Each statement has a default value; however, you may nd it necessary to provide a
value for each statement explicitly.

Examples: IMPORT Procedure

Example 1: Importing a Delimited External File

Procedure features:

PROC IMPORT statement arguments:

DATAFILE=
OUT=
DBMS=
REPLACE
Data source statements:
DELIMITER=
GETNAMES=
Other features:

PRINT procedure

This example imports the following delimited external le and creates a temporary
SAS data set named WORK.MYDATA:
Region&State&Month&Expenses&Revenue
Southern&GA&JAN2001&2000&8000
Southern&GA&FEB2001&1200&6000
Southern&FL&FEB2001&8500&11000
Northern&NY&FEB2001&3000&4000
Northern&NY&MAR2001&6000&5000
Southern&FL&MAR2001&9800&13500
Northern&MA&MAR2001&1500&1000

Program

Specify the input le.

proc import datafile="C:\My Documents\myfiles\delimiter.txt"

The IMPORT Procedure

Program

515

Identify the output SAS data set.

out=mydata

Specify that the input le is a delimited external le.

dbms=dlm

Overwrite the data set if it exists.

replace;

Specify the delimiter. The DELIMITER= option species that an & (ampersand) delimits data
elds in the input le. The delimiter separates the columns of data in the input le.
delimiter=&;

Generate the variable names from the rst row of data in the input le.
getnames=yes;
run;

Print the WORK.MYDATA data set. PROC PRINT produces a simple listing.
options nodate ps=60 ls=80;
proc print data=mydata;
run;

516

SAS Log

Chapter 26

SAS Log
The SAS log displays information about the successful import. For this example,
PROC IMPORT generates a SAS DATA step, as shown in the partial log that follows.

/**********************************************************************
79
*
PRODUCT:
SAS
80
*
VERSION:
9.00
81
*
CREATOR:
External File Interface
82
*
DATE:
24JAN02
83
*
DESC:
Generated SAS Datastep Code
84
*
TEMPLATE SOURCE: (None Specified.)
85
***********************************************************************/
86
data MYDATA
;
87
%let _EFIERR_ = 0; /* set the ERROR detection macro variable */
88
infile C:\My Documents\myfiles\delimiter.txt delimiter = & MISSOVER
88 ! DSD lrecl=32767 firstobs=2 ;
89
informat Region $8. ;
90
informat State $2. ;
91
informat Month $7. ;
92
informat Expenses best32. ;
93
informat Revenue best32. ;
94
format Region $8. ;
95
format State $2. ;
96
format Month $7. ;
97
format Expenses best12. ;
98
format Revenue best12. ;
99
input
100
Region $
101
State $
102
Month $
103
Expenses
104
Revenue
105
;
106
if _ERROR_ then call symput(_EFIERR_,1); /* set ERROR detection
106! macro variable */
107
run;
NOTE: Numeric values have been converted to character
values at the places given by: (Line):(Column).
106:44
NOTE: The infile C:\My Documents\myfiles\delimiter.txt is:
File Name=C:\My Documents\myfiles\delimiter.txt,
RECFM=V,LRECL=32767
NOTE: 7 records were read from the infile C:\My
Documents\myfiles\delimiter.txt.
The minimum record length was 29.
The maximum record length was 31.
NOTE: The data set WORK.MYDATA has 7 observations and 5 variables.
NOTE: DATA statement used (Total process time):
real time
0.04 seconds
cpu time
0.05 seconds

7 rows created in MYDATA

Documents\myfiles\delimiter.txt.

NOTE: .MYDATA was successfully created.

from C:\My

The IMPORT Procedure

Program

517

Output
This output lists the output data set, MYDATA, created by PROC IMPORT from the
delimited external le.

The SAS System

Obs
1
2
3
4
5
6
7

Region

State

Southern
Southern
Southern
Northern
Northern
Southern
Northern

GA
GA
FL
NY
NY
FL
MA

Month

Expenses

Revenue

2000
1200
8500
3000
6000
9800
1500

8000
6000
11000
4000
5000
13500
1000

JAN2001
FEB2001
FEB2001
FEB2001
MAR2001
MAR2001
MAR2001

Example 2: Importing a Specic Spreadsheet from an Excel Workbook

Procedure features:

PROC IMPORT statement arguments:

DATAFILE=
OUT=
Data source statements:
SHEET=
GETNAMES=
Other features:

PRINT procedure option:

OBS=

This example imports a specic spreadsheet from an Excel workbook, which contains
multiple spreadsheets, and creates a new, permanent SAS data set named
SASUSER.ACCOUNTS.

Program

Specify the input le. The lename contains the extension .XLS, which PROC IMPORT
recognizes as identifying an Excel 2000 spreadsheet.
proc import datafile="c:\myfiles\Accounts.xls"

518

Output

Chapter 26

Identify the output SAS data set.

out=sasuser.accounts;

Import only the sheet PRICES that is contained in the le ACCOUNTS.XLS.

sheet=Prices;

Do not generate the variable names from the input le. PROC IMPORT will use default
variable names.
getnames=no;
run;

Print the SASUSER.ACCOUNTS data set. PROC PRINT produces a simple listing. The
OBS= data set option limits the output to the rst 10 observations.
proc print data=sasuser.accounts(obs=10);
run;

Output
The following output displays the rst 10 observations of the output data set,
SASUSER.ACCOUNTS:

The SAS System

OBS
1
2
3
4
5
6
7
8
9
10

Dharamsala Tea
Tibetan Barley Beer
Licorice Syrup
Chef Antons Cajun Seasoning
Chef Antons Gumbo Mix
Grandmas Boysenberry Spread
Uncle Bobs Organic Dried Pears
Northwoods Cranberry Sauce
Mishi Kobe Beef
Fish Roe

10
24
12
48
36
12
12
12
18
12

F3
boxes x 20 bags
- 12 oz bottles
- 550 ml bottles
- 6 oz jars
boxes
- 8 oz jars
- 1 lb pkgs.
- 12 oz jars
- 500 g pkgss.
- 200 ml jars

Example 3: Importing a Subset of Records from an Excel Spreadsheet

Procedure features:

PROC IMPORT statement arguments:

DATAFILE=
OUT=

18.00
19.00
10.00
22.00
21.35
25.00
30.00
40.00
97.00
31.00

The IMPORT Procedure

Example 4: Importing a Microsoft Access Table

519

This example imports a subset of an Excel spreadsheet and creates a temporary SAS
data set. The WHERE= SAS data set option is specied in order to import only a subset
of records from the Excel spreadsheet.

Program
Specify the input le.
proc import datafile=c:\Myfiles\Class.xls

Identify the output SAS data set, and request that only a subset of the records be
imported.
out=work.femaleclass (where=(sex=F));
run;

Print the new SAS data set. PROC PRINT produces a simple listing.
proc print data=work.femaleclass;
run;

Output
The following output displays the output SAS data set, WORK.FEMALECLASS:

The SAS System

Obs
1
2
3
4
5
6
7
8
9

Name
Alice
Barbara
Carol
Jane
Janet
Joyce
Judy
Louise
Mary

Sex

Age

Height

Weight

F
F
F
F
F
F
F
F
F

13
13
14
12
15
11
14
12
15

56.5
65.3
62.8
59.8
62.5
51.3
64.3
56.3
66.5

84.0
98.0
102.5
84.5
112.5
50.5
90.0
77.0
112.0

Example 4: Importing a Microsoft Access Table

Procedure features:

PROC IMPORT statement arguments:

TABLE=
OUT=
DBMS=
Data source Statements:
DATABASE=
PWD=
UID=
WGDB=

520

Program

Chapter 26

This example imports a Microsoft Access 97 table and creates a permanent SAS data
set named SASUSER.CUST. The Access table has user-level security, so it is necessary
to specify values for the PWD=, UID=, and WGDB= statements.

Program

Specify the input DBMS table name.

proc import table="customers"

Identify the output SAS data set.

out=sasuser.cust

Specify that the input le is a Microsoft Access 97 table.

dbms=access97;

Identify the user ID to the DBMS.

uid="userid";

Specify the DBMS password to access the table.

pwd="mypassword";

Specify the path and lename of the database that contains the table.
database="c:\myfiles\east.mdb";

Specify the workgroup (security) database name that contains the user ID and
password data for the Microsoft Access table.
wgdb="c:\winnt\system32\security.mdb";

Print the SASUSER.CUST data set. PROC PRINT produces a simple listing. The OBS= data
set option limits the output to the rst ve observations.
proc print data=sasuser.cust(obs=5);
run;

The IMPORT Procedure

Example 5: Importing a Specic Spreadsheet from an Excel Workbook on a PC Server

521

Output
The following output displays the rst ve observations of the output data set,
SASUSER.CUST.

The SAS System

Obs
1
2
3
4
5

Name

Street

Zipcode

David Taylor
Theo Barnes
Lydia Stirog
Anton Niroles
Cheryl Gaspar

124 Oxbow Street

2412 McAllen Avenue
12550 Overton Place
486 Gypsum Street
36 E. Broadway

72511
72513
72516
72511
72515

Example 5: Importing a Specic Spreadsheet from an Excel Workbook on a

PC Server
Procedure features:

PROC IMPORT statement arguments:

DATAFILE=
OUT=
Data Source Statements:
SERVER=
SERVICE=
SHEET=
GETNAMES=
Other features:

PRINT procedure option:

OBS=

522

Program

Chapter 26

This example imports a specic spreadsheet from an Excel workbook on a PC server,

which contains multiple spreadsheets, and creates a new, permanent SAS data set
named WORK.PRICES.

Program
proc import dbms=excelcs
datafile="c:\myfiles\Invoice.xls"
out=work.prices;
server=Sales;
service=pcfiles;
sheet=Prices;
getnames=yes;
usedate=no;
run;
proc print data=work.prices(obs=10);
run;

523

CHAPTER

27
The MEANS Procedure
Overview: MEANS Procedure 524
What Does the MEANS Procedure Do? 524
What Types of Output Does PROC MEANS Produce? 524
Syntax: MEANS Procedure 526
PROC MEANS Statement 527
BY Statement 535
CLASS Statement 536
FREQ Statement 539
ID Statement 540
OUTPUT Statement 540
TYPES Statement 546
VAR Statement 547
WAYS Statement 548
WEIGHT Statement 549
Concepts: MEANS Procedure 550
Using Class Variables 550
Using TYPES and WAYS Statements 550
Ordering the Class Values 551
Computational Resources 552
Statistical Computations: MEANS Procedure 553
Computation of Moment Statistics 553
Condence Limits 553
Students t Test 554
Quantiles 555
Results: MEANS Procedure 556
Missing Values 556
Column Width for the Output 556
The N Obs Statistic 556
Output Data Set 557
Examples: MEANS Procedure 558
Example 1: Computing Specic Descriptive Statistics 558
Example 2: Computing Descriptive Statistics with Class Variables 560
Example 3: Using the BY Statement with Class Variables 562
Example 4: Using a CLASSDATA= Data Set with Class Variables 564
Example 5: Using Multilabel Value Formats with Class Variables 567
Example 6: Using Preloaded Formats with Class Variables 570
Example 7: Computing a Condence Limit for the Mean 573
Example 8: Computing Output Statistics 575
Example 9: Computing Different Output Statistics for Several Variables 577
Example 10: Computing Output Statistics with Missing Class Variable Values
Example 11: Identifying an Extreme Value with the Output Statistics 580

578

524

Overview: MEANS Procedure

Chapter 27

Example 12: Identifying the Top Three Extreme Values with the Output Statistics
References 587

583

Overview: MEANS Procedure

What Does the MEANS Procedure Do?
The MEANS procedure provides data summarization tools to compute descriptive
statistics for variables across all observations and within groups of observations. For
example, PROC MEANS

3 calculates descriptive statistics based on moments

3 estimates quantiles, which includes the median
3 calculates condence limits for the mean
3 identies extreme values
3 performs a t test.
By default, PROC MEANS displays output. You can also use the OUTPUT statement to
store the statistics in a SAS data set.
PROC MEANS and PROC SUMMARY are very similar; see Chapter 46, The
SUMMARY Procedure, on page 1177 for an explanation of the differences.

What Types of Output Does PROC MEANS Produce?

PROC MEANS Default Output

Output 27.1 shows the default output that PROC MEANS displays. The data set
that PROC MEANS analyzes contains the integers 1 through 10. The output reports
the number of observations, the mean, the standard deviation, the minimum value, and
the maximum value. The statements that produce the output follow:
proc means data=OnetoTen;
run;

Output 27.1

The Default Descriptive Statistics

The SAS System
The MEANS Procedure
Analysis Variable : Integer

N
Mean
Std Dev
Minimum
Maximum
-----------------------------------------------------------------10
5.5000000
3.0276504
1.0000000
10.0000000
------------------------------------------------------------------

The MEANS Procedure

What Types of Output Does PROC MEANS Produce?

525

PROC MEANS Customized Output

Output 27.2 shows the results of a more extensive analysis of two variables,
MoneyRaised and HoursVolunteered. The analysis data set contains information about
the amount of money raised and the number of hours volunteered by high-school
students for a local charity. PROC MEANS uses six combinations of two categorical
variables to compute the number of observations, the mean, and the range. The rst
variable, School, has two values and the other variable, Year, has three values. For an
explanation of the program that produces the output, see Example 11 on page 580.

Output 27.2

Specied Statistics for Class Levels and Identication of Maximum Values

Summary of Volunteer Work by School and Year

The MEANS Procedure

N
School
Year Obs Variable
N
Mean
Range
----------------------------------------------------------------------------Kennedy
1992
15 MoneyRaised
15
29.0800000
39.7500000
HoursVolunteered
15
22.1333333
30.0000000
1993

MoneyRaised
HoursVolunteered

20
20

28.5660000
19.2000000

23.5600000
20.0000000

1994

MoneyRaised
HoursVolunteered

18
18

31.5794444
24.2777778

65.4400000
15.0000000

1992

MoneyRaised
HoursVolunteered

16
16

28.5450000
18.8125000

48.2700000
38.0000000

1993

MoneyRaised
HoursVolunteered

12
12

28.0500000
15.8333333

52.4600000
21.0000000

1994

Monroe

MoneyRaised
28
29.4100000
73.5300000
HoursVolunteered
28
19.1428571
26.0000000
-----------------------------------------------------------------------------

Best Results: Most Money Raised and Most Hours Worked

Obs
1
2
3
4
5
6
7
8
9
10
11
12

School

Year

_TYPE_

Kennedy
Monroe
Kennedy
Kennedy
Kennedy
Monroe
Monroe
Monroe

.
1992
1993
1994
.
.
1992
1993
1994
1992
1993
1994

0
1
1
1
2
2
3
3
3
3
3
3

_FREQ_
109
31
32
46
53
56
15
20
18
16
12
28

Most
Cash

Most
Time

Willard
Tonya
Cameron
Willard
Luther
Willard
Thelma
Bill
Luther
Tonya
Cameron
Willard

Tonya
Tonya
Amy
L.T.
Jay
Tonya
Jay
Amy
Che-Min
Tonya
Myrtle
L.T.

Money
Raised
78.65
55.16
65.44
78.65
72.22
78.65
52.63
42.23
72.22
55.16
65.44
78.65

Hours
Volunteered
40
40
31
33
35
40
35
31
33
40
26
33

In addition to the report, the program also creates an output data set (located on
page 2 of the output) that identies the students who raised the most money and who
volunteered the most time over all the combinations of School and Year and within the
combinations of School and Year:

526

Syntax: MEANS Procedure

Chapter 27

3 The rst observation in the data set shows the students with the maximum values
overall for MoneyRaised and HoursVolunteered.

3 Observations 2 through 4 show the students with the maximum values for each
year, regardless of school.
3 Observations 5 and 6 show the students with the maximum values for each school,
regardless of year.
3 Observations 7 through 12 show the students with the maximum values for each
school-year combination.

Syntax: MEANS Procedure

Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Name: Summary
Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See
Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.
Tip:

PROC MEANS < option(s)> <statistic-keyword(s)>;

BY <DESCENDING> variable-1 < < DESCENDING> variable-n><NOTSORTED>;
CLASS variable(s) </ option(s)>;
FREQ variable;
ID variable(s);
OUTPUT <OUT=SAS-data-set> <output-statistic-specication(s)>
<id-group-specication(s)> <maximum-id-specication(s)>
<minimum-id-specication(s)> </ option(s)> ;
TYPES request(s);
VAR variable(s) < / WEIGHT=weight-variable>;
WAYS list;
WEIGHT variable;

To do this

Use this statement

Calculate separate statistics for each BY group

Identify variables whose values dene subgroups for the analysis

CLASS

Identify a variable whose values represent the frequency of each

observation

FREQ

Include additional identication variables in the output data set

Create an output data set that contains specied statistics and

identication variables

OUTPUT

Identify specic combinations of class variables to use to subdivide

the data

TYPES

Identify the analysis variables and their order in the results

VAR

The MEANS Procedure

PROC MEANS Statement

To do this

Use this statement

Specify the number of ways to make unique combinations of class

variables

WAYS

Identify a variable whose values weight each observation in the

statistical calculations

WEIGHT

PROC MEANS Statement

See also: Chapter 46, The SUMMARY Procedure, on page 1177

PROC MEANS <option(s)> <statistic-keyword(s)>;

To do this

Use this option

Specify the input data set

DATA=

Disable oating point exception recovery

NOTRAP

Specify the amount of memory to use for data summarization with

class variables

SUMSIZE=

Override the SAS system option THREADS | NOTHREADS

THREADS | NOTHREADS

Control the classication levels

Specify a secondary data set that contains the combinations of
class variables to analyze

CLASSDATA=

Create all possible combinations of class variable values

COMPLETETYPES

Exclude from the analysis all combinations of class variable

values that are not in the CLASSDATA= data set

EXCLUSIVE

Use missing values as valid values to create combinations of

class variables

MISSING

Control the statistical analysis

Specify the condence level for the condence limits

ALPHA=

Exclude observations with nonpositive weights from the

analysis

EXCLNPWGTS

Specify the sample size to use for the P2 quantile estimation

method

QMARKERS=

Specify the quantile estimation method

QMETHOD=

Specify the mathematical denition used to compute quantiles

QNTLDEF=

Select the statistics

statistic-keyword

Specify the variance divisor

VARDEF=

Control the output

527

528

PROC MEANS Statement

Chapter 27

To do this

Use this option

Specify the eld width for the statistics

FW=

Specify the number of decimal places for the statistics

MAXDEC=

Suppress reporting the total number of observations for each

unique combination of the class variables

NONOBS

Suppress all displayed output

NOPRINT

Order the values of the class variables according to the

specied order

ORDER=

Display the output

Display the analysis for all requested combinations of class

variables

PRINTALLTYPES

Display the values of the ID variables

PRINTIDVARS

Control the output data set

Specify that the _TYPE_ variable contain character values.

CHARTYPE

Order the output data set by descending _TYPE_ value

DESCENDTYPES

Select ID variables based on minimum values

IDMIN

Limit the output statistics to the observations with the highest

_TYPE_ value

NWAY

Options
ALPHA=value

species the condence level to compute the condence limits for the mean. The
percentage for the condence limits is (1value)2100. For example, ALPHA=.05
results in a 95% condence limit.
Default: .05
Range:

between 0 and 1

Interaction: To compute condence limits specify the statistic-keyword CLM,

LCLM, or UCLM.
See also: Condence Limits on page 553
Featured in:

Example 7 on page 573

CHARTYPE

species that the _TYPE_ variable in the output data set is a character
representation of the binary value of _TYPE_. The length of the variable equals the
number of class variables.
Main discussion: Output Data Set on page 557
Interaction: When you specify more than 32 class variables, _TYPE_ automatically

becomes a character variable.

Featured in:

Example 10 on page 578

CLASSDATA=SAS-data-set

species a data set that contains the combinations of values of the class variables
that must be present in the output. Any combinations of values of the class variables

The MEANS Procedure

PROC MEANS Statement

529

that occur in the CLASSDATA= data set but not in the input data set appear in the
output and have a frequency of zero.
Restriction: The CLASSDATA= data set must contain all class variables. Their

data type and format must match the corresponding class variables in the input
data set.
Interaction: If you use the EXCLUSIVE option, then PROC MEANS excludes any
observation in the input data set whose combination of class variables is not in the
CLASSDATA= data set.
Tip: Use the CLASSDATA= data set to lter or to supplement the input data set.
Featured in:

Example 4 on page 564

COMPLETETYPES

creates all possible combinations of class variables even if the combination does not
occur in the input data set.
Interaction: The PRELOADFMT option in the CLASS statement ensures that
PROC MEANS writes all user-dened format ranges or values for the
combinations of class variables to the output, even when a frequency is zero.
Using COMPLETETYPES does not increase the memory requirements.
Featured in: Example 6 on page 570
Tip:

DATA=SAS-data-set

identies the input SAS data set.

Main discussion:

Input Data Sets on page 19

DESCENDTYPES

orders observations in the output data set by descending _TYPE_ value.

Alias:

DESCENDING | DESCEND

Interaction: Descending has no effect if you specify NWAY.

Use DESCENDTYPES to make the overall total (_TYPE_=0) the last

observation in each BY group.

Tip:

Example 9 on page 577

EXCLNPWGTS

excludes observations with nonpositive weight values (zero or negative) from the
analysis. By default, PROC MEANS treats observations with negative weights like
those with zero weights and counts them in the total number of observations.
Alias: EXCLNPWGT
See also: WEIGHT= on page 548 and WEIGHT Statement on page 549
EXCLUSIVE

excludes from the analysis all combinations of the class variables that are not found
in the CLASSDATA= data set.
Requirement: If a CLASSDATA= data set is not specied, then this option is
ignored.
Featured in:

Example 4 on page 564

FW=eld-width

species the eld width to display the statistics in printed or displayed output. FW=
has no effect on statistics that are saved in an output data set.
Default: 12

If PROC MEANS truncates column labels in the output, then increase the eld
width.

Tip:

530

PROC MEANS Statement

Chapter 27

Featured in:

Example 1 on page 558, Example 4 on page 564, and Example 5 on

page 567
IDMIN

species that the output data set contain the minimum value of the ID variables.
Interaction: Specify PRINTIDVARS to display the value of the ID variables in the

output.
See also: ID Statement on page 540
MAXDEC=number

species the maximum number of decimal places to display the statistics in the
printed or displayed output. MAXDEC= has no effect on statistics that are saved in
an output data set.
Default: BEST. width for columnar format, typically about 7.
Range:

0-8

Featured in:

Example 2 on page 560 and Example 4 on page 564

MISSING

considers missing values as valid values to create the combinations of class variables.
Special missing values that represent numeric values (the letters A through Z and
the underscore (_) character) are each considered as a separate value.
Default: If you omit MISSING, then PROC MEANS excludes the observations with

a missing class variable value from the analysis.

See also: SAS Language Reference: Concepts for a discussion of missing values that

have special meaning.

Featured in:

Example 6 on page 570

NONOBS

suppresses the column that displays the total number of observations for each unique
combination of the values of the class variables. This column corresponds to the
_FREQ_ variable in the output data set.
See also: The N Obs Statistic on page 556
Featured in:

Example 5 on page 567 and Example 6 on page 570

NOPRINT

See PRINT | NOPRINT on page 531.

NOTHREADS

See THREADS | NOTHREADS on page 534.

NOTRAP

disables oating point exception (FPE) recovery during data processing. By default,
PROC MEANS traps these errors and sets the statistic to missing.
In operating environments where the overhead of FPE recovery is signicant,
NOTRAP can improve performance. Note that normal SAS FPE handling is still in
effect so that PROC MEANS terminates in the case of math exceptions.
NWAY

species that the output data set contain only statistics for the observations with the
highest _TYPE_ and _WAY_ values. When you specify class variables, this
corresponds to the combination of all class variables.
Interaction: If you specify a TYPES statement or a WAYS statement, then PROC

MEANS ignores this option.

See also: Output Data Set on page 557
Featured in:

Example 10 on page 578

The MEANS Procedure

PROC MEANS Statement

531

ORDER=DATA | FORMATTED | FREQ | UNFORMATTED

species the sort order to create the unique combinations for the values of the class
variables in the output, where
DATA
orders values according to their order in the input data set.
Interaction: If you use PRELOADFMT in the CLASS statement, then the order

for the values of each class variable matches the order that PROC FORMAT
uses to store the values of the associated user-dened format. If you use the
CLASSDATA= option, then PROC MEANS uses the order of the unique values
of each class variable in the CLASSDATA= data set to order the output levels.
If you use both options, then PROC MEANS rst uses the user-dened formats
to order the output. If you omit EXCLUSIVE, then PROC MEANS appends
after the user-dened format and the CLASSDATA= values the unique values of
the class variables in the input data set based on the order in which they are
encountered.
Tip: By default, PROC FORMAT stores a format denition in sorted order. Use

the NOTSORTED option to store the values or ranges of a user dened format
in the order that you dene them.
FORMATTED
orders values by their ascending formatted values. This order depends on your
operating environment.
Alias: FMT | EXTERNAL

FREQ
orders values by descending frequency count so that levels with the most
observations are listed rst.
Interaction: For multiway combinations of the class variables, PROC MEANS

determines the order of a class variable combination from the individual class
variable frequencies.
Interaction: Use the ASCENDING option in the CLASS statement to order

values by ascending frequency count.

UNFORMATTED
orders values by their unformatted values, which yields the same order as PROC
SORT. This order depends on your operating environment.
Alias: UNFMT | INTERNAL
Default: UNFORMATTED
See also: Ordering the Class Values on page 551
PCTLDEF=

See QNTLDEF= on page 532.

PRINT | NOPRINT

species whether PROC MEANS displays the statistical analysis. NOPRINT

suppresses all the output.
Default: PRINT
Tip:

Use NOPRINT when you want to create only an OUT= output data set.

For an example of NOPRINT, see Example 8 on page 575 and

Example 12 on page 583

Featured in:

532

PROC MEANS Statement

Chapter 27

PRINTALLTYPES

displays all requested combinations of class variables (all _TYPE_ values) in the
printed or displayed output. Normally, PROC MEANS shows only the NWAY type.
Alias:

PRINTALL

Interaction: If you use the NWAY option, the TYPES statement, or the WAYS

statement, then PROC MEANS ignores this option.

Featured in:

Example 4 on page 564

PRINTIDVARS

displays the values of the ID variables in printed or displayed output.

Alias:

PRINTIDS

Interaction: Specify IDMIN to display the minimum value of the ID variables.

See also: ID Statement on page 540
QMARKERS=number

species the default number of markers to use for the P2 quantile estimation method.
The number of markers controls the size of xed memory space.
Default: The default value depends on which quantiles you request. For the median

(P50), number is 7. For the quartiles (P25 and P50), number is 25. For the
quantiles P1, P5, P10, P90, P95, or P99, number is 105. If you request several
quantiles, then PROC MEANS uses the largest value of number.
Range:

an odd integer greater than 3

Increase the number of markers above the defaults settings to improve the
accuracy of the estimate; reduce the number of markers to conserve memory and
computing time.

Tip:

Main Discussion: Quantiles on page 555

QMETHOD=OS|P2|HIST

species the method that PROC MEANS uses to process the input data when it
computes quantiles. If the number of observations is less than or equal to the
QMARKERS= value and QNTLDEF=5, then both methods produce the same results.
OS
uses order statistics. This is the same method that PROC UNIVARIATE uses.
Note: This technique can be very memory-intensive.

P2|HIST
uses the P2 method to approximate the quantile.
Default: OS
Restriction: When QMETHOD=P2, PROC MEANS will not compute weighted

quantiles.
When QMETHOD=P2, reliable estimations of some quantiles (P1,P5,P95,P99)
may not be possible for some data sets.

Tip:

Main Discussion: Quantiles on page 555

QNTLDEF=1|2|3|4|5

species the mathematical denition that PROC MEANS uses to calculate quantiles
when QMETHOD=OS. To use QMETHOD=P2, you must use QNTLDEF=5.
Default: 5
Alias:

PCTLDEF=

Main discussion: Quantile and Related Statistics on page 1345

The MEANS Procedure

PROC MEANS Statement

533

statistic-keyword(s)

species which statistics to compute and the order to display them in the output.
The available keywords in the PROC statement are
Descriptive statistic keywords
CLM

RANGE

CSS

SKEWNESS|SKEW

STDDEV|STD

KURTOSIS|KURT

STDERR

LCLM

SUM

MAX

SUMWGT

MEAN

UCLM

MIN

USS

VAR

NMISS
Quantile statistic keywords
MEDIAN|P50

Q3|P75

P90

P95

P10

P99

Q1|P25

QRANGE

Hypothesis testing keywords

PROBT

Default: N, MEAN, STD, MIN, and MAX

To compute standard error, condence limits for the mean, and the
Students t-test, you must use the default value of the VARDEF= option, which is
DF. To compute skewness or kurtosis, you must use VARDEF=N or VARDEF=DF.

Requirement:

Use CLM or both LCLM and UCLM to compute a two-sided condence limit
for the mean. Use only LCLM or UCLM, to compute a one-sided condence limit.

Tip:

The denitions of the keywords and the formulas for the

associated statistics are listed in Keywords and Formulas on page 1340.

Main discussion:
Featured in:

Example 1 on page 558 and Example 3 on page 562

SUMSIZE=value

species the amount of memory that is available for data summarization when you
use class variables. value may be one of the following:
n|nK| nM| nG
species the amount of memory available in bytes, kilobytes, megabytes, or
gigabytes, respectively. If n is 0, then PROC MEANS use the value of the SAS
system option SUMSIZE=.
MAXIMUM|MAX
species the maximum amount of memory that is available.

534

PROC MEANS Statement

Chapter 27

Default: The value of the SUMSIZE= system option.

For best results, do not make SUMSIZE= larger than the amount of physical
memory that is available for the PROC step. If additional space is needed, then
PROC MEANS uses utility les.

Tip:

See also: The SAS system option SUMSIZE= in SAS Language Reference:

Dictionary.
Main discussion: Computational Resources on page 552
THREADS | NOTHREADS

a BY statement is specied or the value of the SAS system option CPUCOUNT is

less than 2. You can use THREADS in the PROC MEANS statement to force
PROC MEANS to use parallel processing in these situations.
VARDEF=divisor

species the divisor to use in the calculation of the variance and standard deviation.
Table 27.1 on page 534 shows the possible values for divisor and associated divisors.
Table 27.1

Possible Values for VARDEF=

Value

Divisor

Formula for Divisor

degrees of freedom

number of observations

WDF

sum of weights minus one

WEIGHT | WGT

sum of weights

P (x 0

n1
n
(6i wi) 1

6i wi

CSS=divisor

CSS

The procedure computes the variance as

, where
is the corrected
x)2 . When you weight the analysis variables,
sums of squares and equals
i
CSS equals wi (xi xw )2 , where xw is the weighted mean.

Default: DF

To compute the standard error of the mean, condence limits for the
mean, or the Students t-test, use the default value of VARDEF=.
Tip: When you use the WEIGHT statement and VARDEF=DF, the variance is an
estimate of 2 , where the variance of the ith observation is var (xi) = 2 =wi and
wi is the weight for the ith observation. This yields an estimate of the variance of
an observation with unit weight.
Tip: When you use the WEIGHT statement and VARDEF=WGT, the computed
variance is asymptotically (for large n) an estimate of 2 =w, where w is the
average weight. This yields an asymptotic estimate of the variance of an
observation with average weight.
Requirement:

The MEANS Procedure

BY Statement

535

BY Statement
Produces separate statistics for each BY group.
Main discussion:

BY on page 58

BY <DESCENDING> variable-1 << DESCENDING> variable-n> <NOTSORTED>;

Required Arguments
variable

species the variable that the procedure uses to form BY groups. You can specify
more than one variable. If you omit the NOTSORTED option in the BY statement,
then the observations in the data set either must be sorted by all the variables that
you specify or must be indexed appropriately. Variables in a BY statement are called
BY variables.

Options
DESCENDING

species that the observations are sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The observations are sorted in another way, for example, chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations with the same values for the BY variables
are not contiguous, then the procedure treats each contiguous set as a separate BY
group.

Using the BY Statement with the SAS System Option NOBYLINE

If you use the BY statement with the SAS system option NOBYLINE, which
suppresses the BY line that normally appears in output that is produced with BY-group
processing, then PROC MEANS always starts a new page for each BY group. This
behavior ensures that if you create customized BY lines by putting BY-group
information in the title and suppressing the default BY lines with NOBYLINE, then the
information in the titles matches the report on the pages. (See Creating Titles That
Contain BY-Group Information on page 20 and Suppressing the Default BY Line on
page 20.)

536

CLASS Statement

Chapter 27

CLASS Statement
Species the variables whose values dene the subgroup combinations for the analysis.
You can use multiple CLASS statements.
Tip: Some CLASS statement options are also available in the PROC MEANS
statement. They affect all CLASS variables. Options that you specify in a CLASS
statement apply only to the variables in that CLASS statement.
See also: For information about how the CLASS statement groups formatted values, see
Formatted Values on page 25.
Featured in: Example 2 on page 560, Example 4 on page 564, Example 5 on page 567,
Example 6 on page 570, and Example 10 on page 578
Tip:

CLASS variable(s) < / options>;

Required Arguments
variable(s)

species one or more variables that the procedure uses to group the data. Variables
in a CLASS statement are referred to as class variables. Class variables are numeric
or character. Class variables can have continuous values, but they typically have a
few discrete values that dene levels of the variable. You do not have to sort the data
by class variables.
Interaction: Use the TYPES statement or the WAYS statement to control which
class variables that PROC MEANS uses to group the data.
Tip: To reduce the number of class variable levels, use a FORMAT statement to
combine variable values. When a format combines several internal values into one
formatted value, PROC MEANS outputs the lowest internal value.
See also: Using Class Variables on page 550

Options
ASCENDING

species to sort the class variable levels in ascending order.

Alias: ASCEND
Interaction: PROC MEANS issues a warning message if you specify both
ASCENDING and DESCENDING and ignores both options.
Featured in: Example 10 on page 578
DESCENDING

species to sort the class variable levels in descending order.

Alias: DESCEND
Interaction: PROC MEANS issues a warning message if you specify both
ASCENDING and DESCENDING and ignores both options.
EXCLUSIVE

excludes from the analysis all combinations of the class variables that are not found
in the preloaded range of user-dened formats.

The MEANS Procedure

Requirement:

CLASS Statement

537

You must specify PRELOADFMT to preload the class variable

formats.
Featured in:

Example 6 on page 570

GROUPINTERNAL

species not to apply formats to the class variables when PROC MEANS groups the
values to create combinations of class variables.
Interaction: If you specify the PRELOADFMT option, then PROC MEANS ignores

the GROUPINTERNAL option and uses the formatted values.

Interaction: If you specify the ORDER=FORMATTED option, then PROC MEANS

ignores the GROUPINTERNAL option and uses the formatted values.

This option saves computer resources when the numeric class variables contain
discrete values.

Tip:

a missing class variable value from the analysis.

SAS Language Reference: Concepts for a discussion of missing values with
special meanings.

See also:

Featured in:

Example 10 on page 578

MLF

enables PROC MEANS to use the primary and secondary format labels for a given
range or overlapping ranges to create subgroup combinations when a multilabel
format is assigned to a class variable.
You must use PROC FORMAT and the MULTILABEL option in the
VALUE statement to create a multilabel format.

Requirement:

Interaction: If you use the OUTPUT statement with MLF, then the class variable

contains a character string that corresponds to the formatted value. Because the
formatted value becomes the internal value, the length of this variable is the
number of characters in the longest format label.
Interaction: Using MLF with ORDER=FREQ may not produce the order that you

expect for the formatted values.

If you omit MLF, then PROC MEANS uses the primary format labels, which
corresponds to using the rst external format value, to determine the subgroup
combinations.

Tip:

See also: The MULTILABEL option in the VALUE statement of the FORMAT

procedure on page 440.

Featured in:

Example 5 on page 567

Note: When the formatted values overlap, one internal class variable value maps
to more than one class variable subgroup combination. Therefore, the sum of the N
statistics for all subgroups is greater than the number of observations in the data set
(the overall N statistic). 4
ORDER=DATA | FORMATTED | FREQ | UNFORMATTED

species the order to group the levels of the class variables in the output, where
DATA
orders values according to their order in the input data set.

538

CLASS Statement

Chapter 27

Interaction: If you use PRELOADFMT, then the order of the values of each class

variable matches the order that PROC FORMAT uses to store the values of the
associated user-dened format. If you use the CLASSDATA= option in the
PROC statement, then PROC MEANS uses the order of the unique values of
each class variable in the CLASSDATA= data set to order the output levels. If
you use both options, then PROC MEANS rst uses the user-dened formats to
order the output. If you omit EXCLUSIVE in the PROC statement, then PROC
MEANS appends after the user-dened format and the CLASSDATA= values
the unique values of the class variables in the input data set based on the order
in which they are encountered.
Tip: By default, PROC FORMAT stores a format denition in sorted order. Use

the NOTSORTED option to store the values or ranges of a user dened format
in the order that you dene them.
Featured in: Example 10 on page 578

FORMATTED
orders values by their ascending formatted values. This order depends on your
operating environment. If no format has been assigned to a class variable, then
the default format, BEST12., is used.
Alias: FMT | EXTERNAL
Featured in: Example 5 on page 567

FREQ
orders values by descending frequency count so that levels with the most
observations are listed rst.
Interaction: For multiway combinations of the class variables, PROC MEANS
determines the order of a level from the individual class variable frequencies.
Interaction: Use the ASCENDING option to order values by ascending frequency

count.
Featured in: Example 5 on page 567

UNFORMATTED
orders values by their unformatted values, which yields the same order as PROC
SORT. This order depends on your operating environment. This sort sequence is
particularly useful for displaying dates chronologically.
Alias: UNFMT | INTERNAL
Default: UNFORMATTED

By default, all orders except FREQ are ascending. For descending orders, use
the DESCENDING option.

Tip:

The MEANS Procedure

FREQ Statement

539

Comparison of the BY and CLASS Statements

Using the BY statement is similar to using the CLASS statement and the NWAY
option in that PROC MEANS summarizes each BY group as an independent subset of
the input data. Therefore, no overall summarization of the input data is available.
However, unlike the CLASS statement, the BY statement requires that you previously
sort BY variables.
When you use the NWAY option, PROC MEANS might encounter insufcient
memory for the summarization of all the class variables. You can move some class
variables to the BY statement. For maximum benet, move class variables to the BY
statement that are already sorted or that have the greatest number of unique values.
You can use the CLASS and BY statements together to analyze the data by the levels
of class variables within BY groups. See Example 3 on page 562.

How PROC MEANS Handles Missing Values for Class Variables

By default, if an observation contains a missing value for any class variable, then
PROC MEANS excludes that observation from the analysis. If you specify the
MISSING option in the PROC statement, then the procedure considers missing values
as valid levels for the combination of class variables.
Specifying the MISSING option in the CLASS statement allows you to control the
acceptance of missing values for individual class variables.

Computer Resources
The total of unique class values that PROC MEANS allows depends on the amount of
computer memory that is available. See Computational Resources on page 552 for
more information.
The GROUPINTERNAL option can improve computer performance because the
grouping process is based on the internal values of the class variables. If a numeric
class variable is not assigned a format and you do not specify GROUPINTERNAL, then
PROC MEANS uses the default format, BEST12., to format numeric values as
character strings. Then PROC MEANS groups these numeric variables by their
character values, which takes additional time and computer memory.

FREQ Statement
Species a numeric variable that contains the frequency of each observation.
Main discussion:

FREQ on page 61

FREQ variable;

Required Arguments
variable

540

ID Statement

Chapter 27

then SAS truncates it. If n is less than 1 or is missing, then the procedure does not
use that observation to calculate statistics.
The sum of the frequency variable represents the total number of observations.
Note: The FREQ variable does not affect how PROC MEANS identies multiple
extremes when you use the IDGROUP syntax in the OUTPUT statement. 4

ID Statement
Includes additional variables in the output data set.
See Also: Discussion of id-group-specication in OUTPUT Statement on page 540.

ID variable(s);

Required Arguments
variable(s)

identies one or more variables from the input data set whose maximum values for
groups of observations PROC MEANS includes in the output data set.
Interaction: Use IDMIN in the PROC statement to include the minimum value of
the ID variables in the output data set.
Tip: Use the PRINTIDVARS option in the PROC statement to include the value of
the ID variable in the displayed output.

Selecting the Values of the ID Variables

When you specify only one variable in the ID statement, the value of the ID variable
for a given observation is the maximum (minimum) value found in the corresponding
group of observations in the input data set. When you specify multiple variables in the
ID statement, PROC MEANS selects the maximum value by processing the variables in
the ID statement in the order that you list them. PROC MEANS determines which
observation to use from all the ID variables by comparing the values of the rst ID
variable. If more than one observation contains the same maximum (minimum) ID
value, then PROC MEANS uses the second and subsequent ID variable values as
tiebreakers. In any case, all ID values are taken from the same observation for any
given BY group or classication level within a type.
See Sorting Orders for Character Variables on page 1014 for information on how
PROC MEANS compares character values to determine the maximum value.

OUTPUT Statement
Writes statistics to a new SAS data set.
Tip: You can use multiple OUTPUT statements to create several OUT= data sets.
Featured in: Example 8 on page 575, Example 9 on page 577, Example 10 on page 578,

Example 11 on page 580, and Example 12 on page 583

The MEANS Procedure

OUTPUT Statement

541

OUTPUT <OUT=SAS-data-set> <output-statistic-specication(s)>

<id-group-specication(s)> <maximum-id-specication(s)>
<minimum-id-specication(s)> </ option(s)>;

Options
OUT=SAS-data-set

names the new output data set. If SAS-data-set does not exist, then PROC MEANS
creates it. If you omit OUT=, then the data set is named DATAn, where n is the
smallest integer that makes the name unique.
Default: DATAn

You can use data set options with the OUT= option. See Data Set Options on
page 18 for a list.

Tip:

output-statistic-specication(s)

species the statistics to store in the OUT= data set and names one or more
variables that contain the statistics. The form of the output-statistic-specication is
statistic-keyword<(variable-list)>=<name(s)>
where
statistic-keyword
species which statistic to store in the output data set. The available statistic
keywords are
Descriptive statistics keyword
CSS

RANGE

SKEWNESS|SKEW

KURTOSIS|KURT

STDDEV |STD

LCLM

STDERR

MAX

SUM

MEAN

SUMWGT

MIN

UCLM

USS

NMISS

VAR

Quantile statistics keyword

MEDIAN|P50

Q3|P75

P90

P95

P10

P99

Q1|P25

QRANGE

Hypothesis testing keyword

PROBT

542

OUTPUT Statement

Chapter 27

By default the statistics in the output data set automatically inherit the
analysis variables format, informat, and label. However, statistics computed for
N, NMISS, SUMWGT, USS, CSS, VAR, CV, T, PROBT, SKEWNESS, and
KURTOSIS will not inherit the analysis variables format because this format may
be invalid for these statistics (for example, dollar or datetime formats).
Restriction: If you omit variable and name(s), then PROC MEANS allows the

statistic-keyword only once in a single OUTPUT statement, unless you also use
the AUTONAME option.
Featured in: Example 8 on page 575, Example 9 on page 577, Example 11 on

page 580, and Example 12 on page 583

variable-list
species the names of one or more numeric analysis variables whose statistics you
want to store in the output data set.
Default: all numeric analysis variables

name(s)
species one or more names for the variables in output data set that will contain
the analysis variable statistics. The rst name contains the statistic for the rst
analysis variable; the second name contains the statistic for the second analysis
variable; and so on.
Default: the analysis variable name. If you specify AUTONAME, then the default

is the combination of the analysis variable name and the statistic-keyword.

Interaction: If you specify variable-list, then PROC MEANS uses the order in

which you specify the analysis variables to store the statistics in the output
data set variables.
Featured in: Example 8 on page 575
Default: If you use the CLASS statement and an OUTPUT statement without an

output-statistic-specication, then the output data set contains ve observations

for each combination of class variables: the value of N, MIN, MAX, MEAN, and
STD. If you use the WEIGHT statement or the WEIGHT option in the VAR
statement, then the output data set also contains an observation with the sum of
weights (SUMWGT) for each combination of class variables.
Use the AUTONAME option to have PROC MEANS generate unique names
for multiple variables and statistics.

Tip:

id-group-specication

combines the features and extends the ID statement, the IDMIN option in the PROC
statement, and the MAXID and MINID options in the OUTPUT statement to create
an OUT= data set that identies multiple extreme values. The form of the
id-group-specication is
IDGROUP (<MIN|MAX (variable-list-1) <MIN|MAX (variable-list-n)>>
<<MISSING> <OBS> < LAST>> OUT <[n]>
(id-variable-list)=< name(s)>)
MIN|MAX(variable-list)
species the selection criteria to determine the extreme values of one or more
input data set variables specied in variable-list. Use MIN to determine the
minimum extreme value and MAX to determine the maximum extreme value.
When you specify multiple selection variables, the ordering of observations for
the selection of n extremes is done the same way that PROC SORT sorts data with
multiple BY variables. PROC MEANS concatenates the variable values into a
single key. The MAX(variable-list) selection criterion is similar to using PROC
SORT and the DESCENDING option in the BY statement.

The MEANS Procedure

OUTPUT Statement

543

Default: If you do not specify MIN or MAX, then PROC MEANS uses the

observation number as the selection criterion to output observations.

Restriction: If you specify criteria that are contradictory, then PROC MEANS
uses only the rst selection criterion.
Interaction: When multiple observations contain the same extreme values in all
the MIN or MAX variables, PROC MEANS uses the observation number to
resolve which observation to write to the output. By default, PROC MEANS
uses the rst observation to resolve any ties. However, if you specify the LAST
option, then PROC MEANS uses the last observation to resolve any ties.
LAST
species that the OUT= data set contains values from the last observation (or the
last n observations, if n is specied). If you do not specify LAST, then the OUT=
data set contains values from the rst observation (or the rst n observations, if n
is specied). The OUT= data set might contain several observations because in
addition to the value of the last (rst) observation, the OUT= data set contains
values from the last (rst) observation of each subgroup level that is dened by
combinations of class variable values.
Interaction: When you specify MIN or MAX and when multiple observations
contain the same extreme values, PROC MEANS uses the observation number
to resolve which observation to save to the OUT= data set. If you specify LAST,
then PROC MEANS uses the later observations to resolve any ties. If you do
not specify LAST, then PROC MEANS uses the earlier observations to resolve
any ties.
MISSING
species that missing values be used in selection criteria.
Alias: MISS
OBS
includes an _OBS_ variable in the OUT= data set that contains the number of the
observation in the input data set where the extreme value was found.
Interaction: If you use WHERE processing, then the value of _OBS_ might not
correspond to the location of the observation in the input data set.
Interaction: If you use [n] to write multiple extreme values to the output, then
PROC MEANS creates n _OBS_ variables and uses the sufx n to create the
variable names, where n is a sequential integer from 1 to n.
[n]
species the number of extreme values for each variable in id-variable-list to
include in the OUT= data set. PROC MEANS creates n new variables and uses the
sufx _n to create the variable names, where n is a sequential integer from 1 to n.
By default, PROC MEANS determines one extreme value for each level of each
requested type. If n is greater than one, then n extremes are output for each level
of each type. When n is greater than one and you request extreme value selection,
T 3 N 2 n , where T is the number of types
the time complexity is
requested and N is the number of observations in the input data set. By
comparison, to group the entire data set, the time complexity is
N 2N .
Default: 1
Range: an integer between 1 and 100
Example: To output two minimum extreme values for each variable, use

log )

O ( log )

idgroup(min(x) out[2](x y z)=MinX MinY MinZ);

The OUT= data set contains the variables MinX_1, MinX_2, MinY_1, MinY_2,
MinZ_1, and MinZ_2.

544

OUTPUT Statement

Chapter 27

(id-variable-list)
identies one or more input data set variables whose values PROC MEANS
includes in the OUT= data set. PROC MEANS determines which observations to
output by the selection criteria that you specify (MIN, MAX, and LAST).
name(s)
species one or more names for variables in the OUT= data set.
Default: If you omit name, then PROC MEANS uses the names of variables in the

id-variable-list.
Tip: Use the AUTONAME option to automatically resolve naming conicts.
Alias:

IDGRP

You must specify the MIN|MAX selection criteria rst and

OUT(id-variable-list)= after the suboptions MISSING, OBS, and LAST.

Requirement:

You can use id-group-specication to mimic the behavior of the ID statement

and a maximum-id-specication or minimum-id-specication in the OUTPUT
statement.

Tip:

When you want the output data set to contain extreme values along with other
id variables, it is more efcient to include them in the id-variable-list than to
request separate statistics. For example, the statement

Tip:

output idgrp(max(x) out(x a b)= );

is more efcient than the statement

output idgrp(max(x) out(a b)= ) max(x)=;

Featured in:

Example 8 on page 575 and Example 12 on page 583

CAUTION:

The IDGROUP syntax allows you to create output variables with the same name. When
this happens, only the rst variable appears in the output data set. Use the
AUTONAME option to automatically resolve these naming conicts. 4
Note: If you specify fewer new variable names than the combination of analysis
variables and identication variables, then the remaining output variables use the
corresponding names of the ID variables as soon as PROC MEANS exhausts the list
of new variable names. 4
maximum-id-specication(s)

species that one or more identication variables be associated with the maximum
values of the analysis variables. The form of the maximum-id-specication is
MAXID < (variable-1 < (id-variable-list-1)> <variable-n
< (id-variable-list-n)>>)> = name(s)
variable
identies the numeric analysis variable whose maximum values PROC MEANS
determines. PROC MEANS may determine several maximum values for a variable
because, in addition to the overall maximum value, subgroup levels, which are
dened by combinations of class variables values, also have maximum values.
Tip: If you use an ID statement and omit variable, then PROC MEANS uses all

analysis variables.
id-variable-list
identies one or more variables whose values identify the observations with the
maximum values of the analysis variable.
Default: the ID statement variables

The MEANS Procedure

OUTPUT Statement

545

name(s)
species the names for new variables that contain the values of the identication
variable associated with the maximum value of each analysis variable.
Tip: If you use an ID statement, and omit variable and id-variable, then PROC
MEANS associates all ID statement variables with each analysis variable. Thus,
for each analysis variable, the number of variables that are created in the output
data set equals the number of variables that you specify in the ID statement.
Tip: Use the AUTONAME option to automatically resolve naming conicts.
Limitation: If multiple observations contain the maximum value within a class
level, then PROC MEANS saves the value of the ID variable for only the rst of
those observations in the output data set.
Featured in: Example 11 on page 580
CAUTION:

The MAXID syntax allows you to create output variables with the same name. When
this happens, only the rst variable appears in the output data set. Use the
AUTONAME option to automatically resolve these naming conicts. 4
Note: If you specify fewer new variable names than the combination of analysis
variables and identication variables, then the remaining output variables use the
corresponding names of the ID variables as soon as PROC MEANS exhausts the list
of new variable names. 4
minid-specication

See the description of maximum-id-specication on page 544. This option behaves in

exactly the same way, except that PROC MEANS determines the minimum values
instead of the maximum values. The form of the minid-specication is
MINID<(variable-1 <(id-variable-list-1)> <variable-n
< (id-variable-list-n)>>)> = name(s)

AUTOLABEL

species that PROC MEANS appends the statistic name to the end of the variable
label. If an analysis variable has no label, then PROC MEANS creates a label by
appending the statistic name to the analysis variable name.
Featured in: Example 12 on page 583
AUTONAME

species that PROC MEANS creates a unique variable name for an output statistic
when you do not explicitly assign the variable name in the OUTPUT statement. This
is accomplished by appending the statistic-keyword to the end of the input variable
name from which the statistic was derived. For example, the statement
output min(x)=/autoname;

produces the x_Min variable in the output data set.

AUTONAME activates the SAS internal mechanism to automatically resolve
conicts in the variable names in the output data set. Duplicate variables will not
generate errors. As a result, the statement
output min(x)= min(x)=/autoname;

produces two variables, x_Min and x_Min2, in the output data set.
Featured in: Example 12 on page 583
KEEPLEN

species that statistics in the output data set inherit the length of the analysis
variable that PROC MEANS uses to derive them.

546

TYPES Statement

Chapter 27

CAUTION:

You permanently lose numeric precision when the length of the analysis variable causes
PROC MEANS to truncate or round the value of the statistic. However, the precision of
the statistic will match that of the input. 4
LEVELS

includes a variable named _LEVEL_ in the output data set. This variable contains a
value from 1 to n that indicates a unique combination of the values of class variables
(the values of _TYPE_ variable).
Main discussion: Output Data Set on page 557
Featured in: Example 8 on page 575
NOINHERIT

species that the variables in the output data set that contain statistics do not
inherit the attributes (label and format) of the analysis variables which are used to
derive them.
Tip: By default, the output data set includes an output variable for each analysis
variable and for ve observations that contain N, MIN, MAX, MEAN, and
STDDEV. Unless you specify NOINHERIT, this variable inherits the format of the
analysis variable, which may be invalid for the N statistic (for example, datetime
formats).
WAYS

includes a variable named _WAY_ in the output data set. This variable contains a
value from 1 to the maximum number of class variables that indicates how many
class variables PROC MEANS combines to create the TYPE value.
Main discussion: Output Data Set on page 557
See also: WAYS Statement on page 548
Featured in: Example 8 on page 575

TYPES Statement
Identies which of the possible combinations of class variables to generate.
Main discussion: Output Data Set on page 557
Requirement:
Featured in:

CLASS statement
Example 2 on page 560, Example 5 on page 567, and Example 12 on page

583

TYPES request(s);

Required Arguments
request(s)

species which of the 2k combinations of class variables PROC MEANS uses to

create the types, where k is the number of class variables. A request is composed of
one class variable name, several class variable names separated by asterisks, or ().

The MEANS Procedure

VAR Statement

547

To request class variable combinations quickly, use a grouping syntax by placing

parentheses around several variables and joining other variables or variable
combinations. For example, the following statements illustrate grouping syntax:
Request

Equivalent to

types A*(B C);

types AB AC;

types (A

B)*(C D);

types AC AD BC BD;

types (A

B C)*D;

types AD BD C*D;

Interaction The CLASSDATA= option places constraints on the NWAY type. PROC

MEANS generates all other types as if derived from the resulting NWAY type.
Tip: Use ( ) to request the overall total (_TYPE_=0).
Tip: If you do not need all types in the output data set, then use the TYPES
statement to specify particular subtypes rather than applying a WHERE clause to
the data set. Doing so saves time and computer memory.

Order of Analyses in the Output

The analyses are written to the output in order of increasing values of the _TYPE_
variable, which is calculated by PROC MEANS. The _TYPE_ variable has a unique
value for each combination of class variables; the values are determined by how you
specify the CLASS statement, not the TYPES statement. Therefore, if you specify
class A B C;
types (A B)*C;

then the B*C analysis (_TYPE_=3) is written rst, followed by the A*C analysis
(_TYPE_=5). However, if you specify
class B A C;
types (A B)*C;

then the A*C analysis comes rst.

The _TYPE_ variable is calculated even if no output data set is requested. For more
information about the _TYPE_ variable, see Output Data Set on page 557.

VAR Statement
Identies the analysis variables and their order in the output.
If you omit the VAR statement, then PROC MEANS analyzes all numeric
variables that are not listed in the other statements. When all variables are character
variables, PROC MEANS produces a simple count of observations.
Tip: You can use multiple VAR statements.
See also: Chapter 46, The SUMMARY Procedure, on page 1177
Featured in: Example 1 on page 558
Default:

VAR variable(s) </ WEIGHT=weight-variable>;

548

WAYS Statement

Chapter 27

Required Arguments
variable(s)

identies the analysis variables and species their order in the results.

Option
WEIGHT=weight-variable

species a numeric variable whose values weight the values of the variables that are
specied in the VAR statement. The variable does not have to be an integer. If the
value of the weight variable is
Weight value

PROC MEANS

counts the observation in the total number of observations

less than 0

converts the value to zero and counts the observation in the total
number of observations

missing

excludes the observation

To exclude observations that contain negative and zero weights from the analysis,
use EXCLNPWGT. Note that most SAS/STAT procedures, such as PROC GLM,
exclude negative and zero weights by default.
The weight variable does not change how the procedure determines the range,
extreme values, or number of missing values.
Restriction: To compute weighted quantiles, use QMETHOD=OS in the PROC

statement.
Restriction: Skewness and kurtosis are not available with the WEIGHT option.

When you use the WEIGHT option, consider which value of the VARDEF=
option is appropriate. See the discussion of VARDEF= on page 534.

Tip:

Use the WEIGHT option in multiple VAR statements to specify different

weights for the analysis variables.

Tip:

Note: Prior to Version 7 of SAS, the procedure did not exclude the observations
with missing weights from the count of observations. 4

WAYS Statement
Species the number of ways to make unique combinations of class variables.
Tip:

Use the TYPES statement to specify additional combinations of class variables.

Featured in:

WAYS list;

Example 6 on page 570

The MEANS Procedure

WEIGHT Statement

549

Required Arguments
list

species one or more integers that dene the number of class variables to combine to
form all the unique combinations of class variables. For example, you can specify 2
for all possible pairs and 3 for all possible triples. The list can be specied in the
following ways:
m
m1 m2 mn
m1,m2,,mn
m TO n <BY increment>
m1,m2, TO m3 <BY increment>, m4
Range: 0 to maximum number of class variables
Example: To create the two-way types for the classication variables A, B, and C,
use
class A B C ;
ways 2;

This WAYS statement is equivalent to specifying a*b, a*c, and b*c in the TYPES
statement.
See also: WAYS option on page 546

WEIGHT Statement
Species weights for observations in the statistical calculations.
See also: For information on how to calculate weighted statistics and for an example
that uses the WEIGHT statement, see WEIGHT on page 63

WEIGHT variable;

Required Arguments
variable

species a numeric variable whose values weight the values of the analysis variables.
The values of the variable do not have to be integers. If the value of the weight
variable is
Weight value

PROC MEANS

counts the observation in the total number of observations

less than 0

converts the value to zero and counts the observation in the

total number of observations

missing

excludes the observation

550

Concepts: MEANS Procedure

Chapter 27

statement.
Restriction: Skewness and kurtosis are not available with the WEIGHT statement.
Interaction: If you use the WEIGHT= option in a VAR statement to specify a

weight variable, then PROC MEANS uses this variable instead to weight those
VAR statement variables.
When you use the WEIGHT statement, consider which value of the VARDEF=
option is appropriate. See the discussion of VARDEF= on page 534 and the
calculation of weighted statistics in Keywords and Formulas on page 1340 for
more information.

Tip:

Note: Prior to Version 7 of SAS, the procedure did not exclude the observations
with missing weights from the count of observations. 4
CAUTION:

Single extreme weight values can cause inaccurate results. When one (and only one)
weight value is many orders of magnitude larger than the other weight values (for
14
example, 49 weight values of 1 and one weight value of 1210 ), certain statistics
might not be within acceptable accuracy limits. The affected statistics are those that
are based on the second moment (such as standard deviation, corrected sum of
squares, variance, and standard error of the mean). Under certain circumstances, no
warning is written to the SAS log. 4

Concepts: MEANS Procedure

Using Class Variables

Using TYPES and WAYS Statements

The TYPES statement controls which of the available class variables PROC MEANS
uses to subgroup the data. The unique combinations of these active class variable
values that occur together in any single observation of the input data set determine the
data subgroups. Each subgroup that PROC MEANS generates for a given type is called
a level of that type. Note that for all types, the inactive class variables can still affect
the total observation count of the rejection of observations with missing values.
When you use a WAYS statement, PROC MEANS generates types that correspond to
every possible unique combination of n class variables chosen from the complete set of
class variables. For example
proc means;
class a b c d e;
ways 2 3;
run;

The MEANS Procedure

Using Class Variables

551

is equivalent to
proc means;
class a b c d e;
types a*b a*c a*d a*e b*c b*d b*e c*d c*e d*e
a*b*c a*b*d a*b*e a*c*d a*c*e a*d*e
b*c*d b*c*e c*d*e;
run;

If you omit the TYPES statement and the WAYS statement, then PROC MEANS uses
all class variables to subgroup the data (the NWAY type) for displayed output and
computes all types (2k ) for the output data set.

Ordering the Class Values

PROC MEANS determines the order of each class variable in any type by examining
the order of that class variable in the corresponding one-way type. You see the effect of
this behavior in the options ORDER=DATA or ORDER=FREQ. When PROC MEANS
subdivides the input data set into subsets, the classication process does not apply the
options ORDER=DATA or ORDER=FREQ independently for each subgroup. Instead,
one frequency and data order is established for all output based on an nonsubdivided
view of the entire data set. For example, consider the following statements:
data pets;
input Pet $ Gender $;
datalines;
dog m
dog f
dog f
dog f
cat m
cat m
cat f
;
proc means data=pets order=freq;
class pet gender;
run;

The statements produce this output.

The SAS System

The MEANS Procedure

N
Pet
Gender
Obs
--------------------------dog
f
3
m
cat

m
2
---------------------------

552

Computational Resources

Chapter 27

In the example, PROC MEANS does not list male cats before female cats. Instead, it
determines the order of gender for all types over the entire data set. PROC MEANS
found more observations for female pets (f=4, m=3).

Computational Resources
PROC MEANS employs the same memory allocation scheme across all operating
environments. When class variables are involved, PROC MEANS must keep a copy of
each unique value of each class variable in memory. You can estimate the memory
requirements to group the class variable by calculating

Nc1 (Lc1 + K ) + Nc2 (Lc2 + K ) + ::: + Ncn (Lcn + K )

where

Nci
Lci
K

is the number of unique values for the class variable

is the combined unformatted and formatted length of i

is some constant on the order of 32 bytes (64 for 64-bit architectures).

When you use the GROUPINTERNAL option in the CLASS statement,

i is simply
the unformatted length of i .
Each unique combination of class variables, 1i 2j , for a given type forms a level in
that type (see TYPES Statement on page 546). You can estimate the maximum
potential space requirements for all levels of a given type, when all combinations
actually exist in the data (a complete type), by calculating

c c

W 3 Nc1 3 Nc2 3 ::: 3 Ncn

where

Nc1 :::Ncn

is a constant based on the number of variables analyzed and the

number of statistics calculated (unless you request QMETHOD=OS
to compute the quantiles).
are the number of unique levels for the active class variables of the
given type.

Clearly, the memory requirements of the levels overwhelm those of the class variables.
For this reason, PROC MEANS may open one or more utility les and write the levels
of one or more types to disk. These types are either the primary types that PROC
MEANS built during the input data scan or the derived types.
If PROC MEANS must write partially complete primary types to disk while it
processes input data, then one or more merge passes may be required to combine type
levels in memory with those on disk. In addition, if you use an order other than DATA
for any class variable, then PROC MEANS groups the completed types on disk. For this
reason, the peak disk space requirements can be more than twice the memory
requirements for a given type.
When PROC MEANS uses a temporary work le, you will receive the following note
in the SAS log:
Processing on disk occurred during summarization.
Peak disk usage was approximately nnn Mbytes.
Adjusting SUMSIZE may improve performance.

The MEANS Procedure

Condence Limits

553

In most cases processing ends normally.

When you specify class variables in a CLASS statement, the amount of
data-dependent memory that PROC MEANS uses before it writes to a utility le is
controlled by the SAS system option and PROC option SUMSIZE=. Like the system
option SORTSIZE=, SUMSIZE= sets the memory threshold where disk-based
operations begin. For best results, set SUMSIZE= to less than the amount of real
memory that is likely to be available for the task. For efciency reasons, PROC
MEANS may internally round up the value of SUMSIZE=. SUMSIZE= has no effect
unless you specify class variables.
As an alternative, you can set the SAS system option REALMEMSIZE= in the same
way that you would set SUMSIZE=. The value of REALMEMSIZE= indicates the
amount of real (as opposed to virtual) memory that SAS can expect to allocate. PROC
MEANS determines how much data-dependent memory to use before writing to utility
les by calculating the lesser of these two values:

3 the value of REALMEMSIZE=

3 0.8*(M-U), where M is the value of MEMSIZE= and U is the amount of memory
that is already in use.
Operating Environment Information: The REALMEMSIZE= SAS system option is not
available in all operating environments. For details, see the SAS Companion for your
operating environment. 4
If PROC MEANS reports that there is insufcient memory, then increase SUMSIZE=
(or REALMEMSIZE=). A SUMSIZE= (or REALMEMSIZE=) value that is greater than
MEMSIZE= will have no effect. Therefore, you might also need to increase MEMSIZE=.
If PROC MEANS reports insufcient disk space, then increase the WORK space
allocation. See the SAS documentation for your operating environment for more
information on how to adjust your computation resource parameters.
Another way to enhance performance is by carefully applying the TYPES or WAYS
statement, limiting the computations to only those combinations of class variables that
you are interested in. In particular, signicant resource savings can be achieved by not
requesting the combination of all class variables.

Statistical Computations: MEANS Procedure

Computation of Moment Statistics

PROC MEANS uses single-pass algorithms to compute the moment statistics (such
as mean, variance, skewness, and kurtosis). See Keywords and Formulas on page
1340 for the statistical formulas.
The computational details for condence limits, hypothesis test statistics, and
quantile statistics follow.

Condence Limits
With the keywords CLM, LCLM, and UCLM, you can compute condence limits for
the mean. A condence limit is a range, constructed around the value of a sample
statistic, that contains the corresponding true population value with given probability
(ALPHA=) in repeated sampling.

554

Students t Test

Chapter 27

A two-sided 100 (1
limits

0 )% condence interval for the mean has upper and lower

s
x 6 t(10=2;n01) p
n

q P

1
where s is
n01 (xi x) and t(10=2;n01) is the (1 =2) critical value of the
Students t statistics with n 1 degrees of freedom.
A one-sided 100 (1 )% condence interval is computed as

s
x + t(10;n01) p
n
s
x 0 t(10;n01) p
n
A two-sided 100 (1
and upper limits

(upper)
(lower)

0 )% condence interval for the standard deviation has lower

n01
n01
s 2
;s 2
(10=2;n01)
(=2;n01)
2 0=2;n01) and 2=2;n01) are the (1 0 =2) and =2 critical values of the
(1
(
chi-square statistic with n 0 1 degrees of freedom. A one-sided 100 (1 0 )%
condence interval is computed by replacing =2 with .
A 100 (1 0 )% condence interval for the variance has upper and lower limits that

where

are equal to the squares of the corresponding upper and lower limits for the standard
deviation.
When you use the WEIGHT statement or WEIGHT= in a VAR statement and the
default value of VARDEF=, which is DF, the 100 (1 )% condence interval for the
weighted mean has upper and lower limits

yw 6 t(10=2)

sw
sP
n
i=1

where yw is the weighted mean, sw is the weighted standard deviation, wi is the

weight for ith observation, and t(10=2) is the (1 =2) critical value for the
Students t distribution with n 1 degrees of freedom.

Students t Test
PROC MEANS calculates the t statistic as

x 0 0
p
s= n

The MEANS Procedure

Quantiles

555

where x is the sample mean, n is the number of nonmissing values for a variable, and s
is the sample standard deviation. Under the null hypothesis, the population mean
equals 0 . When the data values are approximately normally distributed, the
probability under the null hypothesis of a t statistic as extreme as, or more extreme
than, the observed value (the p-value) is obtained from the t distribution with n 1
degrees of freedom. For large n, the t statistic is asymptotically equivalent to a z test.
When you use the WEIGHT statement or WEIGHT= in a VAR statement and the
default value of VARDEF=, which is DF, the Students t statistic is calculated as

tw =

y w 0 0
n
wi
sw =
i=1

where yw is the weighted mean, sw is the weighted standard deviation, and wi is the
weight for ith observation. The tw statistic is treated as having a Students t
distribution with n 1 degrees of freedom. If you specify the EXCLNPWGT option in
the PROC statement, then n is the number of nonmissing observations when the value
of the WEIGHT variable is positive. By default, n is the number of nonmissing
observations for the WEIGHT variable.

Quantiles
The options QMETHOD=, QNTLDEF=, and QMARKERS= determine how PROC
MEANS calculates quantiles. QNTLDEF= deals with the mathematical denition of a
quantile. See Quantile and Related Statistics on page 1345. QMETHOD= deals with
the mechanics of how PROC MEANS handles the input data. The two methods are
OS
reads all data into memory and sorts it by unique value.
P2
accumulates all data into a xed sample size that is used to approximate the
quantile.
If data set A has 100 unique values for a numeric variable X and data set B has 1000
unique values for numeric variable X, then QMETHOD=OS for data set B will take 10
times as much memory as it does for data set A. If QMETHOD=P2, then both data sets
A and B will require the same memory space to generate quantiles.
The QMETHOD=P2 technique is based on the piecewise-parabolic (P2) algorithm
invented by Jain and Chlamtac (1985). P2 is a one-pass algorithm to determine
quantiles for a large data set. It requires a xed amount of memory for each variable
for each level within the type. However, using simulation studies, reliable estimations
of some quantiles (P1, P5, P95, P99) may not be possible for some data sets such as
those with heavily tailed or skewed distributions.
If the number of observations is less than the QMARKERS= value, then
QMETHOD=P2 produces the same results as QMETHOD=OS when QNTLDEF=5. To
compute weighted quantiles, you must use QMETHOD=OS.

556

Results: MEANS Procedure

Chapter 27

Results: MEANS Procedure

Missing Values
PROC MEANS excludes missing values for the analysis variables before calculating
statistics. Each analysis variable is treated individually; a missing value for an
observation in one variable does not affect the calculations for other variables. The
statements handle missing values as follows:

3 If a class variable has a missing value for an observation, then PROC MEANS
excludes that observation from the analysis unless you use the MISSING option in
the PROC statement or CLASS statement.

3 If a BY or ID variable value is missing, then PROC MEANS treats it like any

other BY or ID variable value. The missing values form a separate BY group.

3 If a FREQ variable value is missing or nonpositive, then PROC MEANS excludes

the observation from the analysis.

3 If a WEIGHT variable value is missing, then PROC MEANS excludes the

observation from the analysis.
PROC MEANS tabulates the number of the missing values. Before the number of
missing values are tabulated, PROC MEANS excludes observations with frequencies
that are nonpositive when you use the FREQ statement and observations with weights
that are missing or nonpositive (when you use the EXCLNPWGT option) when you use
the WEIGHT statement. To report this information in the procedure output use the
NMISS statistical keyword in the PROC statement.

Column Width for the Output

You control the column width for the displayed statistics with the FW= option in the
PROC statement. Unless you assign a format to a numeric class or an ID variable,
PROC MEANS uses the value of the FW= option. When you assign a format to a
numeric class or an ID variable, PROC MEANS determines the column width directly
from the format. If you use the PRELOADFMT option in the CLASS statement, then
PROC MEANS determines the column width for a class variable from the assigned
format.

The N Obs Statistic

By default when you use a CLASS statement, PROC MEANS displays an additional
statistic called N Obs. This statistic reports the total number of observations or the
sum of the observations of the FREQ variable that PROC MEANS processes for each
class level. PROC MEANS might omit observations from this total because of missing
values in one or more class variables or because of the effect of the EXCLUSIVE option
when you use it with the PRELOADFMT option or the CLASSDATA= option. Because
of this and the exclusion of observations when the WEIGHT variable contains missing
values, there is not always a direct relationship between N Obs, N, and NMISS.
In the output data set, the value of N Obs is stored in the _FREQ_ variable. Use the
NONOBS option in the PROC statement to suppress this information in the displayed
output.

The MEANS Procedure

Output Data Set

557

Output Data Set

PROC MEANS can create one or more output data sets. The procedure does not
print the output data set. Use PROC PRINT, PROC REPORT, or another SAS reporting
tool to display the output data set.
Note: By default the statistics in the output data set automatically inherit the
analysis variables format and label. However, statistics computed for N, NMISS,
SUMWGT, USS, CSS, VAR, CV, T, PROBT, SKEWNESS, and KURTOSIS do not inherit
the analysis variables format because this format may be invalid for these statistics.
Use the NOINHERIT option in the OUTPUT statement to prevent the other statistics
from inheriting the format and label attributes. 4
The output data set can contain these variables:

3
3
3
3

the variables specied in the BY statement.

the variables specied in the ID statement.
the variables specied in the CLASS statement.
the variable _TYPE_ that contains information about the class variables. By
default _TYPE_ is a numeric variable. If you specify CHARTYPE in the PROC
statement, then _TYPE_ is a character variable. When you use more than 32 class
variables, _TYPE_ is automatically a character variable.

3 the variable _FREQ_ that contains the number of observations that a given output
level represents.

3 the variables requested in the OUTPUT statement that contain the output
statistics and extreme values.

3 the variable _STAT_ that contains the names of the default statistics if you omit
statistic keywords.

3 the variable _LEVEL_ if you specify the LEVEL option.

3 the variable _WAY_ if you specify the WAYS option.
The value of _TYPE_ indicates which combination of the class variables PROC
MEANS uses to compute the statistics. The character value of _TYPE_ is a series of
zeros and ones, where each value of one indicates an active class variable in the type.
For example, with three class variables, PROC MEANS represents type 1 as 001, type 5
as 101, and so on.
Usually, the output data set contains one observation per level per type. However, if
you omit statistical keywords in the OUTPUT statement, then the output data set
contains ve observations per level (six if you specify a WEIGHT variable). Therefore,
the total number of observations in the output data set is equal to the sum of the levels
for all the types you request multiplied by 1, 5, or 6, whichever is applicable.
If you omit the CLASS statement (_TYPE_= 0), then there is always exactly one level
of output per BY group. If you use a CLASS statement, then the number of levels for
each type that you request has an upper bound equal to the number of observations in
the input data set. By default, PROC MEANS generates all possible types. In this case
the total number of levels for each BY group has an upper bound equal to

m
k

2k 0 1

n+1
m

where is the number of class variables and is the number of observations for the
given BY group in the input data set and
is 1, 5, or 6.

558

Examples: MEANS Procedure

Chapter 27

PROC MEANS determines the actual number of levels for a given type from the
number of unique combinations of each active class variable. A single level is composed
of all input observations whose formatted class values match.
Figure 27.1 on page 558 shows the values of _TYPE_ and the number of observations
in the data set when you specify one, two, and three class variables.

The Effect of Class Variables on the OUTPUT Data Set

th
re
e
tw CLA
o
SS
C
on LAS var
ia
e
CL S v
bl
es
AS ari
ab
S
le
va
s
ria
bl
e

Figure 27.1

Number of observations
of this _TYPE_ and _WAY_
in the data set

Total number of
observations
in the data set

C B

A _WAY_

_TYPE_

Subgroup
defined by

Total

A*B

a*b

A*C

a*c

B*C

b*c

1+a+b+a*b+c

A*B*C

a*b*c

+a*c+b*c+a*b*c

Character binary
equivalent of
_TYPE_
(CHARTYPE
option)

A ,B ,C=CLASS
variables

1+a

1+a+b+a*b

a, b, c,=number of levels of A, B, C,
respectively

Examples: MEANS Procedure

Example 1: Computing Specic Descriptive Statistics

Procedure features:

PROC MEANS statement options:

statistic keywords
FW=
VAR statement

This example
3 species the analysis variables
3 computes the statistics for the specied keywords and displays them in order
3 species the eld width of the statistics.

The MEANS Procedure

Program

559

Program

Create the CAKE data set. CAKE contains data from a cake-baking contest: each
participants last name, age, score for presentation, score for taste, cake avor, and number of
cake layers. The number of cake layers is missing for two observations. The cake avor is
missing for another observation.
data cake;
input LastName $ 1-12 Age 13-14 PresentScore 16-17
TasteScore 19-20 Flavor $ 23-32 Layers 34 ;
datalines;
Orlando
27 93 80 Vanilla
1
Ramey
32 84 72 Rum
2
Goldston
46 68 75 Vanilla
1
Roe
38 79 73 Vanilla
2
Larsen
23 77 84 Chocolate .
Davis
51 86 91 Spice
3
Strickland 19 82 79 Chocolate 1
Nguyen
57 77 84 Vanilla
.
Hildenbrand 33 81 83 Chocolate 1
Byron
62 72 87 Vanilla
2
Sanders
26 56 79 Chocolate 1
Jaeger
43 66 74
1
Davis
28 69 75 Chocolate 2
Conrad
69 85 94 Vanilla
1
Walters
55 67 72 Chocolate 2
Rossburger 28 78 81 Spice
2
Matthew
42 81 92 Chocolate 2
Becker
36 62 83 Spice
2
Anderson
27 87 85 Chocolate 1
Merritt
62 73 84 Chocolate 1
;

Specify the analyses and the analysis options. The statistic keywords specify the statistics
and their order in the output. FW= uses a eld width of eight to display the statistics.
proc means data=cake n mean max min range std fw=8;

Specify the analysis variables. The VAR statement species that PROC MEANS calculate
statistics on the PresentScore and TasteScore variables.
var PresentScore TasteScore;

560

Output

Chapter 27

Specify the title.

title Summary of Presentation and Taste Scores;
run;

Output

PROC MEANS lists PresentScore rst because this is the rst variable that is specied in the
VAR statement. A eld width of eight truncates the statistics to four decimal places.

Summary of Presentation and Taste Scores

The MEANS Procedure

Variable
N
Mean
Maximum
Minimum
Range
Std Dev
-----------------------------------------------------------------------------PresentScore
20
76.1500
93.0000
56.0000
37.0000
9.3768
TasteScore
20
81.3500
94.0000
72.0000
22.0000
6.6116
------------------------------------------------------------------------------

Example 2: Computing Descriptive Statistics with Class Variables

Procedure features:

PROC MEANS statement option:

MAXDEC=
CLASS statement
TYPES statement

This example

3 analyzes the data for the two-way combination of class variables and across all
observations

3 limits the number of decimal places for the displayed statistics.

Program

The MEANS Procedure

Program

561

Create the GRADE data set. GRADE contains each students last name, gender, status of
either undergraduate (1) or graduate (2), expected year of graduation, class section (A or B),
nal exam score, and nal grade for the course.
data grade;
input Name $ 1-8 Gender $ 11 Status $13 Year $ 15-16
Section $ 18 Score 20-21 FinalGrade 23-24;
datalines;
Abbott
F 2 97 A 90 87
Branford M 1 98 A 92 97
Crandell M 2 98 B 81 71
Dennison M 1 97 A 85 72
Edgar
F 1 98 B 89 80
Faust
M 1 97 B 78 73
Greeley
F 2 97 A 82 91
Hart
F 1 98 B 84 80
Isley
M 2 97 A 88 86
Jasper
M 1 97 B 91 93
;

Generate the default statistics and specify the analysis options. Because no statistics are
specied in the PROC MEANS statement, all default statistics (N, MEAN, STD, MIN, MAX) are
generated. MAXDEC= limits the displayed statistics to three decimal places.
proc means data=grade maxdec=3;

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the Score variable.
var Score;

Specify subgroups for the analysis. The CLASS statement separates the analysis into
subgroups. Each combination of unique values for Status and Year represents a subgroup.
class Status Year;

Specify which subgroups to analyze. The TYPES statement requests that the analysis be
performed on all the observations in the GRADE data set as well as the two-way combination
of Status and Year, which results in four subgroups (because Status and Year each have two
unique values).
types () status*year;

Specify the title.

title Final Exam Grades for Student Status and Year of Graduation;
run;

562

Output

Chapter 27

Output

PROC MEANS displays the default statistics for all the observations (_TYPE_=0) and the four
class levels of the Status and Year combination (Status=1, Year=97; Status=1, Year=98;
Status=2, Year=97; Status=2, Year=98).

Final Exam Grades for Student Status and Year of Graduation

The MEANS Procedure

Analysis Variable : Score
N
Obs
N
Mean
Std Dev
Minimum
Maximum
------------------------------------------------------------------------10
10
86.000
4.714
78.000
92.000
-------------------------------------------------------------------------

Analysis Variable : Score

N
Status Year Obs
N
Mean
Std Dev
Minimum
Maximum
----------------------------------------------------------------------------1
97
3
3
84.667
6.506
78.000
91.000
98

88.333

4.041

84.000

92.000

3
3

86.667

4.163

82.000

90.000

98
1
1
81.000
.
81.000
81.000
-----------------------------------------------------------------------------

Example 3: Using the BY Statement with Class Variables

Procedure features:

PROC MEANS statement option:

statistic keywords
BY statement
CLASS statement
Other features:

SORT procedure
Data set:

GRADE on page 561

This example

3 separates the analysis for the combination of class variables within BY values
3 shows the sort order requirement for the BY statement
3 calculates the minimum, maximum, and median.

The MEANS Procedure

Program

563

Program

Sort the GRADE data set. PROC SORT sorts the observations by the variable Section.
Sorting is required in order to use Section as a BY variable in the PROC MEANS step.
proc sort data=Grade out=GradeBySection;
by section;
run;

Specify the analyses. The statistic keywords specify the statistics and their order in the
output.
proc means data=GradeBySection min max median;

Divide the data set into BY groups. The BY statement produces a separate analysis for each
value of Section.
by Section;

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the Score variable.
var Score;

Specify subgroups for the analysis. The CLASS statement separates the analysis by the
values of Status and Year. Because there is no TYPES statement in this program, analyses are
performed for each subgroup, within each BY group.
class Status Year;

Specify the titles.

title1 Final Exam Scores for Student Status and Year of Graduation;
title2 Within Each Section;
run;

564

Output

Chapter 27

Output
Final Exam Scores for Student Status and Year of Graduation
Within Each Section

---------------------------------- Section=A ----------------------------------The MEANS Procedure

Analysis Variable : Score
N
Status
Year
Obs
Minimum
Maximum
Median
--------------------------------------------------------------------1
97
1
85.0000000
85.0000000
85.0000000
98

92.0000000

2
97
3
82.0000000
90.0000000
88.0000000
---------------------------------------------------------------------

---------------------------------- Section=B ----------------------------------Analysis Variable : Score

N
Status
Year
Obs
Minimum
Maximum
Median
--------------------------------------------------------------------1
97
2
78.0000000
91.0000000
84.5000000
98

84.0000000

89.0000000

86.5000000

2
98
1
81.0000000
81.0000000
81.0000000
---------------------------------------------------------------------

Example 4: Using a CLASSDATA= Data Set with Class Variables

Procedure features:

PROC MEANS statement options:

CLASSDATA=
EXCLUSIVE
FW=
MAXDEC=
PRINTALLTYPES
CLASS statement
Data set: CAKE on page 559

This example
3 species the eld width and decimal places of the displayed statistics
3 uses only the values in CLASSDATA= data set as the levels of the combinations of
class variables
3 calculates the range, median, minimum, and maximum
3 displays all combinations of the class variables in the analysis.

The MEANS Procedure

Program

565

Program

Create the CAKETYPE data set. CAKETYPE contains the cake avors and number of layers
that must occur in the PROC MEANS output.
data caketype;
input Flavor $ 1-10
datalines;
Vanilla
1
Vanilla
2
Vanilla
3
Chocolate 1
Chocolate 2
Chocolate 3
;

Layers 12;

Specify the analyses and the analysis options. The FW= option uses a eld width of seven
and the MAXDEC= option uses zero decimal places to display the statistics. CLASSDATA= and
EXCLUSIVE restrict the class levels to the values that are in the CAKETYPE data set.
PRINTALLTYPES displays all combinations of class variables in the output.
proc means data=cake range median min max fw=7 maxdec=0
classdata=caketype exclusive printalltypes;

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the TasteScore variable.
var TasteScore;

Specify subgroups for analysis. The CLASS statement separates the analysis by the values
of Flavor and Layers. Note that these variables, and only these variables, must appear in the
CAKETYPE data set.
class flavor layers;

Specify the title.

title Taste Score For Number of Layers and Cake Flavor;
run;

566

Output

Chapter 27

Output

PROC MEANS calculates statistics for the 13 chocolate and vanilla cakes. Because the
CLASSDATA= data set contains 3 as the value of Layers, PROC MEANS uses 3 as a class value
even though the frequency is zero.

Taste Score For Number of Layers and Cake Flavor

The MEANS Procedure

Analysis Variable : TasteScore
N
Obs
Range
Median
Minimum
Maximum
----------------------------------------------13
22
80
72
94
-----------------------------------------------

Analysis Variable : TasteScore

N
Layers
Obs
Range
Median
Minimum
Maximum
---------------------------------------------------------1
8
19
82
75
94
2

3
0
.
.
.
.
----------------------------------------------------------

Analysis Variable : TasteScore

N
Flavor
Obs
Range
Median
Minimum
Maximum
------------------------------------------------------------Chocolate
8
20
81
72
92
Vanilla
5
21
80
73
94
-------------------------------------------------------------

Analysis Variable : TasteScore

N
Flavor
Layers
Obs
Range
Median
Minimum
Maximum
-----------------------------------------------------------------------Chocolate
1
5
6
83
79
85
2

Vanilla

3
0
.
.
.
.
------------------------------------------------------------------------

The MEANS Procedure

Program

567

Example 5: Using Multilabel Value Formats with Class Variables

Procedure features:

PROC MEANS statement options:

statistic keywords
FW=
NONOBS
CLASS statement options:
MLF
ORDER=
TYPES statement
Other features

FORMAT procedure
FORMAT statement
Data set: CAKE on page 559

This example
3 computes the statistics for the specied keywords and displays them in order
3 species the eld width of the statistics
3 suppresses the column with the total number of observations
3 analyzes the data for the one-way combination of cake avor and the two-way
combination of cake avor and participants age
3 assigns user-dened formats to the class variables
3 uses multilabel formats as the levels of class variables
3 orders the levels of the cake avors by the descending frequency count and orders
the levels of age by the ascending formatted values.

Program

Create the $FLVRFMT. and AGEFMT. formats. PROC FORMAT creates user-dened
formats to categorize the cake avors and ages of the participants. MULTILABEL creates a
multilabel format for Age. A multilabel format is one in which multiple labels can be assigned to
the same value, in this case because of overlapping ranges. Each value is represented in the
output for each range in which it occurs.
proc format;
value $flvrfmt
Chocolate=Chocolate
Vanilla=Vanilla

568

Program

Chapter 27

Rum,Spice=Other Flavor;
value agefmt (multilabel)
15 - 29=below 30 years
30 - 50=between 30 and 50
51 - high=over 50 years
15 - 19=15 to 19
20 - 25=20 to 25
25 - 39=25 to 39
40 - 55=40 to 55
56 - high=56 and above;
run;

Specify the analyses and the analysis options. FW= uses a eld width of six to display the
statistics. The statistic keywords specify the statistics and their order in the output. NONOBS
suppresses the N Obs column.
proc means data=cake fw=6 n min max median nonobs;

Specify subgroups for the analysis. The CLASS statements separate the analysis by values
of Flavor and Age. ORDER=FREQ orders the levels of Flavor by descending frequency count.
ORDER=FMT orders the levels of Age by ascending formatted values. MLF species that
multilabel value formats be used for Age.
class flavor/order=freq;
class age /mlf order=fmt;

Specify which subgroups to analyze. The TYPES statement requests the analysis for the
one-way combination of Flavor and the two-way combination of Flavor and Age.
types flavor flavor*age;

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the TasteScore variable.
var TasteScore;

Format the output. The FORMAT statement assigns user-dened formats to the Age and
Flavor variables for this analysis.
format age agefmt. flavor $flvrfmt.;

Specify the title.

title Taste Score for Cake Flavors and Participants Age;
run;

The MEANS Procedure

Output

569

Output

The one-way combination of class variables appears before the two-way combination. A eld
width of six truncates the statistics to four decimal places. For the two-way combination of Age
and Flavor, the total number of observations is greater than the one-way combination of Flavor.
This situation arises because of the multilabel format for age, which maps one internal value to
more than one formatted value.
The order of the levels of Flavor is based on the frequency count for each level. The order of the
levels of Age is based on the order of the user-dened formats.

Taste Score for Cake Flavors and Participants Age

The MEANS Procedure
Analysis Variable : TasteScore

Flavor
N
Min
Max
Median
-----------------------------------------------Chocolate
9
72.00
92.00
83.00
Vanilla

73.00

94.00

82.00

Other Flavor
4
72.00
91.00
82.00
------------------------------------------------

Analysis Variable : TasteScore

Flavor
Age
N
Min
Max
Median
--------------------------------------------------------------------Chocolate
15 to 19
1
79.00
79.00
79.00
20 to 25

84.00

75.00

85.00

81.00

40 to 55

72.00

92.00

82.00

56 and above

84.00

below 30 years

75.00

85.00

79.00

between 30 and 50

83.00

92.00

87.50

over 50 years

72.00

84.00

78.00

25 to 39

73.00

80.00

76.50

40 to 55

75.00

56 and above

84.00

94.00

87.00

below 30 years

80.00

between 30 and 50

73.00

75.00

74.00

over 50 years
Other Flavor

84.00

25 to 39

Vanilla

84.00

94.00

87.00

25 to 39

72.00

83.00

81.00

40 to 55

91.00

below 30 years

81.00

between 30 and 50

72.00

83.00

77.50

over 50 years

91.00

570

Example 6: Using Preloaded Formats with Class Variables

Chapter 27

Example 6: Using Preloaded Formats with Class Variables

Procedure features:

PROC MEANS statement options:

COMPLETETYPES
FW=
MISSING
NONOBS
CLASS statement options:
EXCLUSIVE
ORDER=
PRELOADFMT
WAYS statement
Other features

FORMAT procedure
FORMAT statement
Data set: CAKE on page 559

This example
3 species the eld width of the statistics
3 suppresses the column with the total number of observations
3 includes all possible combinations of class variables values in the analysis even if
the frequency is zero
3 considers missing values as valid class levels
3 analyzes the one-way and two-way combinations of class variables
3 assigns user-dened formats to the class variables
3 uses only the preloaded range of user-dened formats as the levels of class
variables
3 orders the results by the value of the formatted data.

Program

Create the LAYERFMT. and $FLVRFMT. formats. PROC FORMAT creates user-dened
formats to categorize the number of cake layers and the cake avors. NOTSORTED keeps
$FLVRFMT unsorted to preserve the original order of the format values.
proc format;
value layerfmt 1=single layer
2-3=multi-layer

The MEANS Procedure

Program

571

.=unknown;
value $flvrfmt (notsorted)
Vanilla=Vanilla
Orange,Lemon=Citrus
Spice=Spice
Rum,Mint,Almond=Other Flavor;
run;

Generate the default statistics and specify the analysis options. FW= uses a eld width of
seven to display the statistics. COMPLETETYPES includes class levels with a frequency of zero.
MISSING considers missing values valid values for all class variables. NONOBS suppresses the
N Obs column. Because no specic analyses are requested, all default analyses are performed.
proc means data=cake fw=7 completetypes missing nonobs;

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of Flavor and Layers. PRELOADFMT and EXCLUSIVE restrict the levels to the preloaded
values of the user-dened formats. ORDER=DATA orders the levels of Flavor and Layer by
formatted data values.
class flavor layers/preloadfmt exclusive order=data;

Specify which subgroups to analyze. The WAYS statement requests one-way and two-way
combinations of class variables.
ways 1 2;

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the TasteScore variable.
var TasteScore;

Format the output. The FORMAT statement assigns user-dened formats to the Flavor and
Layers variables for this analysis.
format layers layerfmt. flavor $flvrfmt.;

Specify the title.

title Taste Score For Number of Layers and Cake Flavors;
run;

572

Output

Chapter 27

Output

The one-way combination of class variables appears before the two-way combination. PROC
MEANS reports only the level values that are listed in the preloaded range of user-dened
formats even when the frequency of observations is zero (in this case, citrus). PROC MEANS
rejects entire observations based on the exclusion of any single class value in a given
observation. Therefore, when the number of layers is unknown, statistics are calculated for only
one observation. The other observation is excluded because the avor chocolate was not
included in the preloaded user-dened format for Flavor.
The order of the levels is based on the order of the user-dened formats. PROC FORMAT
automatically sorted the Layers format and did not sort the Flavor format.

Taste Score For Number of Layers and Cake Flavors

The MEANS Procedure
Analysis Variable : TasteScore

Layers
N
Mean
Std Dev
Minimum
Maximum
-------------------------------------------------------------unknown
1
84.000
.
84.000
84.000
single layer
3
83.000
9.849
75.000
94.000
multi-layer
6
81.167
7.548
72.000
91.000
-------------------------------------------------------------Analysis Variable : TasteScore
Flavor
N
Mean
Std Dev
Minimum
Maximum
-------------------------------------------------------------Vanilla
6
82.167
7.834
73.000
94.000
Citrus
0
.
.
.
.
Spice
3
85.000
5.292
81.000
91.000
Other Flavor
1
72.000
.
72.000
72.000
-------------------------------------------------------------Analysis Variable : TasteScore
Flavor
Layers
N
Mean
Std Dev
Minimum
Maximum
-----------------------------------------------------------------------------Vanilla
unknown
1
84.000
.
84.000
84.000
single layer

75.000

94.000

80.000

9.899

73.000

87.000

unknown

multi-layer

unknown

single layer

multi-layer
Other Flavor

9.849

single layer

Spice

83.000

multi-layer
Citrus

85.000

5.292

81.000

91.000

unknown

single layer

multi-layer

72.000

The MEANS Procedure

Program

573

Example 7: Computing a Condence Limit for the Mean

Procedure features:

PROC MEANS statement options:

ALPHA=
FW=
MAXDEC=
CLASS statement

This example

3 species the eld width and number of decimal places of the statistics
3 computes a two-sided 90 percent condence limit for the mean values of
MoneyRaised and HoursVolunteered for the three years of data.
If this data is representative of a larger population of volunteers, then the condence
limits provide ranges of likely values for the true population means.

Program

Create the CHARITY data set. CHARITY contains information about high-school students
volunteer work for a charity. The variables give the name of the high school, the year of the
fund-raiser, the rst name of each student, the amount of money each student raised, and the
number of hours each student volunteered. A DATA step on page 1378 creates this data set.
data charity;
input School $ 1-7 Year 9-12 Name $ 14-20 MoneyRaised 22-26
HoursVolunteered 28-29;
datalines;
Monroe 1992 Allison 31.65 19
Monroe 1992 Barry
23.76 16
Monroe 1992 Candace 21.11 5
. . . more data lines . . .
Kennedy 1994 Sid
Kennedy 1994 Will
Kennedy 1994 Morty
;

27.45 25
28.88 21
34.44 25

Specify the analyses and the analysis options. FW= uses a eld width of eight and
MAXDEC= uses two decimal places to display the statistics. ALPHA=0.1 species a 90%
condence limit, and the CLM keyword requests two-sided condence limits. MEAN and STD
request the mean and the standard deviation, respectively.
proc means data=charity fw=8 maxdec=2 alpha=0.1 clm mean std;

574

Output

Chapter 27

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of Year.
class Year;

Specify the analysis variables. The VAR statement species that PROC MEANS calculate
statistics on the MoneyRaised and HoursVolunteered variables.
var MoneyRaised HoursVolunteered;

Specify the titles.

title Confidence Limits for Fund Raising Statistics;
title2 1992-94;
run;

Output

PROC MEANS displays the lower and upper condence limits for both variables for each year.

Confidence Limits for Fund Raising Statistics

1992-94

The MEANS Procedure

N
Lower 90%
Upper 90%
Year Obs Variable
CL for Mean CL for Mean
Mean
Std Dev
----------------------------------------------------------------------------1992
31 MoneyRaised
25.21
32.40
28.80
11.79
HoursVolunteered
17.67
23.17
20.42
9.01
1993

1994

MoneyRaised
HoursVolunteered

25.17
15.86

31.58
20.02

28.37
17.94

10.69
6.94

MoneyRaised
26.73
33.78
30.26
14.23
HoursVolunteered
19.68
22.63
21.15
5.96
-----------------------------------------------------------------------------

The MEANS Procedure

Program

575

Example 8: Computing Output Statistics

Procedure features:

PROC MEANS statement option:

NOPRINT
CLASS statement
OUTPUT statement options
statistic keywords
IDGROUP
LEVELS
WAYS
Other features:

PRINT procedure
Data set:

GRADE on page 561

This example

3 suppresses the display of PROC MEANS output

3 stores the average nal grade in a new variable
3 stores the name of the student with the best nal exam scores in a new variable
3 stores the number of class variables are that are combined in the _WAY_ variable
3 stores the value of the class level in the _LEVEL_ variable
3 displays the output data set.

Program

Specify the analysis options. NOPRINT suppresses the display of all PROC MEANS output.
proc means data=Grade noprint;

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of Status and Year.
class Status Year;

576

Output

Chapter 27

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the FinalGrade variable.
var FinalGrade;

Specify the output data set options. The OUTPUT statement creates the SUMSTAT data
set and writes the mean value for the nal grade to the new variable AverageGrade. IDGROUP
writes the name of the student with the top exam score to the variable BestScore and the
observation number that contained the top score. WAYS and LEVELS write information on how
the class variables are combined.
output out=sumstat mean=AverageGrade
idgroup (max(score) obs out (name)=BestScore)
/ ways levels;
run;

Print the output data set WORK.SUMSTAT. The NOOBS option suppresses the observation
numbers.
proc print data=sumstat noobs;
title1 Average Undergraduate and Graduate Course Grades;
title2 For Two Years;
run;

Output

The rst observation contains the average course grade and the name of the student with the
highest exam score over the two-year period. The next four observations contain values for each
class variable value. The remaining four observations contain values for the Year and Status
combination. The variables _WAY_, _TYPE_, and _LEVEL_ show how PROC MEANS created
the class variable combinations. The variable _OBS_ contains the observation number in the
GRADE data set that contained the highest exam score.

Average Undergraduate and Graduate Course Grades

For Two Years

Status

Year

97
98
1
2
1
1
2
2

97
98
97
98

_WAY_

_TYPE_

_LEVEL_

_FREQ_

Average
Grade

0
1
1
1
1
2
2
2
2

0
1
1
2
2
3
3
3
3

1
1
2
1
2
1
2
3
4

10
6
4
6
4
3
3
3
1

83.0000
83.6667
82.0000
82.5000
83.7500
79.3333
85.6667
88.0000
71.0000

Best
Score
Branford
Jasper
Branford
Branford
Abbott
Jasper
Branford
Abbott
Crandell

_OBS_
2
10
2
2
1
10
2
1
3

The MEANS Procedure

Program

577

Example 9: Computing Different Output Statistics for Several Variables

Procedure features:

PROC MEANS statement options:

DESCEND
NOPRINT
CLASS statement
OUTPUT statement options:
statistic keywords
Other features:

PRINT procedure
WHERE= data set option
Data set:

GRADE on page 561

This example

3 suppresses the display of PROC MEANS output

3 stores the statistics for the class level and combinations of class variables that are
specied by WHERE= in the output data set

3 orders observations in the output data set by descending _TYPE_ value

3 stores the mean exam scores and mean nal grades without assigning new
variables names

3 stores the median nal grade in a new variable

3 displays the output data set.

Program

Specify the analysis options. NOPRINT suppresses the display of all PROC MEANS output.
DESCEND orders the observations in the OUT= data set by descending _TYPE_ value.
proc means data=Grade noprint descend;

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of Status and Year.
class Status Year;

578

Output

Chapter 27

Specify the analysis variables. The VAR statement species that PROC MEANS calculate
statistics on the Score and FinalGrade variables.
var Score FinalGrade;

Specify the output data set options. The OUTPUT statement writes the mean for Score and
FinalGrade to variables of the same name. The median nal grade is written to the variable
MedianGrade. The WHERE= data set option restricts the observations in SUMDATA. One
observation contains overall statistics (_TYPE_=0). The remainder must have a status of 1.
output out=Sumdata (where=(status=1 or _type_=0))
mean= median(finalgrade)=MedianGrade;
run;

Print the output data set WORK.SUMDATA.

proc print data=Sumdata;
title Exam and Course Grades for Undergraduates Only;
title2 and for All Students;
run;

Output

The rst three observations contain statistics for the class variable levels with a status of 1.
The last observation contains the statistics for all the observations (no subgroup). Score
contains the mean test score and FinalGrade contains the mean nal grade.

Exam and Course Grades for Undergraduates Only

and for All Students

Obs
1
2
3
4

Status
1
1
1

Year
97
98

_TYPE_
3
3
2
0

_FREQ_

Score

Final
Grade

Median
Grade

3
3
6
10

84.6667
88.3333
86.5000
86.0000

79.3333
85.6667
82.5000
83.0000

73
80
80
83

Example 10: Computing Output Statistics with Missing Class Variable Values
Procedure features:

PROC MEANS statement options:

CHARTYPE
NOPRINT
NWAY
CLASS statement options:

The MEANS Procedure

Program

579

ASCENDING
MISSING
ORDER=
OUTPUT statement
Other features:

PRINT procedure
Data set:

CAKE on page 559

This example

3 suppresses the display of PROC MEANS output

3 considers missing values as valid level values for only one class variable
3 orders observations in the output data set by the ascending frequency for a single
class variable

3
3
3
3

stores observations for only the highest _TYPE_ value

stores _TYPE_ as binary character values
stores the maximum taste score in a new variable
displays the output data set.

Program

Specify the analysis options. NWAY prints observations with the highest _TYPE_ value.
NOPRINT suppresses the display of all PROC MEANS output.
proc means data=cake nway noprint;

Specify subgroups for the analysis. The CLASS statements separate the analysis by Flavor
and Layers. ORDER=FREQ and ASCENDING order the levels of Flavor by ascending
frequency. MISSING uses missing values of Layers as a valid class level value.
class flavor /order=freq ascending;
class layers /missing;

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the TasteScore variable.
var TasteScore;

580

Output

Chapter 27

Specify the output data set options. The OUTPUT statement creates the CAKESTAT data
set and outputs the maximum value for the taste score to the new variable HighScore.
output out=cakestat max=HighScore;
run;

Print the output data set WORK.CAKESTAT.

proc print data=cakestat;
title Maximum Taste Score for Flavor and Cake Layers;
run;

Output

The CAKESTAT output data set contains only observations for the combination of both class
variables, Flavor and Layers. Therefore, the value of _TYPE_ is 3 for all observations. The
observations are ordered by ascending frequency of Flavor. The missing value in Layers is a
valid value for this class variable. PROC MEANS excludes the observation with the missing
avor because it is an invalid value for Flavor.

Maximum Taste Score for Flavor and Cake Layers

Obs
1
2
3
4
5
6
7
8
9

Flavor
Rum
Spice
Spice
Vanilla
Vanilla
Vanilla
Chocolate
Chocolate
Chocolate

Layers
2
2
3
.
1
2
.
1
2

_TYPE_
3
3
3
3
3
3
3
3
3

_FREQ_
1
2
1
1
3
2
1
5
3

High
Score
72
83
91
84
94
87
84
85
92

Example 11: Identifying an Extreme Value with the Output Statistics

Procedure features:

CLASS statement
OUTPUT statement options:
statistic keyword
MAXID
Other features:

PRINT procedure
Data set:

CHARITY on page 573

The MEANS Procedure

Program

581

This example

3 identies the observations with maximum values for two variables

3 creates new variables for the maximum values
3 displays the output data set.

Program

Specify the analyses. The statistic keywords specify the statistics and their order in the
output. CHARTYPE writes the _TYPE_ values as binary characters in the output data set
proc means data=Charity n mean range chartype;

Specify subgroups for the analysis. The CLASS statement separates the analysis by School
and Year.
class School Year;

Specify the analysis variables. The VAR statement species that PROC MEANS calculate
statistics on the MoneyRaised and HoursVolunteered variables.
var MoneyRaised HoursVolunteered;

Specify the output data set options. The OUTPUT statement writes the new variables,
MostCash and MostTime, which contain the names of the students who collected the most
money and volunteered the most time, respectively, to the PRIZE data set.
output out=Prize maxid(MoneyRaised(name)
HoursVolunteered(name))= MostCash MostTime
max= ;

Specify the title.

title Summary of Volunteer Work by School and Year;
run;

582

Output

Chapter 27

Print the WORK.PRIZE output data set.

proc print data=Prize;
title Best Results: Most Money Raised and Most Hours Worked;
run;

Output

The rst page of output shows the output from PROC MEANS with the statistics for six class
levels: one for Monroe High for the years 1992, 1993, and 1994; and one for Kennedy High for
the same three years.

Summary of Volunteer Work by School and Year

The MEANS Procedure

MoneyRaised
HoursVolunteered

20
20

28.5660000
19.2000000

23.5600000
20.0000000

1994

MoneyRaised
HoursVolunteered

18
18

31.5794444
24.2777778

65.4400000
15.0000000

1992

MoneyRaised
HoursVolunteered

16
16

28.5450000
18.8125000

48.2700000
38.0000000

1993

Monroe

MoneyRaised
HoursVolunteered

12
12

28.0500000
15.8333333

52.4600000
21.0000000

1994

MoneyRaised
28
29.4100000
73.5300000
HoursVolunteered
28
19.1428571
26.0000000
-----------------------------------------------------------------------------

The MEANS Procedure

Example 12: Identifying the Top Three Extreme Values with the Output Statistics

583

The output from PROC PRINT shows the maximum MoneyRaised and HoursVolunteered values
and the names of the students who are responsible for them. The rst observation contains the
overall results, the next three contain the results by year, the next two contain the results by
school, and the nal six contain the results by School and Year.

Best Results: Most Money Raised and Most Hours Worked

Obs
1
2
3
4
5
6
7
8
9
10
11
12

School

Year

_TYPE_

Kennedy
Monroe
Kennedy
Kennedy
Kennedy
Monroe
Monroe
Monroe

.
1992
1993
1994
.
.
1992
1993
1994
1992
1993
1994

00
01
01
01
10
10
11
11
11
11
11
11

_FREQ_
109
31
32
46
53
56
15
20
18
16
12
28

Most
Cash

Most
Time

Willard
Tonya
Cameron
Willard
Luther
Willard
Thelma
Bill
Luther
Tonya
Cameron
Willard

Tonya
Tonya
Amy
L.T.
Jay
Tonya
Jay
Amy
Che-Min
Tonya
Myrtle
L.T.

Money
Raised
78.65
55.16
65.44
78.65
72.22
78.65
52.63
42.23
72.22
55.16
65.44
78.65

Hours
Volunteered
40
40
31
33
35
40
35
31
33
40
26
33

Example 12: Identifying the Top Three Extreme Values with the Output
Statistics
Procedure features:

PROC MEANS statement option:

NOPRINT
CLASS statement
OUTPUT statement options:
statistic keywords
AUTOLABEL
AUTONAME
IDGROUP
TYPES statement
Other features:

FORMAT procedure
FORMAT statement
PRINT procedure
RENAME = data set option
Data set: CHARITY on page 573

This example
3 suppresses the display of PROC MEANS output
3 analyzes the data for the one-way combination of the class variables and across all
observations

584

Program

Chapter 27

3 stores the total and average amount of money raised in new variables
3 stores in new variables the top three amounts of money raised, the names of the
three students who raised the money, the years when it occurred, and the schools
the students attended

3 automatically resolves conicts in the variable names when names are assigned to
the new variables in the output data set

3 appends the statistic name to the label of the variables in the output data set that
contain statistics that were computed for the analysis variable.

3 assigns a format to the analysis variable so that the statistics that are computed
from this variable inherit the attribute in the output data set

3 renames the _FREQ_ variable in the output data set

3 displays the output data set and its contents.

Program

Create the YRFMT. and $SCHFMT. formats. PROC FORMAT creates user-dened formats
that assign the value of All to the missing levels of the class variables.
proc format;
value yrFmt . = " All";
value $schFmt = "All
run;

Generate the default statistics and specify the analysis options. NOPRINT suppresses
the display of all PROC MEANS output.
proc means data=Charity noprint;

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of School and Year.
class School Year;

Specify which subgroups to analyze. The TYPES statement requests the analysis across all
the observations and for each one-way combination of School and Year.
types () school year;

The MEANS Procedure

Program

585

Specify the analysis variable. The VAR statement species that PROC MEANS calculate
statistics on the MoneyRaised variable.
var MoneyRaised;

Specify the output data set options. The OUTPUT statement creates the TOP3LIST data
set. RENAME= renames the _FREQ_ variable that contains frequency count for each class level.
SUM= and MEAN= specify that the sum and mean of the analysis variable (MoneyRaised) are
written to the output data set. IDGROUP writes 12 variables that contain the top three
amounts of money raised and the three corresponding students, schools, and years.
AUTOLABEL appends the analysis variable name to the label for the output variables that
contain the sum and mean. AUTONAME resolves naming conicts for these variables.
output out=top3list(rename=(_freq_=NumberStudents))sum= mean=
idgroup( max(moneyraised) out[3] (moneyraised name
school year)=)/autolabel autoname;

Format the output. The LABEL statement assigns a label to the analysis variable
MoneyRaised. The FORMAT statement assigns user-dened formats to the Year and School
variables and a SAS dollar format to the MoneyRaised variable.
label MoneyRaised=Amount Raised;
format year yrfmt. school $schfmt.
moneyraised dollar8.2;
run;

Print the output data set WORK.TOP3LIST.

proc print data=top3list;
title1 School Fund Raising Report;
title2 Top Three Students;
run;

Display information about the TOP3LIST data set. PROC DATASETS displays the contents
of the TOP3LIST data set. NOLIST suppresses the directory listing for the WORK data library.
proc datasets library=work nolist;
contents data=top3list;
title1 Contents of the PROC MEANS Output Data Set;
run;

586

Output

Chapter 27

Output

The output from PROC PRINT shows the top three values of MoneyRaised, the names of the
students who raised these amounts, the schools the students attended, and the years when the
money was raised. The rst observation contains the overall results, the next three contain the
results by year, and the nal two contain the results by school. The missing class levels for
School and Year are replaced with the value ALL.
The labels for the variables that contain statistics that were computed from MoneyRaised
include the statistic name at the end of the label.

School Fund Raising Report

Top Three Students

Obs School
1
2
3
4
5
6

All
All
All
1992
All
1993
All
1994
Kennedy All
Monroe
All

Obs Name_1
1
2
3
4
5
6

Number
Year _TYPE_ Students

Willard
Tonya
Cameron
Willard
Luther
Willard

0
1
1
1
2
2

109
31
32
46
53
56

Money
Raised_
Sum

Money
Raised_ Money
Money
Money
Mean Raised_1 Raised_2 Raised_3

$3192.75
$892.92
$907.92
$1391.91
$1575.95
$1616.80

$29.29
$28.80
$28.37
$30.26
$29.73
$28.87

$78.65
$55.16
$65.44
$78.65
$72.22
$78.65

$72.22
$53.76
$47.33
$72.22
$52.63
$65.44

$65.44
$52.63
$42.23
$56.87
$43.89
$56.87

Name_2

Name_3

School_1 School_2 School_3 Year_1 Year_2 Year_3

Luther
Edward
Myrtle
Luther
Thelma
Cameron

Cameron
Thelma
Bill
L.T.
Jenny
L.T.

Monroe
Monroe
Monroe
Monroe
Kennedy
Monroe

Kennedy
Monroe
Monroe
Kennedy
Kennedy
Monroe

Monroe
Kennedy
Kennedy
Monroe
Kennedy
Monroe

1994
1992
1993
1994
1994
1994

1994
1992
1993
1994
1992
1993

1993
1992
1993
1994
1992
1994

The MEANS Procedure

References

Contents of the PROC MEANS Output Data Set

587

The DATASETS Procedure

Data Set Name
Member Type
Engine
Created
Last Modified
Protection
Data Set Type
Label
Data Representation
Encoding

WORK.TOP3LIST
DATA
V9
18:59 Thursday, March 14, 2002
18:59 Thursday, March 14, 2002

WINDOWS
wlatin1

Observations
Variables
Indexes
Observation Length
Deleted Observations
Compressed
Sorted

6
18
0
144
0
NO
NO

Western (Windows)

Engine/Host Dependent Information

Data Set Page Size
Number of Data Set Pages
First Data Page
Max Obs per Page
Obs in First Data Page
Number of Data Set Repairs
File Name
Release Created
Host Created

12288
1
1
85
6
0
filename
9.0000B0
WIN_PRO

Alphabetic List of Variables and Attributes

#
7
8
9
6
5
10
11
12
4
1
13
14
15
2
16
17
18
3

Variable

Type

Len

MoneyRaised_1
MoneyRaised_2
MoneyRaised_3
MoneyRaised_Mean
MoneyRaised_Sum
Name_1
Name_2
Name_3
NumberStudents
School
School_1
School_2
School_3
Year
Year_1
Year_2
Year_3
_TYPE_

Num
Num
Num
Num
Num
Char
Char
Char
Num
Char
Char
Char
Char
Num
Num
Num
Num
Num

8
8
8
8
8
7
7
7
8
7
7
7
7
8
8
8
8
8

Format

Label

DOLLAR8.2
DOLLAR8.2
DOLLAR8.2
DOLLAR8.2
DOLLAR8.2

Amount
Amount
Amount
Amount
Amount

Raised
Raised
Raised
Raised_Mean
Raised_Sum

$SCHFMT.
$SCHFMT.
$SCHFMT.
$SCHFMT.
YRFMT.
YRFMT.
YRFMT.
YRFMT.

See the TEMPLATE procedure in SAS Output Delivery System: Users Guide for an
example of how to create a custom table denition for this output data set.

References
Jain R. and Chlamtac I., (1985) The P2 Algorithm for Dynamic Calculation of
Quantiles and Histograms Without Sorting Observations, Communications of the
Association of Computing Machinery, 28:10.

588

589

CHAPTER

28
The MIGRATE Procedure
Information about the MIGRATE Procedure

589

Information about the MIGRATE Procedure

The MIGRATE procedure is available specically for migrating a SAS data library
from a previous release to the most recent release. For complete information about
PROC MIGRATE, see the Migration Community at http://support.sas.com/rnd/
migration.

See:

590

591

CHAPTER

29
The OPTIONS Procedure
Overview: OPTIONS Procedure 591
What Does the OPTIONS Procedure Do? 591
What Types of Output Does PROC OPTIONS Produce? 591
Displaying the Settings of a Group of Options 593
Syntax: OPTIONS Procedure 595
PROC OPTIONS Statement 595
Results: OPTIONS Procedure 596
Examples: OPTIONS Procedure 597
Example 1: Producing the Short Form of the Options Listing
Example 2: Displaying the Setting of a Single Option 598

597

Overview: OPTIONS Procedure

What Does the OPTIONS Procedure Do?
The OPTIONS procedure lists the current settings of SAS system options. The
results are displayed in the SAS log.
SAS system options control how the SAS System formats output, handles les,
processes data sets, interacts with the operating environment, and does other tasks
that are not specic to a single SAS program or data set. You can change the settings of
SAS system options

3
3
3
3
3
3

in the SAS command

in a conguration or autoexec le
in the SAS OPTIONS statement
by using the OPTLOAD and OPTSAVE procedures
through the SAS System Options window
in other ways, depending on your operating environment. See the companion for
your operating environment for details.

For information about SAS system options, see the section on SAS system options in
SAS Language Reference: Dictionary.

What Types of Output Does PROC OPTIONS Produce?

The log that results from running PROC OPTIONS can show both the portable and
host system options, their settings, and short descriptions.

592

What Types of Output Does PROC OPTIONS Produce?

Chapter 29

The following example shows a partial log that displays the settings of portable
options.
proc options;
run;

Output 29.1

Log Showing a Partial Listing of SAS System Options

Portable Options:
APPLETLOC=(system-specific pathname)
Location of Java applets
ARMAGENT=
ARM Agent to use to collect ARM records
ARMLOC=ARMLOC.LOG Identify location where ARM records are to be written
ARMSUBSYS=(ARM_NONE)
Enable/Disable ARMing of SAS subsystems
NOASYNCHIO
Do not enable asynchronous input/output
AUTOSAVELOC=
Identifies the location where program editor contents are
auto saved
NOAUTOSIGNON
SAS/CONNECT remote submit will not automatically attempt
to SIGNON
NOBATCH
Do not use the batch set of default values for SAS system
options
BINDING=DEFAULT
Controls the binding edge for duplexed output
BOTTOMMARGIN=0.000
Bottom margin for printed output
BUFNO=1
Number of buffers for each SAS data set
BUFSIZE=0
Size of buffer for page of SAS data set
BYERR
Set the error flag if a null data set is input to the SORT
procedure
BYLINE
Print the by-line at the beginning of each by-group
BYSORTED
Require SAS data set observations to be sorted for BY
processing
NOCAPS
Do not translate source input to uppercase
NOCARDIMAGE
Do not process SAS source and data lines as 80-byte records
CATCACHE=0
Number of SAS catalogs to keep in cache memory
CBUFNO=0
Number of buffers to use for each SAS catalog
CENTER
Center SAS procedure output
NOCHARCODE
Do not use character combinations as substitute for
special characters not on the keyboard
CLEANUP
Attempt recovery from out-of-resources condition
NOCMDMAC
Do not support command-style macros
CPMLIB=
Identify previously compiled libraries of CMP subroutines
to use when linking
CMPOPT=(NOEXTRAMATH NOMISSCHECK NOPRECISE NOGUARDCHECK)
Enable SAS compiler performance optimizations
NOCOLLATE
Do not collate multiple copies of printed output
COLORPRINTING
Print in color if printer supports color
COMAMID=TCP
Specifies the communication access method to be used for
SAS distributed products
COMPRESS=NO
Specifies whether to compress observations in output SAS
data sets

To view the setting of a particular option, you can use the option parameter on PROC
OPTIONS. The following example shows a log that PROC OPTIONS produces for a
single SAS system option.
options pagesize=60;
proc options option=pagesize;
run;

The OPTIONS Procedure

Output 29.2
25
26
27

Displaying the Settings of a Group of Options

The Setting of a Single SAS System Option

options pagesize=60;
proc options option=pagesize;
run;
SAS (r) Proprietary Software Release XXX

PAGESIZE=60

Number of lines printed per page of output

Displaying the Settings of a Group of Options

You can display the settings of a group of SAS system options that have a specic
functionality, such as error handling, by using the GROUP= option.
proc options group=errorhandling;
run;

Output 29.3
6
7

Sample Output Using the GROUP= Option

proc options group=errorhandling;

run;
SAS (r) Proprietary Software Release XXX
BYERR

Set the error flag if a null data set is input to the SORT
procedure
CLEANUP
Attempt recovery from out-of-resources condition
NODMSSYNCHK
Do not enable syntax check, in windowing mode, for a
submitted statement block
DSNFERR
Generate error when SAS data set not found condition occurs
NOERRORABEND
Do not abend on error conditions
NOERRORBYABEND
Do not abend on By-group error condition
ERRORCHECK=NORMAL Level of special error processing to be performed
ERRORS=20
Maximum number of observations for which complete error
messages are printed
FMTERR
Treat missing format or informat as an error
QUOTELENMAX
Enable warning for quoted string length max
VNFERR
Treat variable not found on _NULL_ SAS data set as an error

The following table lists the values that are available when you use the GROUP=
option with PROC OPTIONS.
Values for Use with GROUP=
COMMUNICATIONS

GRAPHICS

MACRO

DATAQUALITY

HELP

MEMORY

INPUTCONTROL

Displaying the Settings of a Group of Options

Chapter 29

Values for Use with GROUP=

ERRORHANDLING

LISTCONTROL

SASFILES

EXECMODES

LOG_LISTCONTROL

SORT

EXTFILES

LOGCONTROL

The following table lists operating environmentspecic values that might be

available when you use the GROUP= option with PROC OPTIONS.
Possible Operating EnvironmentSpecic Values for Use with GROUP=
ADABAS

IDMS

ORACLE

DATACOM

IMS

REXX

DB2

ISPF

Operating Environment Information: Refer to the SAS documentation for your

operating environment for more information about these host-specic options. 4

The OPTIONS Procedure

PROC OPTIONS Statement

595

Syntax: OPTIONS Procedure

See:

OPTIONS procedure in the documentation for your operating environment.

PROC OPTIONS < option(s)>;

PROC OPTIONS Statement

PROC OPTIONS < option(s)>;

To do this

Use this option

Choose the format of the listing

Specify the long form

LONG

Specify the short form

SHORT

Display the options description, type and group

DEFINE

Display the options value and scope

VALUE

Restrict the number of options displayed

Display options belonging to a group

GROUP=

Display host options only

HOST

Display portable options only

NOHOST | PORT

Display a single option

OPTION=

Options
DEFINE

displays the short description of the option, the option group, and the option type. It
displays information about when the option can be set, whether an option can be
restricted, and whether the PROC OPTSAVE will save the option.
Interaction: This option has no effect when SHORT is specied.
GROUP=group-name

displays the options in the group specied by group-name. For more information on
options groups, see Displaying the Settings of a Group of Options on page 593.
HOST | NOHOST

displays only host options (HOST) or displays only portable options (NOHOST).
Alias:

PORTABLE is an alias for NOHOST.

596

Results: OPTIONS Procedure

Chapter 29

LONG | SHORT

species the format for displaying the settings of the SAS system options. LONG
lists each option on a separate line with a description; SHORT produces a
compressed listing without the descriptions.
Default: LONG
Featured in:

Example 1 on page 597

NOHOST | PORT

See HOST | NOHOST on page 595.

OPTION=option-name

displays a short description and the value (if any) of the option specied by
option-name. DEFINE and VALUE provide additional information about the option.
option-name
species the option to use as input to the procedure.
If a SAS system option uses an equals sign, such as PAGESIZE=, do
not include the equals sign when specifying the option to OPTION=.

Requirement:
Featured in:

Example 2 on page 598

SHORT

See LONG | SHORT.

VALUE

displays the option value and scope, as well as how the value was set.
Interaction: This option has no effect when SHORT is specied.

Note: SAS options that are passwords, such as EMAILPW and METAPASS,
return the value xxxxxxxx and not the actual password. 4

Results: OPTIONS Procedure

SAS writes the options list to the SAS log. SAS system options of the form option |
NOoption are listed as either option or NOoption, depending on the current setting, but
they are always sorted by the positive form. For example, NOCAPS would be listed
under the Cs.
Operating Environment Information: PROC OPTIONS produces additional
information that is specic to the environment under which you are running the SAS
System. Refer to the SAS documentation for your operating environment for more
information about this and for descriptions of host-specic options. 4

The OPTIONS Procedure

Program

597

Examples: OPTIONS Procedure

Example 1: Producing the Short Form of the Options Listing

Procedure features:

PROC OPTIONS statement option:

SHORT

This example shows how to generate the short form of the listing of SAS system
option settings. Compare this short form with the long form that is shown in
Overview: OPTIONS Procedure on page 591.

Program

List all options and their settings. SHORT lists the SAS system options and their settings
without any descriptions.
proc options short;
run;

598

Log (partial)

Chapter 29

Log (partial)
1
2

proc options short;

run;
SAS (r) Proprietary Software Release XXX

Portable Options:
APPLETLOC=(system-specific pathname) ARMAGENT= ARMLOC=ARMLOC.LOG ARMSUBSYS=
(ARM_NONE) NOASYNCHIO AUTOSAVELOC= NOAUTOSIGNON NOBATCH BINDING=DEFAULT
BOTTOMMARGIN=0.000 IN BUFNO=1 BUFSIZE=0 BYERR BYLINE BYSORTED NOCAPS
NOCARDIMAGE CATCACHE=0 CBUFNO=0 CENTER NOCHARCODE CLEANUP NOCMDMACCMPLIB=
CMPOPT=(NOEXTRAMATH NOMISSCHECK NOPRECISE NOGUARDCHECK) NOCOLLATE COLORPRINTING
COMAMID=TCP COMPRESS=NO CONNECTPERSIST CONNECTREMOTE= CONNECTSTATUS CONNECTWAIT
CONSOLELOG= COPIES=1 CPUCOUNT=1 CPUID DATASTMTCHK=COREKEYWORDS DATE DATESTYLE=MDY
DBSLICEPARM=(THREADED_APPS, 2) DBSRVTP=NONE NODETAILS DEVICE= DFLANG=ENGLISH
DKRICOND=ERROR DKROCOND=WARN DLDMGACTION=REPAIR NODMR DMS NODMSEXP DMSLOGSIZE=99999
DMSOUTSIZE=99999 NODMSSYNCHK DQLOCALE= DQSETUPLOC= DSNFERR NODTRESET NODUPLEX
NOECHOAUTO EMAILAUTHPROTOCOL=NONE EMAILHOST=LOCALHOST EMAILID= EMAILPORT=25 EMAILPW=
ENGINE=V9 NOERRORABEND NOERRORBYABEND ERRORCHECK=NORMAL ERRORS=20 NOEXPLORER
FIRSTOBS=1 FMTERR FMTSEARCH=(WORK LIBRARY) FONTSLOC=(system-specific pathname)
FORMCHAR=$<>\^_{|}~+=|-/\<>* FORMDLIM= FORMS=DEFAULT GISMAPS= GWINDOW HELPENCMD
HELPINDEX=(/help/common.hlp/index.txt /help/common.hlp/keywords.htm common.hhk)
HELPTOC=(/help/helpnav.hlp/config.txt /help/common.hlp/toc.htm common.hhc)
IBUFSIZE=0 NOIMPLMAC INITCMD= INITSTMT= INVALIDDATA=. LABEL LEFTMARGIN=0.000 IN
LINESIZE=97 LOGPARM= MACRO MAPS=(system-specific pathname) NOMAUTOLOCDISPLAY
MAUTOSOURCE MAXSEGRATIO=75 MCOMPILENOTE=NONE MERGENOBY=NOWARN MERROR
METAAUTORESOURCES= METACONNECT= METAENCRYPTALG=NONE METAENCRYPTLEVEL=EVERYTHING
METAID= METAPASS= METAPORT=0 METAPROFILE= METAPROTOCOL=BRIDGE METAREPOSITORY=Default
METASERVER= METAUSER= NOMFILE MINDELIMITER= MINPARTSIZE=0 MISSING=. NOMLOGIC
NOMLOGICNEST NOMPRINT NOMPRINTNEST NOMRECALL MSGLEVEL=N NOMSTORED MSYMTABMAX=4194304
NOMULTENVAPPL MVARSIZE=4096 NONETENCRYPT NETENCRYPTALGORITHM= NETENCRYPTKEYLEN=0
NETMAC NEWS= NOTES NUMBER NOOBJECTSERVER OBS=9223372036854775807 ORIENTATION=PORTRAIT
NOOVP NOPAGEBREAKINITIAL PAGENO=1 PAGESIZE=55 PAPERDEST= PAPERSIZE=LETTER
PAPERSOURCE= PAPERTYPE=PLAIN PARM= PARMCARDS=FT15F001 PRINTERPATH= NOPRINTINIT
PRINTMSGLIST QUOTELENMAX REPLACE REUSE=NO RIGHTMARGIN=0.000 IN NORSASUSER S=0
S2=0 SASAUTOS=(system-specific pathname) SASCMD= SASFRSCR=
SASHELP=(system-specific pathname) SASMSTORE= SASSCRIPT=
SASUSER=(system-specific pathname) SEQ=8 SERROR NOSETINIT SIGNONWAIT SKIP=0
SOLUTIONS SORTDUP=PHYSICAL SORTEQUALS SORTSEQ= SORTSIZE=2097152 SOURCE NOSOURCE2
SPDEINDEXSORTSIZE=33554432 SPDEMAXTHREADS=0 SPDESORTSIZE=33554432 SPDEUTILLOC=
SPDEWHEVAL=COST NOSPOOL NOSSLCLIENTAUTH NOSSLCRLCHECK STARTLIB SUMSIZE=0
NOSYMBOLGEN SYNCHIO SYNTAXCHECK SYSPARM= SYSPRINTFONT= NOSYSRPUTSYNC TBUFSIZE=0
TCPPORTFIRST=0 TCPPORTLAST=0 TERMINAL TERMSTMT= TEXTURELOC=\\dntsrc\sas\m901\ods\misc
THREADS TOOLSMENU TOPMARGIN=0.000 IN TRAINLOC= TRANTAB= UNIVERSALPRINT USER= UTILLOC=
UUIDCOUNT=100 UUIDGENDHOST= V6CREATEUPDATE=NOTE VALIDFMTNAME=LONG VALIDVARNAME=V7
VIEWMENU VNFERR WORK=(system-specific pathname) WORKINIT WORKTERM YEARCUTOFF=1920
_LAST_=_NULL_

Example 2: Displaying the Setting of a Single Option

Procedure features:

PROC OPTIONS statement option:

OPTION=
DEFINE
VALUE

The OPTIONS Procedure

Program

599

This example shows how to display the setting of a single SAS system option. The
log shows the current setting of the SAS system option CENTER. The DEFINE and
VALUE options display additional information.

Program

Set the CENTER SAS system option.OPTION=CENTER displays option value information.
DEFINE and VALUE display additional information.

proc options option=center define value;

run;

Output 29.4
29
30

Log Output from Specifying the CENTER Option

proc options option=center define value;

run;
SAS (r) Proprietary Software Release XXX

Option Value Information For SAS Option CENTER

Option Value: CENTER
Option Scope: Default
How option value set: Shipped Default
Option Definition Information for SAS Option CENTER
Group= LISTCONTROL
Group Description: Procedure output and display settings
Description: Center SAS procedure output
Type: The option value is of type BOOLEAN
When Can Set: Startup or anytime during the SAS Session
Restricted: Your Site Administrator can restrict modification of this
option.
Optsave: Proc Optsave or command Dmoptsave will save this option.

600

601

CHAPTER

30
The OPTLOAD Procedure
Overview: OPTLOAD Procedure 601
What Does the OPTLOAD Procedure Do?
Syntax: OPTLOAD Procedure 601
PROC OPTLOAD Statement 602

601

Overview: OPTLOAD Procedure

What Does the OPTLOAD Procedure Do?
The OPTLOAD procedure reads SAS system option settings that are stored in the
SAS registry or a SAS data set and puts them into effect.
You can load SAS system option settings from a SAS data set or registry key by using

3 the DMOPTLOAD command from a command line in the SAS windowing

environment. For example, DMOPTLOAD key= core\options.

3 the PROC OPTLOAD statement.

When an option is restricted by the site administrator, and the option value that is
being set by PROC OPTLOAD differs from the option value that was established by the
site administrator, SAS issues a Warning message to the log.
Some SAS options will not be saved with PROC OPTSAVE and therefore cannot be
loaded with OPTLOAD. The following is a list of these options:

3
3
3
3
3
3
3
3
3
3

ARMAGENT system option

ARMLOC system option
ARMSUBSYS system option
AWSDEF system option (for Windows only)
FONTALIAS system option (for Windows only)
SORTMSG system option (for z/OS only)
STIMER system option
TCPSEC system option
all SAS system options that can be specied only during startup
all SAS system options that identify a password.

Syntax: OPTLOAD Procedure

PROC OPTLOAD <options>;

602

PROC OPTLOAD Statement

Chapter 30

PROC OPTLOAD Statement

PROC OPTLOAD <options>;

To do this

Use this option

Load SAS system option settings from an existing registry key

KEY=

Load SAS system option settings from an existing data set

DATA=

Options
DATA=libref.dataset

species the library and data set name from where SAS system option settings are
loaded. The SAS variable OPTNAME contains the character value of the SAS system
option name, and the SAS variable OPTVALUE contains the character value of the
SAS system option setting.
Requirement:

The SAS library and data set must exist.

Default: If you omit the DATA= option and the KEY= option, the procedure will use

the default SAS library and data set. The default library is where the current user
prole resides. Unless you specify a library, the default library is SASUSER. If
SASUSER is being used by another active SAS session, then the temporary
WORK library is the default location from which the data set is loaded. The
default data set name is MYOPTS.
KEY=SAS registry key

species the location in the SAS registry of stored SAS system option settings. The
registry is retained in SASUSER. If SASUSER is not available, then the temporary
WORK library is used. For example, KEY="OPTIONS".
Requirement:

SAS registry key must be an existing SAS registry key.

You must use quotation marks around the SAS registry key name.
Separate the names in a sequence of key names with a backslash (\). For
example, KEY=CORE\OPTIONS.

Requirement:

603

CHAPTER

31
The OPTSAVE Procedure
Overview: OPTSAVE Procedure 603
What Does the OPTSAVE Procedure Do?
Syntax: OPTSAVE Procedure
603
PROC OPTSAVE Statement 604

603

Overview: OPTSAVE Procedure

What Does the OPTSAVE Procedure Do?
PROC OPTSAVE saves the current SAS system option settings in the SAS registry
or in a SAS data set.
SAS system options can be saved across SAS sessions. You can save the settings of
the SAS system options in a SAS data set or registry key by using

3 the DMOPTSAVE command from a command line in the SAS windowing

environment. Use the command like this: DMOPTSAVE <save-location>.

3 the PROC OPTSAVE statement.

Some SAS options will not be saved with PROC OPTSAVE. The following is a list of
these options:

3
3
3
3
3
3
3
3
3
3

ARMAGENT system option

ARMLOC system option
ARMSUBSYS system option
AWSDEF system option
FONTALIAS system option
SORTMSG system option
STIMER system option
TPSEC system option
All SAS system options that can be specied only during startup
All SAS system options that identify a password.

Syntax: OPTSAVE Procedure

Tip: The only statement that is used with the OPTSAVE procedure is the PROC
statement.

604

PROC OPTSAVE Statement

Chapter 31

PROC OPTSAVE <options>;

PROC OPTSAVE Statement

PROC OPTSAVE <options >;

To do this

Use this option

Save SAS system option settings to a registry key

KEY=

Save SAS system option settings to a SAS data set

OUT=

Options
KEY=SAS registry key

Separate the names in a sequence of key names with a backslash

(\). Individual key names can contain any character except a backslash.

Requirement:

The length of a key name cannot exceed 255 characters (including

the backslashes).

Requirement:
Requirement:
Tip:

You must use quotation marks around the SAS registry key name.

To specify a subkey, enter multiple key names starting with the root key.

If the key already exists, it will be overwritten. If the specied key does
not already exist in the current SAS registry, then the key is automatically created
when option settings are saved in the SAS registry.

Caution:

OUT=libref.dataset

species the names of the library and data set where SAS system option settings are
saved. The SAS variable OPTNAME contains the character value of the SAS system
option name. The SAS variable OPTVALUE contains the character value of the SAS
system option setting.
Caution:

If the data set already exists, it will be overwritten.

Default: If you omit the OUT= and the KEY= options, the procedure will use the

default SAS library and data set. The default SAS library is where the current
user prole resides. Unless you specify a SAS library, the default library is
SASUSER. If SASUSER is in use by another active SAS session, then the
temporary WORK library is the default location where the data set is saved. The
default data set name is MYOPTS.

605

CHAPTER

607

Plotting Two Sets of Values at Once

Plot of Highs and Lows
for the Dow Jones Industrial Average

Plot of HighYear. Symbol used is .

Plot of Low*Year. Symbol used is o.
---+---------+---------+---------+---------+---------+--4000 +
*
+
|
*
|
|
* o
|
|
*oo
|
High |
*
|
|
* *
|
|
o
|
|
*oo
|
2000 +
* o
+
|
o
|
|
*o
|
|
**o
|
|
****** ************oo
|
|
*****oooooo*o o oooooooo
|
|
*****oooo
o
|
|
o
|
0 +
+
---+---------+---------+---------+---------+---------+--1950
1960
1970
1980
1990
2000
Year
NOTE: 7 obs hidden.

PROC PLOT can also label points on a plot with the values of a variable, as shown in
Output 32.3. The plotted data represents population density and crime rates for
selected U.S. states. The SAS code that produces Output 32.3 is shown in Example 11
on page 654.

608

Syntax: PLOT Procedure

Output 32.3

Chapter 32

Labeling Points on a Plot

A Plot of Population Density and Crime Rates
Plot of Density*CrimeRate$State.

Symbol is value of State.

---+------------+------------+------------+------------+------------+------------+------------+--Density |
|
500 +
|

+
|

|
|
|

Maryland
M

|
|
|

Pennsylvania

|
250 +
|

|
|
|

|
+
|

Illinois

|
|
|
|
|

Delaware
D
Ohio

New
Hampshire

West

| Virginia
|
W
|
|
|
0 +

Florida
F

North Carolina
South
Alabama N
Carolina

N
T
S
Mississippi
A Tennessee
M Vermont V
M Missouri

California
C
G Georgia
Oklahoma

South
Arkansas A
M Minnesota
Dakota
I Idaho
S N North Dakota

O
Nevada
N

Washington Texas
W
T
Oregon
O

|
|
|
|
|
|
|
|
|
|
+

---+------------+------------+------------+------------+------------+------------+------------+--2000
3000
4000
5000
6000
7000
8000
9000
CrimeRate

Syntax: PLOT Procedure

At least one PLOT statement is required.
Supports RUN-group processing
Tip: Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Names: See: ODS Table Names on page 629
Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See
Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.
Requirement:
Tip:

PROC PLOT < option(s)>;

BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
PLOT plot-request(s) </ option(s)>;

The PLOT Procedure

To do this

Describe the plots you want

609

Use this statement

Produce a separate plot for each BY group

PROC PLOT Statement

PLOT

PROC PLOT Statement

Reminder: You can use data set options with the DATA= option. See Data Set Options

on page 18 for a list.

PROC PLOT <option(s)>;

To do this

Use this option

Specify the input data set

DATA=

Control the axes

Include missing character variable values

MISSING

Exclude observations with missing values

NOMISS

Uniformly scale axes across BY groups

UNIFORM

Control the appearance of the plot

Specify the characters that construct the borders
of the plot

FORMCHAR=

Suppress the legend at the top of the plot

NOLEGEND

Specify the aspect ratio of the characters on the

output device

VTOH=

Control the size of the plot

Specify the percentage of the available
horizontal space for each plot

HPERCENT=

Specify the percentage of the available vertical

space for each plot

VPERCENT=

Options
DATA=SAS-data-set

species the input SAS data set.

Main discussion:

Procedures."

See Chapter 2, "Fundamental Concepts for Using Base SAS

610

PROC PLOT Statement

Chapter 32

FORMCHAR <(position(s))>=formatting-character(s)

denes the characters to use for constructing the borders of the plot.
position(s)
identies the position of one or more characters in the SAS formatting-character
string. A space or a comma separates the positions.
Default: Omitting (position(s)) is the same as specifying all twenty possible SAS

formatting characters, in order.

Range: PROC PLOT uses formatting characters 1, 2, 3, 5, 7, 9, and 11. The

following table shows the formatting characters that PROC PLOT uses.

Position

Default

Used to draw

vertical separators

horizontal separators

35911

corners

intersection of vertical
and horizontal separators

formatting-character(s)
lists the characters to use for the specied positions. PROC PLOT assigns
characters in formatting-character(s) to position(s), in the order that they are
listed. For instance, the following option assigns the asterisk (*) to the third
formatting character, the pound sign (#) to the seventh character, and does not
alter the remaining characters:
formchar(3,7)=*#

Interaction: The SAS system option FORMCHAR= species the default formatting

characters. The system option denes the entire string of formatting characters.
The FORMCHAR= option in a procedure can redene selected characters.
You can use any character in formatting-characters, including hexadecimal
characters. If you use hexadecimal characters, then you must put an x after the
closing quotation mark. For instance, the following option assigns the hexadecimal
character 2D to the third formatting character, the hexadecimal character 7C to
the seventh character, and does not alter the remaining characters:

Tip:

formchar(3,7)=2D7Cx

Specifying all blanks for formatting-character(s) produces plots with no

borders, for example

Tip:

formchar (1,2,7)=

HPERCENT=percent(s)

species one or more percentages of the available horizontal space to use for each
plot. HPERCENT= enables you to put multiple plots on one page. PROC PLOT tries
to t as many plots as possible on a page. After using each of the percent(s), PROC
PLOT cycles back to the beginning of the list. A zero in the list forces PROC PLOT to
go to a new page even if it could t the next plot on the same page.
hpercent=33

prints three plots per page horizontally; each plot is one-third of a page wide.

The PLOT Procedure

PROC PLOT Statement

611

hpercent=50 25 25

prints three plots per page; the rst is twice as wide as the other two.
hpercent=33 0

produces plots that are one-third of a page wide,; each plot is on a separate page.
hpercent=300

produces plots three pages wide.

At the beginning of every BY group and after each RUN statement, PROC PLOT
returns to the beginning of the percent(s) and starts printing a new page.
Alias:

HPCT=

Default: 100
Featured in:

Example 4 on page 636

MISSING

includes missing character variable values in the construction of the axes. It has no
effect on numeric variables.
Interaction: overrides the NOMISS option for character variables
NOLEGEND

suppresses the legend at the top of each plot. The legend lists the names of the
variables being plotted and the plotting symbols used in the plot.
NOMISS

excludes observations for which either variable is missing from the calculation of the
axes. Normally, PROC PLOT draws an axis based on all the values of the variable
being plotted, including points for which the other variable is missing.
Interaction: The HAXIS= option overrides the effect of NOMISS on the horizontal

axis. The VAXIS= option overrides the effect on the vertical axis.
Interaction: NOMISS is overridden by MISSING for character variables.
Featured in:

Example 10 on page 652

UNIFORM

uniformly scales axes across BY groups. Uniform scaling enables you to directly
compare the plots for different values of the BY variables.
Restriction: You cannot use PROC PLOT with the UNIFORM option with an

engine that supports concurrent access if another user is updating the data set at
the same time.
VPERCENT=percent(s)

species one or more percentages of the available vertical space to use for each plot.
If you use a percentage greater than 100, then PROC PLOT prints sections of the
plot on successive pages.
Alias:

VPCT=

Default: 100
Featured in:

Example 4 on page 636

options in the PLOT statement.

See also: HAXIS= on page 616 for a way to equate axes so that the given distance

represents the same data range on both axes.

BY Statement
Produces a separate plot and starts a new page for each BY group.
Main discussion: BY on page 58
Featured in:

Example 8 on page 646

BY < DESCENDING> variable-1

<< DESCENDING> variable-n>
<NOTSORTED>;

Required Arguments
variable

species the variable that the procedure uses to form BY groups. You can specify
more than one variable. If you do not use the NOTSORTED option in the BY
statement, then the observations in the data set must either be sorted by all the
variables that you specify or be indexed appropriately. Variables in a BY statement
are called BY variables.

Options
DESCENDING

species that the observations are sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data is grouped in another way, for example, chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations with the same values for the BY variables
are not contiguous, then the procedure treats each contiguous set as a separate BY
group.

The PLOT Procedure

PLOT Statement

PLOT Statement
Requests the plots to be produced by PROC PLOT.
Tip:

You can use multiple PLOT statements.

PLOT plot-request(s) </ option(s)>;

To do this

Use this option

Control the axes

Specify the tick-mark values

HAXIS= and VAXIS=

Expand the axis

HEXPAND and VEXPAND

Specify the number of print positions

HPOS=
and
VPOS=

Reverse the order of the values

HREVERSE and VREVERSE

Specify the number of print positions between

tick marks

HSPACE= and VSPACE=

Assign a value of zero to the rst tick mark

HZERO
and
VZERO

Specify reference lines

Draw a line perpendicular to the specied
values on the axis

HREF=
and
VREF=

Specify a character to use to draw the reference

line

HREFCHAR= and VREFCHAR=

Put a box around the plot

BOX

Overlay plots

OVERLAY

Produce a contour plot

Draw a contour plot

CONTOUR

Specify the plotting symbol for one contour level

Scontour-level=

Specify the plotting symbol for multiple contour

levels

SLIST=

Label points on a plot

List the penalty and the placement state of the
points

LIST=

Force the labels away from the origin

OUTWARD=

Change default penalties

PENALTIES=

Specify locations for the placement of the labels

PLACEMENT=

613

614

PLOT Statement

Chapter 32

To do this

Use this option

Specify a split character for the label

SPLIT=

List all placement states in effect

STATES

Required Arguments
plot-request(s)

species the variables (vertical and horizontal) to plot and the plotting symbol to use
to mark the points on the plot.
Each form of plot-request(s) supports a label variable. A label variable is preceded
by a dollar sign ($) and species a variable whose values label the points on the plot.
For example,
plot y*x $ label-variable
plot y*x=* $ label-variable

See Labeling Plot Points with Values of a Variable on page 625 for more
information. In addition, see Example 9 on page 649 and all the examples that follow
it.
The plot-request(s) can be one or more of the following:
vertical*horizontal <$ label-variable>
species the variable to plot on the vertical axis and the variable to plot on the
horizontal axis.
For example, the following statement requests a plot of Y by X:
plot y*x;

Y appears on the vertical axis, X on the horizontal axis.

This form of the plot request uses the default method of choosing a plotting
symbol to mark plot points. When a point on the plot represents the values of one
observation in the data set, PROC PLOT puts the character A at that point. When
a point represents the values of two observations, the character B appears. When
a point represents values of three observations, the character C appears, and so on
through the alphabet. The character Z is used for the occurrence of 26 or more
observations at the same printing position.
vertical*horizontal=character <$ label-variable>
species the variables to plot on the vertical and horizontal axes and species a
plotting symbol to mark each point on the plot. A single character is used to
represent values from one or more observations.
For example, the following statement requests a plot of Y by X, with each point
on the plot represented by a plus sign (+):
plot y*x=+;

vertical*horizontal=variable <$ label-variable>

species the variables to plot on the vertical and horizontal axes and species a
variable whose values are to mark each point on the plot. The variable can be
either numeric or character. The rst (left-most) nonblank character in the
formatted value of the variable is used as the plotting symbol (even if more than
one value starts with the same letter). When more than one observation maps to
the same plotting position, the value from the rst observation marks the point.

The PLOT Procedure

PLOT Statement

615

For example, in the following statement GENDER is a character variable with

values of FEMALE and MALE; the values F and M mark each observation on the plot.
plot height*weight=gender;

Specifying Variable Lists in Plot Requests

You can use SAS variable lists in plot requests. For example, the following are valid
plot requests:
Plot request

What is plotted

(a - - d)

ab ac ad bc b*d

c*d

(x1 - x4)

x1*x2
x1*x3 x1*x4 x2*x3
x2*x4 x3*x4

(_numeric_)

All combinations of numeric variables

y*(x1 - x4)

y*x1
y*x2 y*x4 y*x4

If both the vertical and horizontal specications request more than one variable and
if a variable appears in both lists, then it will not be plotted against itself. For example,
the following statement does not plot B*B and C*C:
plot (a b c)*(b c d);

Specifying Combinations of Variables

The operator in request is either an asterisk (*) or a colon (:). An asterisk combines
the variables in the lists to produce all possible combinations of x and y variables. For
example, the following plot requests are equivalent:
plot (y1-y2) * (x1-x2);
plot y1*x1 y1*x2 y2*x1 y2*x2;

A colon combines the variables pairwise. Thus, the rst variables of each list
combine to request a plot, as do the second, third, and so on. For example, the following
plot requests are equivalent:
plot (y1-y2) : (x1-x2);
plot y1*x1 y2*x2;

Options
BOX

draws a border around the entire plot, rather than just on the left side and bottom.
Featured in: Example 3 on page 634
CONTOUR<=number-of-levels>

draws a contour plot using plotting symbols with varying degrees of shading where
number-of-levels is the number of levels for dividing the range of variable. The plot

616

PLOT Statement

Chapter 32

request must be of the form vertical*horizontal=variable where variable is a numeric

variable in the data set. The intensity of shading is determined by the values of this
variable.
When you use CONTOUR, PROC PLOT does not plot observations with missing
values for variable.
Overprinting, if it is enabled by the OVP system option, is used to produce the
shading. Otherwise, single characters varying in darkness are used. The CONTOUR
option is most effective when the plot is dense.
Default: 10
Range: 1-10
Featured in: Example 7 on page 642
HAXIS=axis-specication

species the tick-mark values for the horizontal axis.

3 For numeric values, axis-specication is either an explicit list of values, a BY
increment, or a combination of both:
n <n>
BY increment
n TO n BY increment
The values must be in either ascending or descending order. Use a negative
value for increment to specify descending order. The specied values are spaced
evenly along the horizontal axis even if the values are not uniformly
distributed. Numeric values can be specied in the following ways:
HAXIS= value

Comments

10 to 100 by 5

Values appear in increments of 5, starting

at 10 and ending at 100.

by 5

Values are incremented by 5. PROC PLOT

determines the minimum and maximum
values for the tick marks.

10 100 1000 10000

Values are not uniformly distributed. This

specication produces a logarithmic plot. If
PROC PLOT cannot determine the function
implied by the axis specication, it uses
simple linear interpolation between the
points. To determine whether PROC PLOT
correctly interpolates a function, you can
use the DATA step to generate data that
determines the function and see whether it
appears linear when plotted. See Example 5
on page 639 for an example.

1 2 10 to 100
by 5

A combination of the previous specications.

3 For character variables, axis-specication is a list of unique values that are

enclosed in quotation marks:
value-1 <value-n>

The PLOT Procedure

PLOT Statement

617

For example,
haxis=Paris London Tokyo

The character strings are case-sensitive. If a character variable has an

associated format, then axis-specication must specify the formatted value. The
values can appear in any order.
3 For axis variables that contain date-time values, axis-specication is either an
explicit list of values or a starting and an ending value with an increment
specied:
date-time-valuei < date-time-valuei>

date-time-valuei TO <date-time-valuei>
<BY increment>
date-time-valuei
any SAS date, time, or datetime value described for the SAS functions
INTCK and INTNX. The sufx i is one of the following:
D

date

time

datetime

increment
one of the valid arguments for the INTCK or INTNX functions: For dates,
increment can be one of the following:
DAY
WEEK
MONTH
QTR
YEAR
For datetimes, increment can be one of the following:
DTDAY
DTWEEK
DTMONTH
DTQTR
DTYEAR
For times, increment can be one of the following:
HOUR
MINUTE
SECOND
For example,
haxis=01JAN95d to 01JAN96d
by month
haxis=01JAN95d to 01JAN96d
by qtr

Note: You must use a FORMAT statement to print the tick-mark values
in an understandable form. 4
Interaction: You can use the HAXIS= and VAXIS= options with the VTOH= option

to equate axes. If your data is suitable, then use HAXIS=BY n and VAXIS=BY n

618

PLOT Statement

Chapter 32

with the same value for n and specify a value for the VTOH= option. The number
of columns that separate the horizontal tick marks is nearly equal to the number
of lines that separate the vertical tick marks times the value of the VTOH= option.
In some cases, PROC PLOT cannot simultaneously use all three values and
changes one or more of the values.
Featured in: Example 2 on page 632, Example 5 on page 639, and Example 6 on
page 640
HEXPAND

expands the horizontal axis to minimize the margins at the sides of the plot and to
maximize the distance between tick marks, if possible.
HEXPAND causes PROC PLOT to ignore information about the spacing of the
data. Plots produced with this option waste less space but may obscure the nature of
the relationship between the variables.
HPOS=axis-length

species the number of print positions on the horizontal axis. The maximum value of
axis-length that allows a plot to t on one page is three positions less than the value
of the LINESIZE= system option because there must be space for the procedure to
print information next to the vertical axis. The exact maximum depends on the
number of characters that are in the vertical variables values. If axis-length is too
large to t on a line, then PROC PLOT ignores the option.
HREF=value-specication

draws lines on the plot perpendicular to the specied values on the horizontal axis.
PROC PLOT includes the values you specify with the HREF= option on the
horizontal axis unless you specify otherwise with the HAXIS= option.
For the syntax for value-specication, see HAXIS= on page 616.
Featured in: Example 8 on page 646
HREFCHAR=character

species the character to use to draw the horizontal reference line.

Default: vertical bar (|)
See also: FORMCHAR= option on page 610 and HREF= on page 618
HREVERSE

reverses the order of the values on the horizontal axis.

HSPACE=n

species that a tick mark will occur on the horizontal axis at every nth print
position, where n is the value of HSPACE=.
HZERO

assigns a value of zero to the rst tick mark on the horizontal axis.
Interaction: PROC PLOT ignores HZERO if the horizontal variable has negative
values or if the HAXIS= option species a range that does not begin with zero.
LIST<=penalty-value>

lists the horizontal and vertical axis values, the penalty, and the placement state of
all points plotted with a penalty greater than or equal to penalty-value. If no plotted
points have a penalty greater than or equal to penalty-value, then no list is printed.
Tip: LIST is equivalent to LIST=0.
See also: Understanding Penalties on page 626
Featured in: Example 11 on page 654
OUTWARD=character

tries to force the point labels outward, away from the origin of the plot, by protecting
positions next to symbols that match character that are in the direction of the origin

The PLOT Procedure

PLOT Statement

619

(0,0). The algorithm tries to avoid putting the labels in the protected positions, so
they usually move outward.
This option is useful only when you are labeling points with the values of a
variable.

Tip:

OVERLAY

overlays all plots that are specied in the PLOT statement on one set of axes. The
variable names, or variable labels if they exist, from the rst plot are used to label
the axes. Unless you use the HAXIS= or the VAXIS= option, PROC PLOT
automatically scales the axes in the way that best ts all the variables.
When the SAS system option OVP is in effect and overprinting is allowed, the
plots are superimposed; otherwise, when NOOVP is in effect, PROC PLOT uses the
plotting symbol from the rst plot to represent points that appear in more than one
plot. In such a case, the output includes a message telling you how many
observations are hidden.
Featured in:

Example 3 on page 634

PENALTIES<(index-list)>=penalty-list

changes the default penalties. The index-list provides the positions of the penalties in
the list of penalties. The penalty-list contains the values that you are specifying for
the penalties that are indicated in the index-list. The index-list and the penalty-list
can contain one or more integers. In addition, both index-list and penalty-list accept
the form:
value TO value

Example 13 on page 661

PLACEMENT=(expression(s))

controls the placement of labels by specifying possible locations of the labels relative
to their coordinates. Each expression consists of a list of one or more suboptions (H=,
L=, S=, or V=) that are joined by an asterisk (*) or a colon (:). PROC PLOT uses the
asterisk and colon to expand each expression into combinations of values for the four
possible suboptions. The asterisk creates every possible combination of values in the
expression list. A colon creates only pairwise combinations. The colon takes
precedence over the asterisk. With the colon, if one list is shorter than the other,
then the values in the shorter list are reused as necessary.
Use the following suboptions to control the placement:
H=integer(s)
species the number of horizontal spaces (columns) to shift the label relative to
the starting position. Both positive and negative integers are valid. Positive
integers shift the label to the right; negative integers shift it to the left. For
example, you can use the H= suboption in the following way:
place=(h=0 1 -1 2 -2)

You can use the keywords BY ALT in this list. BY ALT produces a series of
numbers whose signs alternate between positive and negative and whose absolute
values change by one after each pair. For instance, the following PLACE=
specications are equivalent:
place=(h=0 -1 to -3 by alt)
place=(h=0 -1 1 -2 2 -3 3)

620

PLOT Statement

Chapter 32

If the series includes zero, then the zero appears twice. For example, the
following PLACE= options are equivalent:
place=(h= 0 to 2 by alt)
place=(h=0 0 1 -1 2 -2)

Default: H=0
Range: 500 to 500

L=integer(s)
species the number of lines onto which the label may be split.
Default: L=1
Range: 1-200

S=start-position(s)
species where to start printing the label. The value for start-position can be one
or more of the following:
CENTER
the procedure centers the label around the plotting symbol.
RIGHT
the label starts at the plotting symbol location and continues to the right.
LEFT
the label starts to the left of the plotting symbol and ends at the plotting symbol
location.
Default: CENTER

V=integer(s)
species the number of vertical spaces (lines) to shift the label relative to the
starting position. V= behaves the same as the H= suboption, described earlier.
A new expression begins when a suboption is not preceded by an operator.
Parentheses around each expression are optional. They make it easier to recognize
individual expressions in the list. However, the entire expression list must be in
parentheses, as shown in the following example. Table 32.1 on page 621 shows how
this expression is expanded and describes each placement state.
place=((v=1)
(s=right left : h=2 -2)
(v=-1)
(h=0 1 to 2 by alt * v=1 -1)
(l=1 to 3 * v=1 to 2 by alt *
h=0 1 to 2 by alt))

Each combination of values is a placement state. The procedure uses the

placement states in the order in which they appear in the placement states list, so
specify your most preferred placements rst. For each label, the procedure tries all
states, then uses the rst state that places the label with minimum penalty. When
all labels are initially placed, the procedure cycles through the plot multiple times,
systematically rening the placements. The renement step tries to both minimize
the penalties and to use placements nearer to the beginning of the states list.
However, PROC PLOT uses a heuristic approach for placements, so the procedure
does not always nd the best set of placements.
Alias:

PLACE=

There are two defaults for the PLACE= option. If you are using a blank
as the plotting symbol, then the default placement state is PLACE=(S=CENTER :

Defaults:

The PLOT Procedure

PLOT Statement

621

V=0 : H=0 : L=1), which centers the label. If you are using anything other than a
blank, then the default is PLACE=((S=RIGHT LEFT : H=2 2) (V=1 1 * H=0 1 -1
2 -2)). The default for labels placed with symbols includes multiple positions
around the plotting symbol so the procedure has exibility when placing labels on
a crowded plot.
Tip: Use the STATES option to print a list of placement states.
See also: Labeling Plot Points with Values of a Variable on page 625
Featured in:
Table 32.1

Example 11 on page 654 and Example 12 on page 658

Expanding an Expression List into Placement States

Expression

Placement state

Meaning

(V=1)

S=CENTER L=1 H=0 V=1

Center the label, relative to the

point, on the line above the
point. Use one line for the label.

(S=RIGHT LEFT : H=2 2)

S=RIGHT L=1 H=2 V=0

Begin the label in the second

column to the right of the point.
Use one line for the label.

S=LEFT L=1 H=2 V=0

End the label in the second

column to the left of the point.
Use one line for the label.

(V=1)

S=CENTER L=1 H=0 V= 1

Center the label, relative to the

point, on the line below the
point. Use one line for the label.

(H=0 1 to 2 BY ALT * V=1 1)

S=CENTER L=1 H=2 V=1

From center, shift the labels two

columns to the right, rst on
the line above the point, then
on the line below.

622

PLOT Statement

Chapter 32

Expression

Placement state

Meaning

S=CENTER L=1 H=2 V=1

From center, shift the labels two

columns to the left, rst on the
line above the point, then on
the line below.

S=CENTER L=1 H=2 V=1

(L=1 to 3 * V=1 to 2 BY ALT * H=0 1 to
2 BY ALT)

S=CENTER L=1 H=0 V=1

Center the label, relative to the

point, on the line above the
point. Use one line for the label.

S=CENTER
S=CENTER
S=CENTER
S=CENTER

From center, shift the label one

or two columns to the right or
left on the line above the point.
Use one line for the label.

L=1
L=1
L=1
L=1

H=1 V=1
H=1 V=1
H=2 V=1
H=2 V=1

S=CENTER L=1 H=0 V=1

Center the label, relative to the

point, on the line below the
point. Use one line for the label.

S=CENTER
S=CENTER
S=CENTER
S=CENTER

From center, shift the label one

or two columns to the right and
the left on the line below the
point.

L=1
L=1
L=1
L=1

H=1 V=1
H=1 V=1
H=2 V=1
H=2 V=1

.
.
.

Use the same horizontal shifts

on the line two lines above the
point and on the line two lines
below the point.

S=CENTER L=1 H= 2 V=2

S=CENTER L=2 H=0 V=1

Repeat the whole process

splitting the label over two
lines. Then repeat it splitting
the label over three lines.

.
.
.
S=CENTER L=3 H= 2 V=2

Scontour-level=character-list

species the plotting symbol to use for a single contour level. When PROC PLOT
produces contour plots, it automatically chooses the symbols to use for each level of
intensity. You can use the S= option to override these symbols and specify your own.
You can include up to three characters in character-list. If overprinting is not
allowed, then PROC PLOT uses only the rst character.
For example, to specify three levels of shading for the Z variable, use the following
statement:
plot y*x=z /
contour=3 s1=A s2=+ s3=X0A;

The PLOT Procedure

PLOT Statement

623

You can also specify the plotting symbols as hexadecimal constants:

plot y*x=z /
contour=3 s1=7Ax

s2=7Fx s3=A6x;

This feature was designed especially for printers where the hexadecimal constants
can represent grey-scale ll characters.
Range: 1 to the highest contour level (determined by the CONTOUR option).
See also: SLIST= and CONTOUR
SLIST=character-list-1 <character-list-n>

species plotting symbols for multiple contour levels. Each character-list species the
plotting symbol for one contour level: the rst character-list for the rst level, the
second character-list for the second level, and so on. For example:
plot y*x=z /
contour=5

slist=. : ! = +O;

Default: If you omit a plotting symbol for each contour level, then PROC PLOT

uses the default symbols:

slist=. , - = + O X
W * #

Restriction: If you use the SLIST= option, then it must be listed last in the PLOT

statement.
See also: Scontour-level= and CONTOUR=
SPLIT=split-character

when labeling plot points, species where to split the label when the label spans two
or more lines. The label is split onto the number of lines that is specied in the L=
suboption to the PLACEMENT= option. If you specify a split character, then the
procedure always splits the label on each occurrence of that character, even if it
cannot nd a suitable placement. If you specify L=2 or more but do not specify a split
character, then the procedure tries to split the label on blanks or punctuation but
will split words if necessary.
PROC PLOT shifts split labels as a block, not as individual fragments (a fragment
is the part of the split label that is contained on one line). For example, to force This
is a label to split after the a , change it to This is a*label and specify
SPLIT=* .
See also: Labeling Plot Points with Values of a Variable on page 625
STATES

lists all the placement states in effect. STATES prints the placement states in the
order that you specify them in the PLACE= option.
VAXIS=axis-specication

species tick mark values for the vertical axis. VAXIS= follows the same rules as
theHAXIS= option on page 616.
Featured in:

Example 7 on page 642 and Example 12 on page 658

VEXPAND

expands the vertical axis to minimize the margins above and below the plot and to
maximize the space between vertical tick marks, if possible.
See also: HEXPAND on page 618
VPOS=axis-length

species the number of print positions on the vertical axis. The maximum value for
axis-length that allows a plot to t on one page is 8 lines less than the value of the

624

Concepts: PLOT Procedure

Chapter 32

SAS system option PAGESIZE= because you must allow room for the procedure to
print information under the horizontal axis. The exact maximum depends on the
titles that are used, whether or not plots are overlaid, and whether or not CONTOUR
is specied. If the value of axis-length species a plot that cannot t on one page,
then the plot spans multiple pages.
See also: HPOS= on page 618
VREF=value-specication

draws lines on the plot perpendicular to the specied values on the vertical axis.
PROC PLOT includes the values you specify with the VREF= option on the vertical
axis unless you specify otherwise with the VAXIS= option. For the syntax for
value-specication, see HAXIS= on page 616.
Featured in:

Example 2 on page 632

VREFCHAR=character

species the character to use to draw the vertical reference lines.

Default: horizontal bar (-)
See also: FORMCHAR= option on page 610, HREFCHAR= on page 618, and

VREF= on page 624

VREVERSE

reverses the order of the values on the vertical axis.

VSPACE=n

species that a tick mark will occur on the vertical axis at every nth print position,
where n is the value of VSPACE=.
VZERO

assigns a value of zero to the rst tick mark on the vertical axis.
Interaction: PROC PLOT ignores the VZERO option if the vertical variable has

negative values or if the VAXIS= option species a range that does not begin with
zero.

Concepts: PLOT Procedure

RUN Groups
PROC PLOT is an interactive procedure. It remains active after a RUN statement is
executed. Usually, SAS terminates a procedure after executing a RUN statement.
When you start the PLOT procedure, you can continue to submit any valid statements
without resubmitting the PROC PLOT statement. Thus, you can easily experiment with
changing labels, values of tick marks, and so forth. Any options submitted in the PROC
PLOT statement remain in effect until you submit another PROC PLOT statement.
When you submit a RUN statement, PROC PLOT executes all the statements
submitted since the last PROC PLOT or RUN statement. Each group of statements is
called a RUN group. With each RUN group, PROC PLOT begins a new page and begins
with the rst item in the VPERCENT= and HPERCENT= lists, if any.
To terminate the procedure, submit a QUIT statement, a DATA statement, or a
PROC statement. Like the RUN statement, each of these statements completes a RUN
group. If you do not want to execute the statements in the RUN group, then use the
RUN CANCEL statement, which terminates the procedure immediately.

The PLOT Procedure

Labeling Plot Points with Values of a Variable

625

You can use the BY statement interactively. The BY statement remains in effect
until you submit another BY statement or terminate the procedure.
See Example 11 on page 654 for an example of using RUN group processing with
PROC PLOT.

Generating Data with Program Statements

When you generate data to be plotted, a good rule is to generate fewer observations
than the number of positions on the horizontal axis. PROC PLOT then uses the
increment of the horizontal variable as the interval between tick marks.
Because PROC PLOT prints one character for each observation, using SAS program
statements to generate the data set for PROC PLOT can enhance the effectiveness of
continuous plots. For example, suppose that you want to generate data in order to plot
the following equation, for x ranging from 0 to 100:

= 2:54 + 3:83x

You can submit these statements:

options linesize=80;
data generate;
do x=0 to 100 by 2;
y=2.54+3.83*x;
output;
end;
run;
proc plot data=generate;
plot y*x;
run;

If the plot is printed with a LINESIZE= value of 80, then about 75 positions are
available on the horizontal axis for the X values. Thus, 2 is a good increment: 51
observations are generated, which is fewer than the 75 available positions on the
horizontal axis.
However, if the plot is printed with a LINESIZE= value of 132, then an increment of
2 produces a plot in which the plotting symbols have space between them. For a
smoother line, a better increment is 1, because 101 observations are generated.

Labeling Plot Points with Values of a Variable

Pointer Symbols
When you are using a label variable and do not specify a plotting symbol or if the
value of the variable you use as the plotting symbol is null (00x), PROC PLOT uses
pointer symbols as plotting symbols. Pointer symbols associate a point with its label by
pointing in the general direction of the label placement. PROC PLOT uses four
different pointer symbols based on the value of the S= and V= suboptions in the
PLACEMENT= option. The table below shows the pointer symbols:

626

Labeling Plot Points with Values of a Variable

Chapter 32

Symbol

LEFT

any

RIGHT

any

CENTER

<=0

If you are using pointer symbols and multiple points coincide, then PROC PLOT uses
the number of points as the plotting symbol if the number of points is between 2 and 9.
If the number of points is more than 9, then the procedure uses an asterisk (*).
Note: Because of character set differences among operating environments, the
pointer symbol for S=CENTER and V>0 may differ from the one shown here. 4

Understanding Penalties
PROC PLOT assesses the quality of placements with penalties. If all labels are
plotted with zero penalty, then no labels collide and all labels are near their symbols.
When it is not possible to place all labels with zero penalty, PROC PLOT tries to
minimize the total penalty. Table 32.2 on page 626 gives a description of the penalty,
the default value of the penalty, the index that you use to reference the penalty, and the
range of values that you can specify if you change the penalties. Each penalty is
described in more detail in Table 32.3 on page 627.
Table 32.2

Penalties Table

Penalty

Default penalty

Index

500

Labeling Plot Points with Values of a Variable

Table 32.3

Index Values for Penalties

a nonblank character in the plot collides with an embedded blank in a label, or there is not a blank or a
plot boundary before or after each label fragment.

a split occurs on a nonblank or nonpunctuation character when you do not specify a split character.

a label is placed with a different number of lines than the L= suboption species, when you specify a
split character.

4-7

627

a label is placed far away from the corresponding point. PROC PLOT calculates the penalty according to
this (integer arithmetic) formula:

[MAX (j

fhs

; 0) +

vsw

2 MAX (j

L fvs V

j0(

vhsd

)) =2; 0)] =

Notice that penalties 4 through 7 are actually just components of the formula used to determine the
penalty. Changing the penalty for a free horizontal or free vertical shift to a large value such as 500 has
the effect of removing any penalty for a large horizontal or vertical shift. Example 6 on page 640
illustrates a case in which removing the horizontal shift penalty is useful.
8

a label may collide with its own plotting symbol. If the plotting symbol is blank, then a collision state
cannot occur. See Collision States on page 628 for more information.

15-214

a label character does not appear in the plot. By default, the penalty for not printing the rst character
is greater than the penalty for not printing the second character, and so on. By default, the penalty for
not printing the fth and subsequent characters is the same.

Note:

Labels can share characters without penalty.

Changing Penalties
You can change the default penalties with the PENALTIES= option in the PLOT
statement. Because PROC PLOT considers penalties when it places labels, changing
the default penalties can change the placement of the labels. For example, if you have
labels that all begin with the same two-letter prex, then you might want to increase
the default penalty for not printing the third, fourth, and fth characters to 11, 10, and
8 and decrease the penalties for not printing the rst and second characters to 2. The
following PENALTIES= option accomplishes this change:
penalties(15 to 20)=2 2 11 10 8 2

This example extends the penalty list. The twentieth penalty of 2 is the penalty for
not printing the sixth through 200th character. When the last index i is greater than
18, the last penalty is used for the (i 14)th character and beyond.
You can also extend the penalty list by just specifying the starting index. For
example, the following PENALTIES= option is equivalent to the one above:
penalties(15)=2 2 11 10 8 2

628

Labeling Plot Points with Values of a Variable

Chapter 32

Collision States
Collision states are placement states that may cause a label to collide with its own
plotting symbol. PROC PLOT usually avoids using collision states because of the large
default penalty of 500 that is associated with them. PROC PLOT does not consider the
actual length or splitting of any particular label when determining if a placement state
is a collision state. The following are the rules that PROC PLOT uses to determine
collision states:

3 When S=CENTER, placement states that do not shift the label up or down
sufciently so that all of the label is shifted onto completely different lines from
the symbol are collision states.

3 When S=RIGHT, placement states that shift the label zero or more positions to the
left without rst shifting the label up or down onto completely different lines from
the symbol are collision states.

3 When S=LEFT, placement states that shift the label zero or more positions to the
right without rst shifting the label up or down onto completely different lines
from the symbol are collision states.

Note:

A collision state cannot occur if you do not use a plotting symbol.

Reference Lines
PROC PLOT places labels and computes penalties before placing reference lines on a
plot. The procedure does not attempt to avoid rows and columns that contain reference
lines.

Hidden Label Characters

In addition to the number of hidden observations and hidden plotting symbols, PROC
PLOT prints the number of hidden label characters. Label characters can be hidden by
plotting symbols or other label characters.

Overlaying Label Plots

When you overlay a label plot and a nonlabel plot, PROC PLOT tries to avoid
collisions between the labels and the characters of the nonlabel plot. When a label
character collides with a character in a nonlabel plot, PROC PLOT adds the usual
penalty to the penalty sum.
When you overlay two or more label plots, all label plots are treated as a single plot
in avoiding collisions and computing hidden character counts. Labels of different plots
never overprint, even with the OVP system option in effect.

Computational Resources Used for Label Plots

This section uses the following variables to discuss how much time and memory
PROC PLOT uses to construct label plots:
n

number of points with labels

len

constant length of labels

number of label pieces, or fragments

number of placement states specied in the PLACE= option.

The PLOT Procedure

ODS Table Names

629

Time
For a given plot size, the time that is required to construct the plot is roughly
proportional to n len. The amount of time required to split the labels is roughly
proportional to ns2 . Generally, the more placement states that you specify, the more
time that PROC PLOT needs to place the labels. However, increasing the number of
horizontal and vertical shifts gives PROC PLOT more exibility to avoid collisions,
often resulting in less time used to place labels.

Memory
PROC PLOT uses 24p bytes of memory for the internal placement state list. PROC
PLOT uses n (84 + 5len + 4s (1 + 1:5 (s + 1))) bytes for the internal list of labels.
PROC PLOT buildsall plots in memory; each printing position uses one byte of memory.
If you run out of memory, then request fewer plots in each PLOT statement and put a
RUN statement after each PLOT statement.

Results: PLOT Procedure

Scale of the Axes

Normally, PROC PLOT looks at the minimum difference between each pair of the ve
lowest ordered values of each variable (the delta) and ensures that there is no more
than one of these intervals per print position on the nal scaled axis, if possible. If
there is not enough room for this interval arrangement, and if PROC PLOT guesses
that the data was articially generated, then it puts a xed number of deltas in each
print position. Otherwise, PROC PLOT ignores the value.

Printed Output
Each plot uses one full page unless the plots size is changed by the VPOS= and
HPOS= options in the PLOT statement, the VPERCENT= or HPERCENT= options in
the PROC PLOT statement, or the PAGESIZE= and LINESIZE= system options. Titles,
legends, and variable labels are printed at the top of each page. Each axis is labeled
with the variables name or, if it exists, the variables label.
Normally, PROC PLOT begins a new plot on a new page. However, the VPERCENT=
and HPERCENT= options enable you to print more than one plot on a page.
VPERCENT= and HPERCENT= are described earlier in PROC PLOT Statement on
page 609.
PROC PLOT always begins a new page after a RUN statement and at the beginning
of a BY group.

ODS Table Names

The PLOT procedure assigns a name to each table that it creates. You can use these
names to reference the table when using the Output Delivery System (ODS) to select
tables and create output data sets. For more information, see SAS Output Delivery
System: Users Guide.

630

Portability of ODS Output with PROC PLOT

Table 32.4

Chapter 32

ODS Tables Produced by the PLOT Procedure

Table Name

Description

The PLOT procedure generates

the table:

Plot

A single plot

when you do not specify the

OVERLAY option.

Overlaid

Two or more plots on a single

set of axes

when you specify the

OVERLAY option.

Portability of ODS Output with PROC PLOT

Under certain circumstances, using PROC PLOT with the Output Delivery System
produces les that are not portable. If the SAS system option FORMCHAR= in your
SAS session uses nonstandard line-drawing characters, then the output might include
strange characters instead of lines in operating environments in which the SAS
Monospace font is not installed. To avoid this problem, specify the following OPTIONS
statement before executing PROC PLOT:
options formchar="|----|+|---+=|-/\<>*";

Missing Values
If values of either of the plotting variables are missing, then PROC PLOT does not
include the observation in the plot. However, in a plot of Y*X, values of X with
corresponding missing values of Y are included in scaling the X axis, unless the
NOMISS option is specied in the PROC PLOT statement.

Hidden Observations
By default, PROC PLOT uses different plotting symbols (A, B, C, and so on) to
represent observations whose values coincide on a plot. However, if you specify your
own plotting symbol or if you use the OVERLAY option, then you may not be able to
recognize coinciding values.
If you specify a plotting symbol, then PROC PLOT uses the same symbol regardless
of the number of observations whose values coincide. If you use the OVERLAY option
and overprinting is not in effect, then PROC PLOT uses the symbol from the rst plot
request. In both cases, the output includes a message telling you how many
observations are hidden.

The PLOT Procedure

Program

631

Examples: PLOT Procedure

Example 1: Specifying a Plotting Symbol

Procedure features:

PLOT statement
plotting symbol in plot request

This example expands on Output 32.1 by specifying a different plotting symbol.

Program

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. NUMBER enables printing of the page number. PAGENO= species the starting
page number. LINESIZE= species the output line length, and PAGESIZE= species the
number of lines on an output page.
options nodate number pageno=1 linesize=80 pagesize=35;

Create the DJIA data set. DJIA contains the high and low closing marks for the Dow Jones
Industrial Average from 1954 to 1994. A DATA step on page 1383 creates this data set.
data djia;
input Year @7 HighDate date7. High @24 LowDate date7. Low;
format highdate lowdate date7.;
datalines;
1954 31DEC54 404.39 11JAN54 279.87
1955 30DEC55 488.40 17JAN55 388.20
...more data lines...
1993 29DEC93 3794.33 20JAN93 3241.95
1994 31JAN94 3978.36 04APR94 3593.35
;

Create the plot. The plot request plots the values of High on the vertical axis and the values of
Year on the horizontal axis. It also species an asterisk as the plotting symbol.
proc plot data=djia;
plot high*year=*;

632

Output

Chapter 32

Specify the titles.

title High Values of the Dow Jones Industrial Average;
title2 from 1954 to 1994;
run;

Output

PROC PLOT determines the tick marks and the scale of both axes.

High Values of the Dow Jones Industrial Average

from 1954 to 1994
Plot of High*Year.

Symbol used is *.

High |
|
4000 +
*
|
*
|
|
*
|
*
3000 +
*
|
* *
|
|
|
*
2000 +
*
|
|
*
|
|
**
1000 +
***** *** *** ***
|
****
*
**
*
|
*****
|
**
|
0 +
|
---+---------+---------+---------+---------+---------+-1950
1960
1970
1980
1990
2000
Year

Example 2: Controlling the Horizontal Axis and Adding a Reference Line

Procedure features:

PLOT statement options:

HAXIS=
VREF=
Data set:

DJIA on page 631

The PLOT Procedure

Program

633

This example species values for the horizontal axis and draws a reference line from
the vertical axis.

Program

Customize the horizontal axis and draw a reference line. HAXIS= species that the
horizontal axis will show the values 1950 to 1995 in ve-year increments. VREF= draws a
reference line that extends from the value 3000 on the vertical axis.
/ haxis=1950 to 1995 by 5 vref=3000;

Specify the titles.

title High Values of Dow Jones Industrial Average;
title2 from 1954 to 1994;
run;

634

Output

Chapter 32

Output
High Values of Dow Jones Industrial Average
from 1954 to 1994
Plot of High*Year.

Symbol used is *.

High |
|
4000 +
*
|
*
|
|
*
|
*
3000 +----------------------------------------------------------------*--------|
* *
|
|
|
*
2000 +
*
|
|
*
|
|
**
1000 +
* ** **
** *
** * * **
|
** **
*
* *
*
|
** ** *
|
* *
|
0 +
|
-+-------+-------+-------+-------+-------+-------+-------+-------+-------+1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
Year

Example 3: Overlaying Two Plots

Procedure features:

PLOT statement options

BOX
OVERLAY
Data set: DJIA on page 631

This example overlays two plots and puts a box around the plot.

Program

The PLOT Procedure

Output

Create the plot.The rst plot request plots High on the vertical axis, plots Year on the
horizontal axis, and species an asterisk as a plotting symbol. The second plot request plots
Low on the vertical axis, plots Year on the horizontal axis, and species an o as a plotting
symbol. OVERLAY superimposes the second plot onto the rst. BOX draws a box around the
plot. OVERLAY and BOX apply to both plot requests.
proc plot data=djia;
plot high*year=*
low*year=o / overlay box;

Specify the titles.

title Plot of Highs and Lows;
title2 for the Dow Jones Industrial Average;
run;

Output
Plot of Highs and Lows
for the Dow Jones Industrial Average

Plot of HighYear. Symbol used is .

635

636

Example 4: Producing Multiple Plots per Page

Chapter 32

Example 4: Producing Multiple Plots per Page

Procedure features:

PROC PLOT statement options

HPERCENT=
VPERCENT=
Data set: DJIA on page 631

This example puts three plots on one page of output.

Program

Specify the plot sizes. VPERCENT= species that 50% of the vertical space on the page of
output is used for each plot. HPERCENT= species that 50% of the horizontal space is used for
each plot.
proc plot data=djia vpercent=50 hpercent=50;

Create the rst plot. This plot request plots the values of High on the vertical axis and the
values of Year on the horizontal axis. It also species an asterisk as the plotting symbol.
plot high*year=*;

Create the second plot.This plot request plots the values of Low on the vertical axis and the
values of Year on the horizontal axis. It also species an asterisk as the plotting symbol.
plot low*year=o;

The PLOT Procedure

Program

637

Create the third plot. The rst plot request plots High on the vertical axis, plots Year on the
horizontal axis, and species an asterisk as a plotting symbol. The second plot request plots
Low on the vertical axis, plots Year on the horizontal axis, and species an o as a plotting
symbol. OVERLAY superimposes the second plot onto the rst. BOX draws a box around the
plot. OVERLAY and BOX apply to both plot requests.
plot high*year=* low*year=o / overlay box;

Specify the titles.

title Plots of the Dow Jones Industrial Average;
title2 from 1954 to 1994;
run;

638

Output

Chapter 32

Output

Plots of the Dow Jones Industrial Average

from 1954 to 1994
Plot of High*Year.

Symbol used is *.

Plot of Low*Year.

4000 +
|
|

*
*
*

|
|
|
|
|
|

********

***

** ***

|
|
|

******
****

|
0 +
-+---------+---------+---------+---------+---------+1950

1960

1970

1980

1990

2000

Year

Plot of High*Year.
Plot of Low*Year.

Symbol used is *.

-+---------+---------+---------+---------+---------+4000 +
|
|

*
*
* o

|
High |

*oo
*

|
|
|

|
|
|
|
|
0 +

* *
o
*oo
* o
o
*o
**o
****** ************oo
*****oooooo*o o oooooooo
*****oooo
o

+
|
|
|
|
|
|
|
+
|
|
|
|
|
|
|
+

-+---------+---------+---------+---------+---------+1950
1960
1970
1980
1990
2000
Year
NOTE: 7 obs hidden.

oo
o
ooo
o oo ooo oo o o
ooo oo o oo
oo o o o
oooo
o

|
o
0 +
-+---------+---------+---------+---------+---------+1950

1960

1970

1980
Year

Symbol used is o.

2000 +
|
|

oo
o

|
|
|

**
*

2000 +
|

*
**

|
|
|

2000 +
|

|
Low |
|

|
|
|

Symbol used is o.

4000 +
|
|

|
High |
|

1990

2000

The PLOT Procedure

Program

639

Example 5: Plotting Data on a Logarithmic Scale

Procedure features:

PLOT statement option

HAXIS=

This example uses a DATA step to generate data. The PROC PLOT step shows two
plots of the same data: one plot without a horizontal axis specication and one plot
with a logarithmic scale specied for the horizontal axis.

Program

Create the EQUA data set. EQUA contains values of X and Y. Each value of X is calculated as
Y
10 .
data equa;
do Y=1 to 3 by .1;
X=10**y;
output;
end;
run;

Specify the plot sizes. HPERCENT= makes room for two plots side-by-side by specifying that
50% of the horizontal space is used for each plot.
proc plot data=equa hpercent=50;

Create the plots. The plot requests plot Y on the vertical axis and X on the horizontal axis.
HAXIS= species a logarithmic scale for the horizontal axis for the second plot.
plot y*x;
plot y*x / haxis=10 100 1000;

Specify the titles.

title Two Plots with Different;
title2 Horizontal Axis Specifications;
run;

640

Output

Chapter 32

Output
Two Plots with Different
Horizontal Axis Specifications
Plot of Y*X.

A=1, B=2, etc.

Y |
|
3.0 +
A
2.9 +
A
2.8 +
A
2.7 +
A
2.6 +
A
2.5 +
A
2.4 +
A
2.3 +
A
2.2 +
A
2.1 +
A
2.0 +
A
1.9 +
A
1.8 + A
1.7 + A
1.6 + A
1.5 + A
1.4 + A
1.3 + A
1.2 + A
1.1 +A
1.0 +A
|
-+---------------+---------------+
0
500
1000

Plot of Y*X.

A=1, B=2, etc.

Y |
|
3.0 +
A
2.9 +
A
2.8 +
A
2.7 +
A
2.6 +
A
2.5 +
A
2.4 +
A
2.3 +
A
2.2 +
A
2.1 +
A
2.0 +
A
1.9 +
A
1.8 +
A
1.7 +
A
1.6 +
A
1.5 +
A
1.4 +
A
1.3 +
A
1.2 +
A
1.1 + A
1.0 +A
|
-+---------------+---------------+
10
100
1000

Example 6: Plotting Date Values on an Axis

Procedure features:

PLOT statement option

HAXIS=

This example shows how you can specify date values on an axis.

The PLOT Procedure

Program

641

Program

Create the EMERGENCY_CALLS data set. EMERGENCY_CALLS contains the number of

telephone calls to an emergency help line for each date.
data emergency_calls;
input Date : date7. Calls @@;
label calls=Number of Calls;
datalines;
1APR94 134
11APR94 384 13FEB94
2MAR94 289
21MAR94 201 14MAR94
3JUN94 184
13JUN94 152 30APR94
4JAN94 179
14JAN94 128 16JUN94
5APR94 360
15APR94 350 24JUL94
6MAY94 245
15DEC94 150 17NOV94
7JUL94 280
16MAY94 240 25AUG94
8AUG94 494
17JUL94 499 26SEP94
9SEP94 309
18AUG94 248 23NOV94
19SEP94 356 24FEB94 201 29JUL94
10OCT94 222 25MAR94 183 30AUG94
11NOV94 294 26APR94 412
2DEC94
27MAY94 294 22DEC94 413 28JUN94
;

488
460
356
480
388
328
280
394
590
330
321
511
309

Create the plot. The plot request plots Calls on the vertical axis and Date on the horizontal
axis. HAXIS= uses a monthly time for the horizontal axis. The notation 1JAN94d is a date
constant. The value 1JAN95d ensures that the axis will have enough room for observations
from December.
proc plot data=emergency_calls;
plot calls*date / haxis=1JAN94d to 1JAN95d by month;

Format the DATE values. The FORMAT statement assigns the DATE7. format to Date.
format date date7.;

Specify the titles.

title Calls to City Emergency Services Number;
title2 Sample of Days for 1994;
run;

642

Output

Chapter 32

Output

PROC PLOT uses the variables labels on the axes.

Calls to City Emergency Services Number

Sample of Days for 1994
Plot of Calls*Date.

Legend: A = 1 obs, B = 2 obs, etc.

|
|
600 +

|
|
|
|
N 500 +
u
m
b

A
A

|
|
|

A
A

e
|
r 400 +
|
o
f

A
A

|
|
|

|
|
100 +

A
A

A
A
A
A

A
A

|
|
200 +
|
|

First 5 Observations Only
Z

46.2
47.2
48.0
48.8
49.4

0
0
0
0
0

0
10
20
30
40

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. NOOVP ensures
that overprinting is not used in the plot.
options nodate pageno=1 linesize=120 pagesize=60 noovp;

Create the plot. The plot request plots Y on the vertical axis, plots X on the horizontal axis,
and species Z as the contour variable. CONTOUR=10 species that the plot will divide the
values of Z into ten increments, and each increment will have a different plotting symbol.
proc plot data=contours;
plot y*x=z / contour=10;

Specify the title.

title A Contour Plot;
run;

The PLOT Procedure

Output

The shadings associated with the values of Z appear at the bottom of the plot. The plotting symbol # shows
where high values of Z occur.

A Contour Plot

Contour plot of Y*X.

Y |
|
350 +
340 +

======++++++OOOOOOOOXXXXXXXXXXXWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWXXXXXXXXXXXOOOOOOOO
====++++++OOOOOOOXXXXXXXXXXWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWXXXXXXXXXXOOOOOOO

330 +
320 +
310 +

=++++++OOOOOOOXXXXXXXXXWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWXXXXXXXXXOOOOO
+++++OOOOOOOXXXXXXXXWWWWWWWWWWWWWW********************WWWWWWWWWWWWWWXXXXXXXXXOOOO
+++OOOOOOXXXXXXXXWWWWWWWWWWWW*****************************WWWWWWWWWWWXXXXXXXXOOOO

300 +
290 +
280 +

+OOOOOOXXXXXXXXWWWWWWWWWW***********************************WWWWWWWWWWXXXXXXXXOOO
OOOOOXXXXXXXWWWWWWWWWW****************************************WWWWWWWWWXXXXXXXOOO
OOOXXXXXXXWWWWWWWWW********************####********************WWWWWWWWWXXXXXXXOO

270 +
260 +
250 +

OXXXXXXXWWWWWWWWW**************##################***************WWWWWWWWXXXXXXXOO
XXXXXXWWWWWWWW*************#########################************WWWWWWWWXXXXXXXOO
XXXXWWWWWWWW************#############################************WWWWWWWWXXXXXXOO

240 +
230 +
220 +

XXXWWWWWWW***********#################################***********WWWWWWWWXXXXXXOO
XWWWWWWWW**********####################################**********WWWWWWWXXXXXXXOO
WWWWWWW**********######################################**********WWWWWWWXXXXXXOOO

210 +
200 +

WWWWWW*********########################################**********WWWWWWWXXXXXXOOO
WWWWW*********#########################################*********WWWWWWWXXXXXXOOOO

190 +
180 +
170 +

WWW**********##########################################*********WWWWWWWXXXXXXOOOO
WW*********###########################################*********WWWWWWWXXXXXXOOOOO
W*********############################################*********WWWWWWWXXXXXXOOOOO

160 +
150 +
140 +

W*********###########################################*********WWWWWWWXXXXXXOOOOO+
*********###########################################*********WWWWWWWXXXXXXOOOOO++
********###########################################*********WWWWWWWXXXXXXOOOOO+++

130 +
120 +
110 +

********##########################################*********WWWWWWWXXXXXXOOOOO++++
********########################################**********WWWWWWWXXXXXXOOOOO+++++
********#######################################**********WWWWWWWXXXXXXOOOOO+++++=

100 +
90 +

********#####################################**********WWWWWWWXXXXXXOOOOOO+++++==
********###################################**********WWWWWWWWXXXXXXOOOOO+++++====

80 +
70 +
60 +

*********################################***********WWWWWWWXXXXXXXOOOOO+++++====**********############################************WWWWWWWWXXXXXXOOOOOO+++++====-************######################**************WWWWWWWWXXXXXXXOOOOO+++++=====---

50 +
40 +
30 +

***************###############***************WWWWWWWWWXXXXXXXOOOOOO+++++====----
W******************************************WWWWWWWWWXXXXXXXOOOOOO+++++=====----
WW**************************************WWWWWWWWWWXXXXXXXOOOOOO+++++=====----

20 +
10 +
0 +

WWWW********************************WWWWWWWWWWWXXXXXXXXOOOOOO++++++====-----.
WWWWWW**************************WWWWWWWWWWWWWXXXXXXXXOOOOOO++++++=====----...
WWWWWWWWWW*****************WWWWWWWWWWWWWWWXXXXXXXXOOOOOOO++++++=====----....

|
---+---------+---------+---------+---------+---------+---------+---------+---------+-0
50
100
150
200
250
300
350
400
X
Symbol
.....

z
2.2 - 8.1
8.1 - 14.0

Symbol
----=====

z
14.0 - 19.9
19.9 - 25.8

Symbol
+++++
OOOOO

z
25.8 - 31.7
31.7 - 37.6

Symbol
XXXXX
WWWWW

z
37.6 - 43.5
43.5 - 49.4

Symbol
*****
#####

z
49.4 - 55.4
55.4 - 61.3

645

646

Example 8: Plotting BY Groups

Chapter 32

Example 8: Plotting BY Groups

Procedure features:

PLOT statement option

HREF=
Other features:

BY statement

This example shows BY group processing in PROC PLOT.

Program

Create the EDUCATION data set. EDUCATION contains educational data* about some U.S.
states. DropoutRate is the percentage of high school dropouts. Expenditures is the dollar
amount the state spends on each pupil. MathScore is the score of eighth-grade students on a
standardized math test. Not all states participated in the math test. A DATA step on page 1384
creates this data set.
data education;
input State $14. +1 Code $ DropoutRate Expenditures MathScore
Region $;
label dropout=Dropout Percentage - 1989
expend=Expenditure Per Pupil - 1989
math=8th Grade Math Exam - 1990;
datalines;
Alabama
AL 22.3 3197 252 SE
Alaska
AK 35.8 7716 .
W
...more data lines...
New York
NY 35.0 .
261 NE
North Carolina NC 31.2 3874 250 SE
North Dakota
ND 12.1 3952 281 MW
Ohio
OH 24.4 4649 264 MW
;

* Source: U.S. Department of Education.

The PLOT Procedure

Program

647

Sort the EDUCATION data set. PROC SORT sorts EDUCATION by Region so that Region
can be used as the BY variable in PROC PLOT.
proc sort data=education;
by region;
run;

Create a separate plot for each BY group. The BY statement creates a separate plot for
each value of Region.
proc plot data=education;
by region;

Create the plot with a reference line. The plot request plots Expenditures on the vertical
axis, plots DropoutRate on the horizontal axis, and species an asterisk as the plotting symbol.
HREF= draws a reference line that extende from 28.6 on the horizontal axis. The reference line
represents the national average.
plot expenditures*dropoutrate=* / href=28.6;

Specify the title.

title Plot of Dropout Rate and Expenditure Per Pupil;
run;

648

Output

Chapter 32

Output

PROC PLOT produces a plot for each BY group. Only the plots for Midwest and Northeast
are shown.

Plot of Dropout Rate and Expenditure Per Pupil

---------------------------------- Region=MW ----------------------------------Plot of Expenditures*DropoutRate.

Symbol used is *.

Expenditures |
|
5500 +
|
|
|
|
|
|
|
|
| *
5000 +
|
|
*
|
|
*
|
|
|
|
*
|
4500 +
|
|
*
*
|
|
**
*
|
|
|
|
|
4000 +
*
|
|
|
|
|
|
|
|
|
3500 +
|
|
|
---+------------+------------+------------+------------+-10
15
20
25
30
Dropout Percentage - 1989

The PLOT Procedure

Program

Plot of Dropout Rate and Expenditure Per Pupil

649

---------------------------------- Region=NE ----------------------------------Plot of Expenditures*DropoutRate.

Symbol used is *.

Expenditures |
|
8000 +
|
|
|
|
*
|
|
|
|
|
7000 +
|
|
*
|
|
|
|
|
|
|
6000 +
*|
|
*
|
|
|
|
*
|
|
5000 +
|
|
*
*
|
|
|
|
|
|
|
4000 +
|
|
|
---+------------+------------+------------+------------+-15
20
25
30
35
Dropout Percentage - 1989
NOTE: 1 obs had missing values.

Example 9: Adding Labels to a Plot

Procedure features:

PLOT statement
label variable in plot request
Data set: EDUCATION on page 646

This example shows how to modify the plot request to label points on the plot with
the values of variables. This example adds labels to the plot shown in Example 8 on
page 646.

Program

650

Program

Chapter 32

Sort the EDUCATION data set. PROC SORT sorts EDUCATION by Region so that Region
can be used as the BY variable in PROC PLOT.
proc sort data=education;
by region;
run;

Create a separate plot for each BY group. The BY statement creates a separate plot for
each value of Region.
proc plot data=education;
by region;

Create the plot with a reference line and a label for each data point. The plot request
plots Expenditures on the vertical axis, plots DropoutRate on the horizontal axis, and species
an asterisk as the plotting symbol. The label variable specication ($ state) in the PLOT
statement labels each point on the plot with the name of the corresponding state. HREF= draws
a reference line that extends from 28.6 on the horizontal axis. The reference line represents the
national average.
plot expenditures*dropoutrate=* $ state / href=28.6;

Specify the title.

title Plot of Dropout Rate and Expenditure Per Pupil;
run;

The PLOT Procedure

Output

651

Output

PROC PLOT produces a plot for each BY group. Only the plots for Midwest and Northeast are
shown.

Plot of Dropout Rate and Expenditure Per Pupil

---------------------------------- Region=MW ----------------------------------Plot of Expenditures*DropoutRate$State.

Symbol used is *.

652

Example 10: Excluding Observations That Have Missing Values

Chapter 32

Plot of Dropout Rate and Expenditure Per Pupil

---------------------------------- Region=NE ----------------------------------Plot of Expenditures*DropoutRate$State.

Symbol used is *.

Example 10: Excluding Observations That Have Missing Values

Procedure features:

PROC PLOT statement option

NOMISS
Data set:

EDUCATION on page 646

This example shows how missing values affect the calculation of the axes.

Program

The PLOT Procedure

Program

653

Sort the EDUCATION data set. PROC SORT sorts EDUCATION by Region so that Region
can be used as the BY variable in PROC PLOT.
proc sort data=education;
by region;
run;

Exclude data points with missing values. NOMISS excludes observations that have a
missing value for either of the axis variables.
proc plot data=education nomiss;

Create a separate plot for each BY group. The BY statement creates a separate plot for
each value of Region.
by region;

Create the plot with a reference line and a label for each data point. The plot request
plots Expenditures on the vertical axis, plots DropoutRate on the horizontal axis, and species
an asterisk as the plotting symbol. The label variable specication ($ state) in the PLOT
statement labels each point on the plot with the name of the corresponding state. HREF= draws
a reference line extending from 28.6 on the horizontal axis. The reference line represents the
national average.
plot expenditures*dropoutrate=* $ state / href=28.6;

Specify the title.

title Plot of Dropout Rate and Expenditure Per Pupil;
run;

654

Output

Chapter 32

Output

PROC PLOT produces a plot for each BY group. Only the plot for the Northeast is shown.
Because New York has a missing value for Expenditures, the observation is excluded and
PROC PLOT does not use the value 35 for DropoutRate to calculate the horizontal axis.
Compare the horizontal axis in this output with the horizontal axis in the plot for Northeast in
Example 9 on page 649.

Plot of Dropout Rate and Expenditure Per Pupil

---------------------------------- Region=NE ----------------------------------Plot of Expenditures*DropoutRate$State.

Symbol used is *.

Example 11: Adjusting Labels on a Plot with the PLACEMENT= Option

Procedure features:

PLOT statement options

label variable in plot request
LIST=
PLACEMENT=
Other features:

RUN group processing

The PLOT Procedure

Program

655

This example illustrates the default placement of labels and how to adjust the
placement of labels on a crowded plot. The labels are values of variable in the data set.*
This example also shows RUN group processing in PROC PLOT.

Program

Create the CENSUS data set. CENSUS contains the variables CrimeRate and Density for
selected states. CrimeRate is the number of crimes per 100,000 people. Density is the population
density per square mile in the 1980 census. A DATA step on page 1377 creates this data set.
data census;
input Density CrimeRate State $ 14-27 PostalCode $ 29-30;
datalines;
263.3 4575.3 Ohio
OH
62.1 7017.1 Washington
WA
...more data lines...
111.6 4665.6 Tennessee
TN
120.4 4649.9 North Carolina NC
;

Create the plot with a label for each data point. The plot request plots Density on the
vertical axis, CrimeRate on the horizontal axis, and uses the rst letter of the value of State as
the plotting symbol. This makes it easier to match the symbol with its label. The label variable
specication ($ state) in the PLOT statement labels each point with the corresponding state
name.
proc plot data=census;
plot density*crimerate=state $ state /

Specify plot options. BOX draws a box around the plot. LIST= lists the labels that have
penalties greater than or equal to 1. HAXIS= and VAXIS= specify increments only. PROC PLOT
uses the data to determine the range for the axes.
box
list=1
haxis=by 1000
vaxis=by 250;

* Source: U.S. Bureau of the Census and the 1987 Uniform Crime Reports, FBI.

656

Program

Chapter 32

Specify the title.

title A Plot of Population Density and Crime Rates;
run;

The labels Tennessee, South Carolina, Arkansas, Minnesota, and South Dakota have penalties. The
default placement states do not provide enough possibilities for PROC PLOT to avoid penalties given the
proximity of the points. Seven label characters are hidden.

A Plot of Population Density and Crime Rates

Plot of Density*CrimeRate$State.

Symbol is value of State.

---+------------+------------+------------+------------+------------+------------+------------+--Density |
|
500 +
|
|

+
|
|

|
|
|

M Maryland

|
|
|

D Delaware
P Pennsylvania

|
|
|

O Ohio

250 +
|

+
|

|
|
|

I Illinois

|
|
|
|
|
|
|
0 +

North Carolina
TennNssee
Georgia
N New Hampshire T
S South Garolina
W West Virginia
Mississippi M

A Alabama
Vermont V
M Missouri
MinneAoArkMnsas

North Dakota
S Nouth Dakota

|
F Florida|
|
C California

|
|
|

Washington W
O Oklahoma

I Idaho

|
|
|

T Texas

O Oregon

|
+

N Nevada

---+------------+------------+------------+------------+------------+------------+------------+--2000
3000
4000
5000
6000
7000
8000
9000
CrimeRate
NOTE: 7 label characters hidden.

A Plot of Population Density and Crime Rates

List of Point Locations, Penalties, and Placement States

Label
Tennessee
South Carolina
Arkansas
Minnesota
South Dakota

Vertical
Axis

Horizontal
Axis

Penalty

111.60
103.40
43.90

4665.6
5161.9
4245.2

2
2
6

51.20
9.10

4615.8
2678.0

7
11

Starting
Position

Lines

Vertical
Shift

Horizontal
Shift

Center
Right
Right

1
1
1

1
0
0

-1
2
2

Left
Right

1
1

0
0

-2
2

The PLOT Procedure

Program

657

Request a second plot. Because PROC PLOT is interactive, the procedure is still running at
this point in the program. It is not necessary to restart the procedure to submit another plot
request. LIST=1 produces no output because there are no penalties of 1 or greater.
plot density*crimerate=state $ state /
box
list=1
haxis=by 1000
vaxis=by 250

Specify placement options. PLACEMENT= gives PROC PLOT more placement states to use
to place the labels. PLACEMENT= contains three expressions. The rst expression species the
preferred positions for the label. The rst expression resolves to placement states centered
above the plotting symbol, with the label on one or two lines. The second and third expressions
resolve to placement states that enable PROC PLOT to place the label in multiple positions
around the plotting symbol.
placement=((v=2
((l=2 2 1
(s=center
h=0 1 to

1 : l=2 1)
: v=0 1 0) * (s=right left : h=2 -2))
right left * l=2 1 * v=0 1 -1 2 *
5 by alt));

Specify the title.

title A Plot of Population Density and Crime Rates;
run;

658

Output

Chapter 32

Output

No collisions occur in the plot.

A Plot of Population Density and Crime Rates

Plot of Density*CrimeRate$State.

Symbol is value of State.

---+------------+------------+------------+------------+------------+------------+------------+--Density |
|
500 +
|
|

+
|
|

Maryland

|
|

|
|
|

Pennsylvania

|
250 +
|

|
+
|

Florida
F

North Carolina
New
Hampshire

West

| Virginia
|
W
|
|
|
0 +

|
|
|

Illinois

|
|
|
|
|

Delaware
D
Ohio

Alabama N

California

South
Carolina

N
T
S
Mississippi
A Tennessee
M Vermont V
M Missouri

|
|

G Georgia
Oklahoma

South
Arkansas A
M Minnesota
Dakota
I Idaho
S N North Dakota

O
Nevada
N

|
|
|

Washington Texas
W
T
Oregon
O

|
|
|
|
|
+

---+------------+------------+------------+------------+------------+------------+------------+--2000
3000
4000
5000
6000
7000
8000
9000
CrimeRate

Example 12: Adjusting Labeling on a Plot with a Macro

Procedure features:

PLOT statement options

label variable in plot request
PLACEMENT=
Data set:

CENSUS on page 655

This example illustrates the default placement of labels and uses a macro to adjust
the placement of labels. The labels are values of a variable in the data set.

The PLOT Procedure

Program

659

Program

Use conditional logic to determine placement. The %PLACE macro provides an alternative
to using the PLACEMENT= option. The higher the value of n, the more freedom PROC PLOT
has to place labels.
%macro place(n);
%if &n > 13 %then %let n = 13;
placement=(
%if &n <= 0 %then (s=center); %else (h=2 -2 : s=right left);
%if &n = 1 %then (v=1 * h=0 -1 to -2 by alt);
%else %if &n = 2 %then (v=1 -1 * h=0 -1 to -5 by alt);
%else %if &n > 2 %then (v=1 to 2 by alt * h=0 -1 to -10 by alt);
%if &n > 3 %then
(s=center right left * v=0 1 to %eval(&n - 2) by alt *
h=0 -1 to %eval(-3 * (&n - 2)) by alt *
l=1 to %eval(2 + (10 * &n - 35) / 30)); )
%if &n > 4 %then penalty(7)=%eval((3 * &n) / 2);
%mend;

Create the plot. The plot request plots Density on the vertical axis, CrimeRate on the
horizontal axis, and uses the rst letter of the value of State as the plotting symbol. The label
variable specication ($ state) in the PLOT statement t labels each point with the
corresponding state name.
proc plot data=census;
plot density*crimerate=state $ state /

660

Output

Chapter 32

Specify the title.

title A Plot of Population Density and Crime Rates;
run;

Output

No collisions occur in the plot.

A Plot of Population Density and Crime Rates

Plot of Density*CrimeRate$State.

Symbol is value of State.

---+------------+------------+------------+------------+------------+------------+------------+--Density |
|
500 +
+
|
|

|
|

|
|
|

M Maryland

|
|
|

D Delaware
P Pennsylvania

|
|
|

O Ohio

250 +
|
|

+
|
|

I Illinois

|
|

F Florida|
|

|
|
|

North Carolina
N Tennessee
N New Hampshire T
S

|
|
|

W West Virginia
Mississippi M

C California

|
|
|

G Georgia

Alabama A South Carolina

Vermont V
M Missouri
Arkansas A
M Minnesota

Washington W
O Oklahoma

|
|
|

T Texas

|
South Dakota
I Idaho
O Oregon
|
0 +
S N North Dakota
N Nevada
+
---+------------+------------+------------+------------+------------+------------+------------+--2000

3000

4000

5000

6000
CrimeRate

7000

8000

9000

The PLOT Procedure

Program

661

Example 13: Changing a Default Penalty

Procedure features:

PLOT statement option

PENALTIES=
Data set: CENSUS on page 655

This example demonstrates how changing a default penalty affects the placement of
labels. The goal is to produce a plot that has labels that do not detract from how the
points are scattered.

Program

Create the plot. The plot request plots Density on the vertical axis, CrimeRate on the
horizontal axis, and uses the rst letter of the value of State as the plotting symbol. The label
variable specication ($ state) in the PLOT statement labels each point with the
corresponding state name.
proc plot data=census;
plot density*crimerate=state $ state /

662

Program

Chapter 32

Specify the placement. PLACEMENT= species that the preferred placement states are 100
columns to the left and the right of the point, on the same line with the point.
placement=(h=100 to 10 by alt * s=left right)

Change the default penalty. PENALTIES(4)= changes the default penalty for a free
horizontal shift to 500, which removes all penalties for a horizontal shift. LIST= shows how far
PROC PLOT shifted the labels away from their respective points.
penalties(4)=500

list=0

Customize the axes. HAXIS= creates a horizontal axis long enough to leave space for the
labels on the sides of the plot. VAXIS= species that the values on the vertical axis be in
increments of 100.
haxis=0 to 13000 by 1000
vaxis=by 100;

Specify the title.

title A Plot of Population Density and Crime Rates;
run;

The PLOT Procedure

Output

A Plot of Population Density and Crime Rates

Plot of Density*CrimeRate$State.

Symbol is value of State.

Density |
500 +
|
|
|
|
400 +

Maryland

|
|
|
|
300 +

|
|
|

Pennsylvania Ohio

|
200 +
|Florida

|Washington Texas
|Oklahoma
|Oregon

Illinois
F

|
|
|
100 +Georgia
|

Delaware

California

T
N

North Carolina Tennessee

New Hampshire South Carolina

Alabama Missouri West Virginia

A M
M

V
A
I

W
O
O

Vermont Minnesota Mississippi

Arkansas
Idaho

0 +
S N
N
North Dakota South Dakota Nevada
---+-------+-------+-------+-------+-------+-------+-------+-------+-------+-------+-------+-------+-------+-0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
11000
12000
13000
CrimeRate
NOTE: 1 obs hidden.

663

664

Output

Chapter 32

A Plot of Population Density and Crime Rates

List of Point Locations, Penalties, and Placement States

Vertical
Axis

Horizontal
Axis

Penalty

Lines

Vertical
Shift

Horizontal
Shift

Maryland
Delaware

428.70
307.60

5477.6
4938.8

0
0

Right
Right

1
1

0
0

55
59

Pennsylvania
Ohio

264.30
263.30

3163.2
4575.3

0
0

Right
Right

1
1

0
0

65
66

Illinois
Florida
California

205.30
180.00
151.40

5416.5
8503.2
6506.4

0
0
0

Right
Left
Right

1
1
1

0
0
0

56
-64
45

Tennessee
North Carolina
New Hampshire

111.60
120.40
102.40

4665.6
4649.9
3371.7

0
0
0

Right
Right
Right

1
1
1

0
0
0

61
46
52

South Carolina
Georgia
West Virginia

103.40
94.10
80.80

5161.9
5792.0
2190.7

0
0
0

Right
Left
Right

1
1
1

0
0
0

52
-42
76

Alabama
Missouri
Mississippi

76.60
71.20
53.40

4451.4
4707.5
3438.6

0
0
0

Right
Right
Right

1
1
1

0
0
0

41
47
68

Vermont
Minnesota

55.20
51.20

4271.2
4615.8

0
0

Right
Right

1
1

0
0

44
49

Washington
Texas
Arkansas

62.10
54.30
43.90

7017.1
7722.4
4245.2

0
0
0

Left
Left
Right

1
1
1

0
0
0

-49
-49
65

Oklahoma
Idaho
Oregon

44.10
11.50
27.40

6025.6
4156.3
6969.9

0
0
0

Left
Right
Left

1
1
1

0
0
0

-43
69
-53

9.10
9.40
7.30

2678.0
2833.0
6371.4

0
0
0

Right
Right
Right

1
1
1

0
0
0

67
52
50

Label

South Dakota
North Dakota
Nevada

Starting
Position

665

CHAPTER

33
The PMENU Procedure
Overview: PMENU Procedure 665
Syntax: PMENU Procedure 666
PROC PMENU Statement 667
CHECKBOX Statement 668
DIALOG Statement 668
ITEM Statement 670
MENU Statement 673
RADIOBOX Statement 675
RBUTTON Statement 675
SELECTION Statement 676
SEPARATOR Statement 677
SUBMENU Statement 677
TEXT Statement 678
Concepts: PMENU Procedure 679
Procedure Execution 679
Initiating the Procedure 679
Ending the Procedure 680
Steps for Building and Using PMENU Catalog Entries 680
Templates for Coding PROC PMENU Steps 681
Examples: PMENU Procedure 682
Example 1: Building a Menu Bar for an FSEDIT Application 682
Example 2: Collecting User Input in a Dialog Box 685
Example 3: Creating a Dialog Box to Search Multiple Variables 688
Example 4: Creating Menus for a DATA Step Window Application 694
Example 5: Associating Menus with a FRAME Application 700

Overview: PMENU Procedure

The PMENU procedure denes menus that can be used in DATA step windows,
macro windows, both SAS/AF and SAS/FSP windows, or in any SAS application that
enables you to specify customized menus.
Menus can replace the command line as a way to execute commands. To activate
menus, issue the PMENU command from any command line. Menus must be activated
in order for them to appear.
When menus are activated, each active window has a menu bar, which lists items
that you can select. Depending upon which item you select, SAS either processes a
command, displays a menu or a submenu, or requests that you complete information in
a dialog box. The dialog box is simply a box of questions or choices that require answers
before an action can be performed. The following gure illustrates features that you can
create with PROC PMENU.

666

Syntax: PMENU Procedure

Figure 33.1

Chapter 33

Menu Bar, Pull-Down Menu, and Dialog Box

Menu bar
Dialog box
File

Edit

Reports

Help

Farm
Industrial...
Manufacturing...
Select a commodity:

Wheat
Corn
Oats

Select a market:

Farmville
Monticello
Plainview

Enter a year from 1950 to 1996:

Check here for double spacing:
pull-down
menu

Cancel

Note: A menu bar in some operating environments may appear as a popup menu or
may appear at the bottom of the window. 4
The PMENU procedure produces no immediately visible output. It simply builds a
catalog entry of type PMENU that can be used later in an application.

Syntax: PMENU Procedure

Restriction:

You must use at least one MENU statement followed by at least one ITEM

statement.
Tip:

Supports RUN group processing

You can also use appropriate global statements with this procedure. See
Chapter 2, Fundamental Concepts for Using Base SAS Procedures, on page 15 for a
list.

Reminder:

See:

PMENU Procedure in the documentation for your operating environment.

PROC PMENU <CATALOG=< libref.>catalog>

<DESC entry-description>;
MENU menu-bar;
ITEM command <option(s)>;
ITEM menu-item <option(s)>;
DIALOG dialog-box command-string
eld-number-specication;
CHECKBOX <ON> #line @column
text-for-selection
<COLOR=color> <SUBSTITUTE=text-for-substitution>;
RADIOBOX DEFAULT=button-number;
RBUTTON <NONE> #line @column
text-for-selection < COLOR=color>
<SUBSTITUTE=text-for-substitution>;
TEXT #line @column eld-description
<ATTR=attribute> <COLOR=color>;

The PMENU Procedure

PROC PMENU Statement

667

MENU pull-down-menu;
SELECTION selection command-string;
SEPARATOR;
SUBMENU submenu-name SAS-le;

To do this

Use this statement

Dene choices a user can make in a dialog box

CHECKBOX

Describe a dialog box that is associated with an item

in a pull-down menu

DIALOG

Identify an item to be listed in a menu bar or in a

pull-down menu

ITEM

Name the catalog entry or dene a pull-down menu

List and dene mutually exclusive choices within a

dialog box

RADIOBOX and RBUTTON

Dene a command that is submitted when an item is

selected

SELECTION

Draw a line between items in a pull-down menu

SEPARATOR

Dene a common submenu associated with an item

SUBMENU

Specify text and the input elds for a dialog box

TEXT

PROC PMENU Statement

Invokes the PMENU procedure and species where to store all PMENU catalog entries that are
created in the PROC PMENU step.
PROC PMENU <CATALOG=< libref.>catalog>
<DESC entry-description>;

Options
CATALOG=<libref.>catalog

species the catalog in which you want to store PMENU entries.

Default: If you omit libref, then the PMENU entries are stored in a catalog in the
SASUSER data library. If you omit CATALOG=, then the entries are stored in the
SASUSER.PROFILE catalog.
Featured in: Example 1 on page 682
DESC entry-description

provides a description for the PMENU catalog entries created in the step.
Default: Menu description

Note: These descriptions are displayed when you use the CATALOG window in
the windowing environment or the CONTENTS statement in the CATALOG
procedure. 4

668

CHECKBOX Statement

Chapter 33

CHECKBOX Statement
Denes choices that a user can make within a dialog box.
Restriction:

Must be used after a DIALOG statement.

CHECKBOX <ON> #line @column

text-for-selection
<COLOR=color> <SUBSTITUTE=text-for-substitution>;

Required Arguments
column

species the column in the dialog box where the check box and text are placed.
line

species the line in the dialog box where the check box and text are placed.
text-for-selection

denes the text that describes this check box. This text appears in the window and,
if the SUBSTITUTE= option is not used, is also inserted into the command in the
preceding DIALOG statement when the user selects the check box.

Options
COLOR=color

denes the color of the check box and the text that describes it.
ON

indicates that by default this check box is active. If you use this option, then you
must specify it immediately after the CHECKBOX keyword.
SUBSTITUTE=text-for-substitution

species the text that is to be inserted into the command in the DIALOG statement.

Check Boxes in a Dialog Box

Each CHECKBOX statement denes a single item that the user can select
independent of other selections. That is, if you dene ve choices with ve CHECKBOX
statements, then the user can select any combination of these choices. When the user
selects choices, the text-for-selection values that are associated with the selections are
inserted into the command string of the previous DIALOG statement at eld locations
prexed by an ampersand (&).

DIALOG Statement
Describes a dialog box that is associated with an item on a pull-down menu.
Restriction: Must be followed by at least one TEXT statement.
Featured in: Example 2 on page 685, Example 3 on page 688, and Example 4 on page 694

The PMENU Procedure

DIALOG Statement

669

DIALOG dialog-box command-string

eld-number-specication;

Required Arguments
command-string

is the command or partial command that is executed when the item is selected. The
limit of the command-string that results after the substitutions are made is the
command-line limit for your operating environment. Typically, the command-line
limit is approximately 80 characters.
The limit for command-string eld-number-specication is 200 characters.
Note: If you are using PROC PMENU to submit any command that is valid only
in the PROGRAM EDITOR window (such as the INCLUDE command), then you
must have the windowing environment running, and you must return control to the
PROGRAM EDITOR window. 4
dialog-box

is the same name specied for the DIALOG= option in a previous ITEM statement.
eld-number-specication

can be one or more of the following:

@1@n
%1%n
&1&n
You can embed the eld numbers, for example @1, %1, or &1, in the command
string and mix different types of eld numbers within a command string. The
numeric portion of the eld number corresponds to the relative position of TEXT,
RADIOBOX, and CHECKBOX statements, not to any actual number in these
statements.
@1@n
are optional TEXT statement numbers that can add information to the command
before it is submitted. Numbers preceded by an at sign (@) correspond to TEXT
statements that use the LEN= option to dene input elds.
%1%n
are optional RADIOBOX statement numbers that can add information to the
command before it is submitted. Numbers preceded by a percent sign (%)
correspond to RADIOBOX statements following the DIALOG statement.
Note: Keep in mind that the numbers correspond to RADIOBOX statements,
not to RBUTTON statements. 4
&1&n
are optional CHECKBOX statement numbers that can add information to the
command before it is submitted. Numbers preceded by an ampersand (&)
correspond to CHECKBOX statements following the DIALOG statement.

Note: To specify a literal @ (at sign), % (percent sign), or & (ampersand) in the
command-string, use a double character: @@ (at signs), %% (percent signs), or &&
(ampersands). 4

670

ITEM Statement

Chapter 33

Details
3 You cannot control the placement of the dialog box. The dialog box is not
3
3
3
3

scrollable. The size and placement of the dialog box are determined by your
windowing environment.
To use the DIALOG statement, specify an ITEM statement with the DIALOG=
option in the ITEM statement.
The ITEM statement creates an entry in a menu bar or in a pull-down menu, and
the DIALOG= option species which DIALOG statement describes the dialog box.
You can use CHECKBOX, RADIOBOX, and RBUTTON statements to dene the
contents of the dialog box.
Figure 33.2 on page 670 shows a typical dialog box. A dialog box can request
information in three ways:
3 Fill in a eld. Fields that accept text from a user are called text elds.
3 Choose from a list of mutually exclusive choices. A group of selections of this
type is called a radio box, and each individual selection is called a radio
button.
3 Indicate whether you want to select other independent choices. For example,
you could choose to use various options by selecting any or all of the listed
selections. A selection of this type is called a check box.

Figure 33.2

A Typical Dialog Box

Radio button

Select a commodity:

Wheat
Corn
Oats

Select a market:

Farmville
Monticello
Plainview

Radio box

Enter a year from 1950 to 1996:

Text field

Check here for double spacing:

Check box

Cancel

Push button

Dialog boxes have two or more buttons, such as OK and Cancel, automatically
built into the box.* A button causes an action to occur.

ITEM Statement
Identies an item to be listed in a menu bar or in a pull-down menu.
Featured in:

Example 1 on page 682

* The actual names of the buttons vary in different windowing environments.

The PMENU Procedure

ITEM Statement

671

ITEM command < option(s)>< action-options>;

ITEM menu-item < option(s)>< action-options>;

To do this

Use this option

Specify the action for the item

Associate the item with a dialog box

DIALOG=

Associate the item with a pull-down menu

MENU=

Associate the item with a command

SELECTION=

Associate the item with a common submenu

SUBMENU=

Specify help text for an item

HELP=

Dene a key that can be used instead of the

pull-down menu

ACCELERATE=

Indicate that the item is not an active choice in the

window

GRAY

Provide an ID number for an item

ID=

Dene a single character that can select the item

MNEMONIC=

Place a check box or a radio button next to an item

STATE=

Required Arguments
command

a single word that is a valid SAS command for the window in which the menu
appears. Commands that are more than one word, such as WHERE CLEAR, must be
enclosed in single quotation marks. The command appears in uppercase letters on
the menu bar.
If you want to control the case of a SAS command on the menu, then enclose the
command in single quotation marks. The case that you use then appears on the
menu.
menu-item

a word or text string, enclosed in quotation marks, that describes the action that
occurs when the user selects this item. A menu item should not begin with a percent
sign (%).

Options
ACCELERATE=name-of-key

denes a key sequence that can be used instead of selecting an item. When the user
presses the key sequence, it has the same effect as selecting the item from the menu
bar or pull-down menu.
Restriction: The functionality of this option is limited to only a few characters. For

details, see the SAS documentation for your operating environment.

672

ITEM Statement

Chapter 33

Restriction: This option is not available in all operating environments. If you

include this option and it is not available in your operating environment, then the
option is ignored.
action-option

is one of the following:

DIALOG=dialog-box
the name of an associated DIALOG statement, which displays a dialog box when
the user selects this item.
Featured in: Example 3 on page 688

MENU=pull-down-menu
the name of an associated MENU statement, which displays a pull-down menu
when the user selects this item.
Featured in: Example 1 on page 682

SELECTION=selection
the name of an associated SELECTION statement, which submits a command
when the user selects this item.
Featured in: Example 1 on page 682

SUBMENU=submenu
the name of an associated SUBMENU statement, which displays a pmenu entry
when the user selects this item.
Featured in: Example 1 on page 682

If no DIALOG=, MENU=, SELECTION=, or SUBMENU= option is specied, then

the command or menu-item text string is submitted as a command-line command
when the user selects the item.
GRAY

indicates that the item is not an active choice in this window. This option is useful
when you want to dene standard lists of items for many windows, but not all items
are valid in all windows. When this option is set and the user selects the item, no
action occurs.
HELP=help-text

species text that is displayed when the user displays the menu item. For example,
if you use a mouse to pull down a menu, then position the mouse pointer over the
item and the text is displayed.
Restriction: This option is not available in all operating environments. If you

include this option and it is not available in your operating environment, then the
option is ignored.
Tip: The place where the text is displayed is operating environment-specic.
ID=integer

a value that is used as an identier for an item in a pull-down menu. This identier
is used within a SAS/AF application to selectively activate or deactivate items in a
menu or to set the state of an item as a check box or a radio button.
Minimum: 3001
Restriction: Integers from 0 to 3000 are reserved for operating environment and

SAS use.
Restriction: This option is not available in all operating environments. If you

include this option and it is not available in your operating environment, then the
option is ignored.
Tip: ID= is useful with the WINFO function in SAS Component Language.

The PMENU Procedure

Tip:

MENU Statement

673

You can use the same ID for more than one item.

Restriction: This option is not available in all operating environments. If you

include this option and it is not available in your operating environment, then the
option is ignored.

Dening Items on the Menu Bar

You must use ITEM statements to name all the items that appear in a menu bar. You
also use the ITEM statement to name the items that appear in any pull-down menus.
The items that you specify in the ITEM statement can be commands that are issued
when the user selects the item, or they can be descriptions of other actions that are
performed by associated DIA0LOG, MENU, SELECTION, or SUBMENU statements.
All ITEM statements for a menu must be placed immediately after the MENU
statement and before any DIALOG, SELECTION, SUBMENU, or other MENU
statements. In some operating environments, you can insert SEPARATOR statements
between ITEM statements to produce lines separating groups of items in a pull-down
menu. See SEPARATOR Statement on page 677 for more information.
Note: If you specify a menu bar that is too long for the window, then it might be
truncated or wrapped to multiple lines. 4

MENU Statement
Names the catalog entry that stores the menus or denes a pull-down menu.
Featured in: Example 1 on page 682

MENU menu-bar;
MENU pull-down-menu;

Required Arguments
One of the following arguments is required:

674

MENU Statement

Chapter 33

menu-bar

names the catalog entry that stores the menus.

pull-down-menu

names the pull-down menu that appears when the user selects an item in the menu
bar. The value of pull-down-menu must match the pull-down-menu name that is
specied in the MENU= option in a previous ITEM statement.

Dening Pull-Down Menus

When used to dene a pull-down menu, the MENU statement must follow an ITEM
statement that species the MENU= option. Both the ITEM statement and the MENU
statement for the pull-down menu must be in the same RUN group as the MENU
statement that denes the menu bar for the PMENU catalog entry.
For both menu bars and pull-down menus, follow the MENU statement with ITEM
statements that dene each of the items that appear on the menu. Group all ITEM
statements for a menu together. For example, the following PROC PMENU step creates
one catalog entry, WINDOWS, which produces a menu bar with two items, Primary
windows and Other windows. When you select one of these items, a pull-down menu is
displayed.
libname proclib SAS-data-library;
proc pmenu cat=proclib.mycat;
/* create catalog entry */
menu windows;
item Primary windows menu=prime;
item Other windows menu=other;
/* create first pull-down menu */
menu prime;
item output;
item manager;
item log;
item pgm;
/* create second pull-down menu */
menu other;
item keys;
item help;
item pmenu;
item bye;
/* end of run group */
run;

The following gure shows the resulting menu selections.

The PMENU Procedure

Figure 33.3

RBUTTON Statement

Pull-Down Menu

Primary windows

Other windows

OUTPUT
MANAGER
LOG
PGM

KEYS
HELP
PMENU
BYE

RADIOBOX Statement
Denes a box that contains mutually exclusive choices within a dialog box.
Restriction:

Must be used after a DIALOG statement.

Must be followed by one or more RBUTTON statements.

Featured in: Example 3 on page 688
Restriction:

RADIOBOX DEFAULT=button-number;

Required Arguments
DEFAULT=button-number

indicates which radio button is the default.

Default: 1

Details
The RADIOBOX statement indicates the beginning of a list of selections.
Immediately after the RADIOBOX statement, you must list an RBUTTON statement
for each of the selections the user can make. When the user makes a choice, the text
value that is associated with the selection is inserted into the command string of the
previous DIALOG statement at eld locations prexed by a percent sign (%).

RBUTTON Statement
Lists mutually exclusive choices within a dialog box.
Restriction:

Must be used after a RADIOBOX statement.

Featured in: Example 3 on page 688

RBUTTON <NONE> #line @column

text-for-selection <COLOR=color> < SUBSTITUTE=text-for-substitution>;

675

676

SELECTION Statement

Chapter 33

Required Arguments
column

species the column in the dialog box where the radio button and text are placed.
line

species the line in the dialog box where the radio button and text are placed.
text-for-selection

denes the text that appears in the dialog box and, if the SUBSTITUTE= option is
not used, denes the text that is inserted into the command in the preceding
DIALOG statement.
Note: Be careful not to overlap columns and lines when placing text and radio
buttons; if you overlap text and buttons, you will get an error message. Also, specify
space between other text and a radio button. 4

Options
COLOR=color

denes the color of the radio button and the text that describes the button.
Restriction: This option is not available in all operating environments. If you

include this option and it is not available in your operating environment, then the
option is ignored.
NONE

denes a button that indicates none of the other choices. Dening this button
enables the user to ignore any of the other choices. No characters, including blanks,
are inserted into the DIALOG statement.
Restriction: If you use this option, then it must appear immediately after the

RBUTTON keyword.
SUBSTITUTE=text-for-substitution

species the text that is to be inserted into the command in the DIALOG statement.
Featured in:

Example 3 on page 688

SELECTION Statement
Denes a command that is submitted when an item is selected.
Restriction:

Must be used after an ITEM statement

Featured in:

Example 1 on page 682 and Example 4 on page 694

SELECTION selection command-string;

The PMENU Procedure

SUBMENU Statement

677

Required Arguments
selection

is the same name specied for the SELECTION= option in a previous ITEM
statement.
command-string

is a text string, enclosed in quotation marks, that is submitted as a command-line

command when the user selects this item. There is a limit of 200 characters for
command-string. However, the command-line limit of approximately 80 characters
cannot be exceeded. The command-line limit differs slightly for various operating
environments.

Details
You dene the name of the item in the ITEM statement and specify the
SELECTION= option to associate the item with a subsequent SELECTION statement.
The SELECTION statement then denes the actual command that is submitted when
the user chooses the item in the menu bar or pull-down menu.
You are likely to use the SELECTION statement to dene a command string. You
create a simple alias by using the ITEM statement, which invokes a longer command
string that is dened in the SELECTION statement. For example, you could include an
item in the menu bar that invokes a WINDOW statement to enable data entry. The
actual commands that are processed when the user selects this item are the commands
to include and submit the application.
Note: If you are using PROC PMENU to issue any command that is valid only in
the PROGRAM EDITOR window (such as the INCLUDE command), then you must
have the windowing environment running, and you must return control to the
PROGRAM EDITOR window. 4

SEPARATOR Statement
Draws a line between items on a pull-down menu.
Restriction:
Restriction:

Must be used after an ITEM statement.

Not available in all operating environments.

SEPARATOR;

SUBMENU Statement
Species the SAS le that contains a common submenu associated with an item.
Featured in: Example 1 on page 682

SUBMENU submenu-name SAS-le;

678

TEXT Statement

Chapter 33

Required Arguments
submenu-name

species a name for the submenu statement. To associate a submenu with a menu
item, submenu-name must match the submenu name specied in the SUBMENU=
action-option in the ITEM statement.
SAS-le

species the name of the SAS le that contains the common submenu.

TEXT Statement
Species text and the input elds for a dialog box.
Restriction:

Can be used only after a DIALOG statement.

Featured in:

Example 2 on page 685

TEXT #line @column eld-description

<ATTR=attribute> <COLOR=color>;

Required Arguments
column

species the starting column for the text or input eld.

eld-description

denes how the TEXT statement is used. The eld-description can be one of the
following:
LEN=eld-length
is the length of an input eld in which the user can enter information. If the
LEN= argument is used, then the information entered in the eld is inserted into
the command string of the previous DIALOG statement at eld locations prexed
by an at sign (@).
Featured in: Example 2 on page 685

text
is the text string that appears inside the dialog box at the location dened by line
and column.
line

species the line number for the text or input eld.

The PMENU Procedure

Procedure Execution

679

Options
ATTR=attribute

denes the attribute for the text or input eld. Valid attribute values are
3 BLINK
3 HIGHLIGH
3 REV_VIDE
3 UNDERLIN
Restriction: This option is not available in all operating environments. If you

include this option and it is not available in your operating environment, then the
option is ignored.
Restriction: Your hardware may not support all of these attributes.
COLOR=color

denes the color for the text or input eld characters. These are the color values that
you can use:
BLACK

BROWN

GRAY

MAGENTA

PINK

WHITE

BLUE

CYAN

GREEN

ORANGE

RED

YELLOW

Restriction: This option is not available in all operating environments. If you

include this option and it is not available in your operating environment, then the
option is ignored.
Restriction: Your hardware may not support all of these colors.

Concepts: PMENU Procedure

Procedure Execution

Initiating the Procedure

You can dene multiple menus by separating their denitions with RUN statements.
A group of statements that ends with a RUN statement is called a RUN group. You
must completely dene a PMENU catalog entry before submitting a RUN statement.
You do not have to restart the procedure after a RUN statement.
You must include an initial MENU statement that denes the menu bar, and you
must include all ITEM statements and any SELECTION, MENU, SUBMENU, and

680

Steps for Building and Using PMENU Catalog Entries

Chapter 33

DIALOG statements as well as statements that are associated with the DIALOG
statement within the same RUN group. For example, the following statements dene
two separate PMENU catalog entries. Both are stored in the same catalog, but each
PMENU catalog entry is independent of the other. In the example, both PMENU
catalog entries create menu bars that simply list windowing environment commands
the user can select and execute:
libname proclib SAS-data-library;
proc pmenu catalog=proclib.mycat;
menu menu1;
item end;
item bye;
run;
menu
item
item
item
item
run;

menu2;
end;
pgm;
log;
output;

When you submit these statements, you receive a message that says that the
PMENU entries have been created. To display one of these menu bars, you must
associate the PMENU catalog entry with a window and then activate the window with
the menus turned on, as described in Steps for Building and Using PMENU Catalog
Entries on page 680.

Ending the Procedure

Submit a QUIT, DATA, or new PROC statement to execute any statements that have
not executed and end the PMENU procedure. Submit a RUN CANCEL statement to
cancel any statements that have not executed and end the PMENU procedure.

Steps for Building and Using PMENU Catalog Entries

In most cases, building and using PMENU entries requires the following steps:
1 Use PROC PMENU to dene the menu bars, pull-down menus and other features
that you want. Store the output of PROC PMENU in a SAS catalog.
2 Dene a window using SAS/AF and SAS/FSP software, or the WINDOW or
%WINDOW statement in base SAS software.
3 Associate the PMENU catalog entry created in step 1 with a window by using one
of the following:
3 the MENU= option in the WINDOW statement in base SAS software. See
Associating a Menu with a Window on page 697.
3 the MENU= option in the %WINDOW statement in the macro facility.
3 the Command Menu eld in the GATTR window in PROGRAM entries in SAS/
AF software.
3 the Keys, Pmenu, and Commands window in a FRAME entry in SAS/AF
software. See Example 5 on page 700.
3 the PMENU function in SAS/AF and SAS/FSP software.
3 the SETPMENU command in SAS/FSP software. See Example 1 on page 682.
4 Activate the window you have created. Make sure that the menus are turned on.

The PMENU Procedure

Templates for Coding PROC PMENU Steps

681

Templates for Coding PROC PMENU Steps

The following coding templates summarize how to use the statements in the PMENU
procedure. Refer to descriptions of the statements for more information:

3 Build a simple menu bar. All items on the menu bar are windowing environment
commands:
proc pmenu;
menu menu-bar;
item command;
...more-ITEM-statements...
run;

3 Create a menu bar with an item that produces a pull-down menu:

proc pmenu;
menu menu-bar;
item menu-item menu=pull-down-menu;
...more-ITEM-statements...
menu pull-down-menu;
...ITEM-statements-for-pull-down-menu...
run;

3 Create a menu bar with an item that submits a command other than that which
appears on the menu bar:
proc pmenu;
menu menu-bar;
item menu-item selection=selection;
...more-ITEM-statements...
selection selection command-string;
run;

3 Create a menu bar with an item that opens a dialog box, which displays
information and requests text input:
proc pmenu;
menu menu-bar;
item menu-item menu=pull-down-menu;
...more-ITEM-statements...
menu pull-down-menu;
item menu-item dialog=dialog-box;
dialog dialog-box command @1;
text #line @column text;
text #line @column LEN=field-length;
run;

3 Create a menu bar with an item that opens a dialog box, which permits one choice
from a list of possible values:
proc pmenu;
menu menu-bar;
item menu-item menu=pull-down-menu;
...more-ITEM-statements...
menu pull-down-menu;
item menu-item dialog=dialog-box;
dialog dialog-box command %1;
text #line @column text;

682

Examples: PMENU Procedure

Chapter 33

radiobox default=button-number;
rbutton #line @column
text-for-selection;
...more-RBUTTON-statements...
run;

3 Create a menu bar with an item that opens a dialog box, which permits several
independent choices:
proc pmenu;
menu menu-bar;
item menu-item menu=pull-down-menu;
...more-ITEM-statements...
menu pull-down-menu;
item menu-item dialog=dialog-box;
dialog dialog-box command &1;
text #line @column text;
checkbox #line @column text;
...more-CHECKBOX-statements...
run;

Examples: PMENU Procedure

The windows in these examples were produced in the UNIX environment and may
appear slightly different from the same windows in other operating environments.
You should know the operating environment-specic system options that can affect
how menus are displayed and merged with existing SAS menus. For details, see the
SAS documentation for your operating environment.

Example 1: Building a Menu Bar for an FSEDIT Application

Procedure features:

PROC PMENU statement option:

CATALOG=
ITEM statement options:
MENU=
SELECTION=
SUBMENU=
MENU statement
SELECTION statement
SUBMENU statement

This example creates a menu bar that can be used in an FSEDIT application to
replace the default menu bar. The selections available on these pull-down menus do not
enable end users to delete or duplicate observations.

The PMENU Procedure

Program

683

Program

Declare the PROCLIB library. The PROCLIB library is used to store menu denitions.
libname proclib SAS-data-library;

Specify the catalog for storing menu denitions. Menu denitions will be stored in the
PROCLIB.MENUCAT catalog.
proc pmenu catalog=proclib.menucat;

Specify the name of the catalog entry. The MENU statement species PROJECT as the
name of the catalog entry. The menus are stored in the catalog entry
PROCLIB.MENUCAT.PROJECT.PMENU.
menu project;

Design the menu bar. The ITEM statements specify the items for the menu bar. The value of
the MENU= option is used in a subsequent MENU statement. The Edit item uses a common
predened submenu; the menus for the other items are dened in this PROC step.
item
item
item
item

File menu=f;
Edit submenu=editmnu;
Scroll menu=s;
Help menu=h;

Design the File menu. This group of statements denes the selections available under File
on the menu bar. The rst ITEM statement species Goback as the rst selection under File.
The value of the SELECTION= option corresponds to the subsequent SELECTION statement,
which species END as the command that is issued for that selection. The second ITEM
statement species that the SAVE command is issued for that selection.
menu f;
item Goback selection=g;
item Save;
selection g end;

Add the EDITMNU submenu. The SUBMENU statement associates a predened submenu
that is located in the SAS le SASHELP.CORE.EDIT with the Edit item on the menu bar. The
name of this SUBMENU statement is EDITMNU, which corresponds with the name in the
SUBMENU= action-option in the ITEM statement for the Edit item.
submenu editmnu sashelp.core.edit;

684

Associating a Menu Bar with an FSEDIT Session

Chapter 33

Design the Scroll menu. This group of statements denes the selections available under
Scroll on the menu bar.
menu s;
item Next Obs selection=n;
item Prev Obs selection=p;
item Top;
item Bottom;
selection n forward;
selection p backward;

Design the Help menu. This group of statements denes the selections available under Help
on the menu bar. The SETHELP command species a HELP entry that contains user-written
information for this FSEDIT application. The semicolon that appears after the HELP entry
name enables the HELP command to be included in the string. The HELP command invokes
the HELP entry.
menu h;
item Keys;
item About this application selection=hlp;
selection hlp sethelp user.menucat.staffhlp.help;help;
quit;

Associating a Menu Bar with an FSEDIT Session

The following SETPMENU command associates the customized menu bar with the
FSEDIT window.
setpmenu proclib.menucat.project.pmenu;pmenu on

You can also specify the menu bar on the command line in the FSEDIT session or by
issuing a CALL EXECCMD command in SAS Component Language (SCL).
See Associating a Menu Bar with an FSEDIT Session on page 691 for other
methods of associating the customized menu bar with the FSEDIT window.
The FSEDIT window shows the menu bar.

The PMENU Procedure

Program

685

Example 2: Collecting User Input in a Dialog Box

Procedure features:

DIALOG statement
TEXT statement option:
LEN=

This example adds a dialog box to the menus created in Example 1 on page 682. The
dialog box enables the user to use a WHERE clause to subset the SAS data set.
Tasks include

3 collecting user input in a dialog box

3 creating customized menus for an FSEDIT application.

Program

Declare the PROCLIB library. The PROCLIB library is used to store menu denitions.
libname proclib SAS-data-library;

Specify the catalog for storing menu denitions. Menu denitions will be stored in the
PROCLIB.MENUCAT catalog.
proc pmenu catalog=proclib.menucat;

Specify the name of the catalog entry. The MENU statement species PROJECT as the
name of the catalog entry. The menus are stored in the catalog entry
PROCLIB.MENUCAT.PROJECT.PMENU.
menu project;

Design the menu bar. The ITEM statements specify the items for the menu bar. The value of
the MENU= option is used in a subsequent MENU statement.
item
item
item
item
item

File menu=f;
Edit menu=e;
Scroll menu=s;
Subset menu=sub;
Help menu=h;

686

Program

Chapter 33

Design the File menu. This group of statements denes the selections under File on the
menu bar. The rst ITEM statement species Goback as the rst selection under File. The
value of the SELECTION= option corresponds to the subsequent SELECTION statement, which
species END as the command that is issued for that selection. The second ITEM statement
species that the SAVE command is issued for that selection.
menu f;
item Goback selection=g;
item Save;
selection g end;

Design the Edit menu. This group of statements denes the selections available under Edit
on the menu bar.
menu e;
item Cancel;
item Add;

Design the Subset menu. This group of statements denes the selections available under
Subset on the menu bar. The value d1 in the DIALOG= option is used in the subsequent
DIALOG statement.
menu sub;
item Where dialog=d1;
item Where Clear;

Design the Help menu. This group of statements denes the selections available under Help
on the menu bar. The SETHELP command species a HELP entry that contains user-written
information for this FSEDIT application. The semicolon enables the HELP command to be
included in the string. The HELP command invokes the HELP entry.
menu h;
item Keys;
item About this application selection=hlp;
selection hlp sethelp proclib.menucat.staffhlp.help;help;

The PMENU Procedure

Associating a Menu Bar with an FSEDIT Window

687

Design the dialog box. The DIALOG statement builds a WHERE command. The arguments
for the WHERE command are provided by user input into the text entry elds described by the
three TEXT statements. The @1 notation is a placeholder for user input in the text eld. The
TEXT statements specify the text in the dialog box and the length of the input eld.
dialog d1 where @1;
text #2 @3 Enter a valid WHERE clause or UNDO;
text #4 @3 WHERE ;
text #4 @10 len=40;
quit;

Associating a Menu Bar with an FSEDIT Window

The following SETPMENU command associates the customized menu bar with the
FSEDIT window.
setpmenu proclib.menucat.project.pmenu;pmenu on

You can also specify the menu bar on the command line in the FSEDIT session or by
issuing a CALL EXECCMD command in SAS Component Language (SCL). Refer to
SAS Component Language: Reference for complete documentation on SCL.
See Associating a Menu Bar with an FSEDIT Session on page 691 for other
methods of associating the customized menu bar with the FSEDIT window.
This dialog box appears when the user chooses Subset and then Where.

688

Example 3: Creating a Dialog Box to Search Multiple Variables

Chapter 33

Example 3: Creating a Dialog Box to Search Multiple Variables

Procedure features:

DIALOG statement
SAS macro invocation
ITEM statement
DIALOG= option
RADIOBOX statement option:
DEFAULT=
RBUTTON statement option:
SUBSTITUTE=
Other features: SAS macro invocation

This example shows how to modify the menu bar in an FSEDIT session to enable a
search for one value across multiple variables. The example creates customized menus
to use in an FSEDIT session. The menu structure is the same as in the preceding
example, except for the WHERE dialog box.
When selected, the menu item invokes a macro. The user input becomes values for
macro parameters. The macro generates a WHERE command that expands to include
all the variables needed for the search.
Tasks include

3 associating customized menus with an FSEDIT session

3 searching multiple variables with a WHERE clause
3 extending PROC PMENU functionality with a SAS macro.

Program

Declare the PROCLIB library. The PROCLIB library is used to store menu denitions.
libname proclib SAS-data-library;

Specify the catalog for storing menu denitions. Menu denitions will be stored in the
PROCLIB.MENUCAT catalog.
proc pmenu catalog=proclib.menucat;

Specify the name of the catalog entry. The MENU statement species STAFF as the name
of the catalog entry. The menus are stored in the catalog entry
PROCLIB.MENUCAT.PROJECT.PMENU.
menu project;

The PMENU Procedure

Program

689

Design the menu bar. The ITEM statements specify the items for the menu bar. The value of
the MENU= option is used in a subsequent MENU statement.
item
item
item
item
item

File menu=f;
Edit menu=e;
Scroll menu=s;
Subset menu=sub;
Help menu=h;

Design the Edit menu. The ITEM statements dene the selections under Edit on the menu
bar.
menu e;
item Cancel;
item Add;

Design the Scroll menu. This group of statements denes the selections under Scroll on the
menu bar. If the quoted string in the ITEM statement is not a valid command, then the
SELECTION= option corresponds to a subsequent SELECTION statement, which species a
valid command.
menu s;
item Next Obs selection=n;
item Prev Obs selection=p;
item Top;
item Bottom;
selection n forward;
selection p backward;

Design the Subset menu. This group of statements denes the selections under Subset on
the menu bar. The DIALOG= option names a dialog box that is dened in a subsequent
DIALOG statement.
menu sub;
item Where dialog=d1;
item Where Clear;

690

Program

Chapter 33

Design the Help menu. This group of statements denes the selections under Help on the
menu bar. The SETHELP command species a HELP entry that contains user-written
information for this FSEDIT application. The semicolon that appears after the HELP entry
name enables the HELP command to be included in the string. The HELP command invokes
the HELP entry.
menu h;
item Keys;
item About this application selection=hlp;
selection hlp sethelp proclib.menucat.staffhlp.help;help;

Design the dialog box. WBUILD is a SAS macro. The double percent sign that precedes
WBUILD is necessary to prevent PROC PMENU from expecting a eld number to follow. The
eld numbers %1, %2, and %3 equate to the values that the user specied with the radio boxes.
The eld number @1 equates to the search value that the user enters. See How the WBUILD
Macro Works on page 693.
dialog d1 %%wbuild(%1,%2,@1,%3);

Add a radio box for region selection. The TEXT statement species text for the dialog box
that appears on line 1 and begins in column 1. The RADIOBOX statement species that a radio
box will appear in the dialog box. DEFAULT= species that the rst radio button (Northeast)
will be selected by default. The RBUTTON statements specify the mutually exclusive choices for
the radio buttons: Northeast, Northwest, Southeast, or Southwest. SUBSTITUTE= gives
the value that is substituted for the %1 in the DIALOG statement above if that radio button is
selected.
text #1 @1 Choose a region:;
radiobox default=1;
rbutton #3 @5 Northeast substitute=NE;
rbutton #4 @5 Northwest substitute=NW;
rbutton #5 @5 Southeast substitute=SE;
rbutton #6 @5 Southwest substitute=SW;

Add a radio box for pollutant selection. The TEXT statement species text for the dialog
box that appears on line 8 (#8) and begins in column 1 (@1). The RADIOBOX statement
species that a radio box will appear in the dialog box. DEFAULT= species that the rst radio
button (Pollutant A) will be selected by default. The RBUTTON statements specify the
mutually exclusive choices for the radio buttons: Pollutant A or Pollutant B.
SUBSTITUTE= gives the value that is substituted for the %2 in the preceding DIALOG
statement if that radio button is selected.
text #8 @1 Choose a contaminant:;
radiobox default=1;
rbutton #10 @5 Pollutant A substitute=pol_a,2;
rbutton #11 @5 Pollutant B substitute=pol_b,4;

The PMENU Procedure

Associating a Menu Bar with an FSEDIT Session

691

Add an input eld. The rst TEXT statement species text for the dialog box that appears on
line 13 and begins in column 1. The second TEXT statement species an input eld that is 6
bytes long that appears on line 13 and begins in column 25. The value that the user enters in
the eld is substituted for the @1 in the preceding DIALOG statement.
text #13 @1 Enter Value for Search:;
text #13 @25 len=6;

Add a radio box for comparison operator selection. The TEXT statement species text for
the dialog box that appears on line 15 and begins in column 1. The RADIOBOX statement
species that a radio box will appear in the dialog box. DEFAULT= species that the rst radio
button (Greater Than or Equal To) will be selected by default. The RBUTTON statements
specify the mutually exclusive choices for the radio buttons. SUBSTITUTE= gives the value that
is substituted for the %3 in the preceding DIALOG statement if that radio button is selected.
text #15 @1 Choose a comparison criterion:;
radiobox default=1;
rbutton #16 @5 Greater Than or Equal To
substitute=GE;
rbutton #17 @5 Less Than or Equal To
substitute=LE;
rbutton #18 @5 Equal To substitute=EQ;
quit;

This dialog box appears when the user selects Subset and then Where.

Associating a Menu Bar with an FSEDIT Session

The SAS data set PROCLIB.LAKES has data about several lakes. Two pollutants,
pollutant A and pollutant B, were tested at each lake. Tests were conducted for

692

Associating a Menu Bar with an FSEDIT Session

Chapter 33

pollutant A twice at each lake, and the results are recorded in the variables POL_A1
and POL_A2. Tests were conducted for pollutant B four times at each lake, and the
results are recorded in the variables POL_B1 - POL_B4. Each lake is located in one of
four regions. The following output lists the contents of PROCLIB.LAKES:

Output 33.1
PROCLIB.LAKES
region
NE
NE
NE
NE
NW
NW
NW
NW
SE
SE
SE
SE
SW
SW
SW
SW

lake
Carr
Duraleigh
Charlie
Farmer
Canyon
Morris
Golf
Falls
Pleasant
Juliette
Massey
Delta
Alumni
New Dam
Border
Red

pol_a1

pol_a2

pol_b1

pol_b2

pol_b3

pol_b4

0.24
0.34
0.40
0.60
0.63
0.85
0.69
0.01
0.16
0.82
1.01
0.84
0.45
0.80
0.51
0.22

0.99
0.01
0.48
0.65
0.44
0.95
0.37
0.02
0.96
0.35
0.77
1.05
0.32
0.70
0.04
0.09

0.95
0.48
0.29
0.25
0.20
0.80
0.08
0.59
0.71
0.09
0.45
0.90
0.45
0.31
0.55
0.02

0.36
0.58
0.56
0.20
0.98
0.67
0.72
0.58
0.35
0.03
0.32
0.09
0.44
0.98
0.35
0.10

0.44
0.12
0.52
0.30
0.19
0.32
0.71
0.67
0.35
0.59
0.55
0.64
0.55
1.00
0.45
0.32

0.67
0.56
0.95
0.64
0.01
0.81
0.32
0.02
0.48
0.90
0.66
0.03
0.12
0.22
0.78
0.01

A DATA step on page 1393 creates PROCLIB.LAKES.

The following statements initiate a PROC FSEDIT session for PROCLIB.LAKES:
proc fsedit data=proclib.lakes screen=proclib.lakes;
run;

To associate the customized menu bar menu with the FSEDIT session, do any one of
the following:
3 enter a SETPMENU command on the command line. The command for this
example is
setpmenu proclib.menucat.project.pmenu

Turn on the menus by entering PMENU ON on the command line.

3 enter the SETPMENU command in a Command window.

3 include an SCL program with the FSEDIT session that uses the customized menus
and turns on the menus, for example:
fseinit:
call execcmd(setpmenu proclib.menucat.project.pmenu;
pmenu on;);
return;
init:
return;
main:
return;
term:
return;

The PMENU Procedure

How the WBUILD Macro Works

693

How the WBUILD Macro Works

Consider how you would learn whether any of the lakes in the Southwest region
tested for a value of .50 or greater for pollutant A. Without the customized menu item,
you would issue the following WHERE command in the FSEDIT window:
where region="SW" and (pol_a1 ge .50 or pol_a2 ge .50);

Using the custom menu item, you would select Southwest, Pollutant A, enter .50
as the value, and choose Greater Than or Equal To as the comparison criterion. Two
lakes, New Dam and Border, meet the criteria.
The WBUILD macro uses the four pieces of information from the dialog box to
generate a WHERE command:

3 One of the values for region, either NE, NW, SE, or SW, becomes the value of the
macro parameter REGION.

3 Either pol_a,2 or pol_b,4 become the values of the PREFIX and NUMVAR
macro parameters. The comma is part of the value that is passed to the WBUILD
macro and serves to delimit the two parameters, PREFIX and NUMVAR.

3 The value that the user enters for the search becomes the value of the macro
parameter VALUE.

3 The operator that the user chooses becomes the value of the macro parameter
OPERATOR.
To see how the macro works, again consider the following example, in which you
want to know if any of the lakes in the southwest tested for a value of .50 or greater for
pollutant A. The values of the macro parameters would be
REGION

PREFIX

pol_a

NUMVAR

VALUE

.50

OPERATOR

The rst %IF statement checks to make sure that the user entered a value. If a
value has been entered, then the macro begins to generate the WHERE command.
First, the macro creates the beginning of the WHERE command:
where region="SW" and (

Next, the %DO loop executes. For pollutant A, it executes twice because
NUMVAR=2. In the macro denition, the period in &prefix.&i concatenates pol_a
with 1 and with 2. At each iteration of the loop, the macro resolves PREFIX,
OPERATOR, and VALUE, and it generates a part of the WHERE command. On the
rst iteration, it generates pol_a1 GE .50
The %IF statement in the loop checks to see if the loop is working on its last
iteration. If it is not working, then the macro makes a compound WHERE command by
putting an OR between the individual clauses. The next part of the WHERE command
becomes OR pol_a2 GE .50
The loop ends after two executions for pollutant A, and the macro generates the end
of the WHERE command:
)

694

Example 4: Creating Menus for a DATA Step Window Application

Chapter 33

Results from the macro are placed on the command line. The following code is the
denition of the WBUILD macro. The underlined code shows the parts of the WHERE
command that are text strings that the macro does not resolve:
%macro wbuild(region,prefix,numvar,value,operator);
/* check to see if value is present */
%if &value ne %then %do;
where region="&region" AND (
/* If the values are character,
*/
/* enclose &value in double quotation marks. */
%do i=1 %to &numvar;
&prefix.&i &operator &value
/* if not on last variable, */
/* generate OR
*/
%if &i ne &numvar %then %do;
OR
%end;
%end;
)
%end;
%mend wbuild;

Example 4: Creating Menus for a DATA Step Window Application

Procedure features:

DIALOG statement
SELECTION statement
Other features: FILENAME statement

This example denes an application that enables the user to enter human resources
data for various departments and to request reports from the data sets that are created
by the data entry.
The rst part of the example describes the PROC PMENU step that creates the
menus. The subsequent sections describe how to use the menus in a DATA step window
application.
Tasks include
3 associating customized menus with a DATA step window
3 creating menus for a DATA step window
3 submitting SAS code from a menu selection
3 creating a pull-down menu selection that calls a dialog box.

Program

Declare the PROCLIB library. The PROCLIB library is used to store menu denitions.
libname proclib SAS-data-library;

The PMENU Procedure

Program

695

Declare the DE and PRT lenames. The FILENAME statements dene the external les in
which the programs to create the windows are stored.
filename de
filename prt

external-file;
external-file;

Specify the catalog for storing menu denitions. Menu denitions will be stored in the
PROCLIB.MENUCAT catalog.
proc pmenu catalog=proclib.menus;

Specify the name of the catalog entry. The MENU statement species SELECT as the name
of the catalog entry. The menus are stored in the catalog entry
PROCLIB.MENUS.SELECT.PMENU.
menu select;

Design the menu bar. The ITEM statements specify the three items on the menu bar. The
value of the MENU= option is used in a subsequent MENU statement.
item File menu=f;
item Data_Entry menu=deptsde;
item Print_Report menu=deptsprt;

Design the File menu. This group of statements denes the selections under File. The value
of the SELECTION= option is used in a subsequent SELECTION statement.
menu f;
item End
item End
selection
selection

this window selection=endwdw;

this SAS session selection=endsas;
endwdw end;
endsas bye;

Design the Data_Entry menu. This group of statements denes the selections under
Data_Entry on the menu bar. The ITEM statements specify that For Dept01 and For
Dept02 appear under Data_Entry. The value of the SELECTION= option equates to a
subsequent SELECTION statement, which contains the string of commands that are actually
submitted. The value of the DIALOG= option equates to a subsequent DIALOG statement,
which describes the dialog box that appears when this item is selected.
menu deptsde;
item For Dept01 selection=de1;
item For Dept02 selection=de2;
item Other Departments dialog=deother;

696

Program

Chapter 33

Specify commands under the Data_Entry menu. The commands in single quotation marks
are submitted when the user selects For Dept01 or For Dept02. The END command ends the
current window and returns to the PROGRAM EDITOR window so that further commands can
be submitted. The INCLUDE command includes the SAS statements that create the data entry
window. The CHANGE command modies the DATA statement in the included program so that
it creates the correct data set. (See Using a Data Entry Program on page 698.) The SUBMIT
command submits the DATA step program.
selection de1 end;pgm;include de;change xx 01;submit;
selection de2 end;pgm;include de;change xx 02;submit;

Design the DEOTHER dialog box. The DIALOG statement denes the dialog box that
appears when the user selects Other Departments. The DIALOG statement modies the
command string so that the name of the department that is entered by the user is used to
change deptxx in the SAS program that is included. (See Using a Data Entry Program on
page 698.) The rst two TEXT statements specify text that appears in the dialog box. The third
TEXT statement species an input eld. The name that is entered in this eld is substituted for
the @1 in the DIALOG statement.
dialog deother end;pgm;include de;c deptxx @1;submit;
text #1 @1 Enter department name;
text #2 @3 in the form DEPT99:;
text #2 @25 len=7;

Design the Print_Report menu. This group of statements denes the choices under the
Print_Report item. These ITEM statements specify that For Dept01 and For Dept02
appear in the pull-down menu. The value of the SELECTION= option equates to a subsequent
SELECTION statement, which contains the string of commands that are actually submitted.
menu deptsprt;
item For Dept01 selection=prt1;
item For Dept02 selection=prt2;
item Other Departments dialog=prother;

Specify commands for the Print_Report menu. The commands in single quotation marks
are submitted when the user selects For Dept01 or For Dept02. The END command ends the
current window and returns to the PROGRAM EDITOR window so that further commands can
be submitted. The INCLUDE command includes the SAS statements that print the report. (See
Printing a Program on page 699.) The CHANGE command modies the PROC PRINT step in
the included program so that it prints the correct data set. The SUBMIT command submits the
PROC PRINT program.
selection prt1
end;pgm;include prt;change xx 01 all;submit;
selection prt2
end;pgm;include prt;change xx 02 all;submit;

The PMENU Procedure

Associating a Menu with a Window

697

Design the PROTHER dialog box. The DIALOG statement denes the dialog box that
appears when the user selects Other Departments. The DIALOG statement modies the
command string so that the name of the department that is entered by the user is used to
change deptxx in the SAS program that is included. (See Printing a Program on page 699.)
The rst two TEXT statements specify text that appears in the dialog box. The third TEXT
statement species an input eld. The name entered in this eld is substituted for the @1 in the
DIALOG statement.
dialog prother end;pgm;include prt;c deptxx @1 all;submit;
text #1 @1 Enter department name;
text #2 @3 in the form DEPT99:;
text #2 @25 len=7;

End this RUN group.

run;

Specify a second catalog entry and menu bar. The MENU statement species ENTRDATA
as the name of the catalog entry that this RUN group is creating. File is the only item on the
menu bar. The selections available are End this window and End this SAS session.
menu entrdata;
item File menu=f;
menu f;
item End this window selection=endwdw;
item End this SAS session selection=endsas;
selection endwdw end;
selection endsas bye;
run;
quit;

Associating a Menu with a Window

The rst group of statements denes the primary window for the application. These
statements are stored in the le that is referenced by the HRWDW leref:
The WINDOW statement creates the HRSELECT window. MENU= associates the
PROCLIB.MENUS.SELECT.PMENU entry with this window.
data _null_;
window hrselect menu=proclib.menus.select
#4 @10 This application allows you to
#6 @13 - Enter human resources data for
#7 @15 one department at a time.
#9 @13 - Print reports on human resources data for
#10 @15 one department at a time.
#12 @13 - End the application and return to the PGM window.
#14 @13 - Exit from the SAS System.
#19 @10 You must have the menus turned on.;

698

Using a Data Entry Program

Chapter 33

The DISPLAY statement displays the window HRSELECT.

display hrselect;
run;

Primary window, HRSELECT.

Using a Data Entry Program

When the user selects Data_Entry from the menu bar in the HRSELECT window, a
pull-down menu is displayed. When the user selects one of the listed departments or
chooses to enter a different department, the following statements are invoked. These
statements are stored in the le that is referenced by the DE leref.
The WINDOW statement creates the HRDATA window. MENU= associates the
PROCLIB.MENUS.ENTRDATA.PMENU entry with the window.
data proclib.deptxx;
window hrdata menu=proclib.menus.entrdata
#5 @10 Employee Number
#8 @10 Salary
#11 @10 Employee Name
#5 @31 empno $4.
#8 @31 salary 10.
#11 @31 name $30.
#19 @10 Press ENTER to add the observation to the data set.;

The PMENU Procedure

Printing a Program

699

The DISPLAY statement displays the HRDATA window.

display hrdata;
run;

The %INCLUDE statement recalls the statements in the le HRWDW. The statements in
HRWDW redisplay the primary window. See the HRSELECT window on page 698.
filename hrwdw external-file;
%include hrwdw;
run;

The SELECTION and DIALOG statements in the PROC PMENU step modify the
DATA statement in this program so that the correct department name is used when the
data set is created. That is, if the user selects Other Departments and enters DEPT05,
then the DATA statement is changed by the command string in the DIALOG statement
to
data proclib.dept05;

Data entry window, HRDATA.

Printing a Program
When the user selects Print_Report from the menu bar, a pull-down menu is
displayed. When the user selects one of the listed departments or chooses to enter a
different department, the following statements are invoked. These statements are
stored in the external le referenced by the PRT leref.

700

Example 5: Associating Menus with a FRAME Application

Chapter 33

PROC PRINTTO routes the output to an external le.

proc printto file=external-file new;
run;

The xxs are changed to the appropriate department number by the CHANGE command in the
SELECTION or DIALOG statement in the PROC PMENU step. PROC PRINT prints that data
set.
libname proclib SAS-data-library;
proc print data=proclib.deptxx;
title Information for deptxx;
run;

This PROC PRINTTO steps restores the default output destination. See Chapter 35, The
PRINTTO Procedure, on page 771 for documentation on PROC PRINTTO.
proc printto;
run;

The %INCLUDE statement recalls the statements in the le HRWDW. The statements in
HRWDW redisplay the primary window.
filename hrwdw external-file;
%include hrwdw;
run;

Example 5: Associating Menus with a FRAME Application

Procedure features:

ITEM statement
MENU statement
Other features: SAS/AF software

This example creates menus for a FRAME entry and gives the steps necessary to
associate the menus with a FRAME entry from SAS/AF software.

Program

Declare the PROCLIB library. The PROCLIB library is used to store menu denitions.
libname proclib SAS-data-library;

The PMENU Procedure

Program

701

Specify the catalog for storing menu denitions. Menu denitions will be stored in the
PROCLIB.MENUCAT catalog.
proc pmenu catalog=proclib.menucat;

Specify the name of the catalog entry. The MENU statement species FRAME as the name
of the catalog entry. The menus are stored in the catalog entry
PROCLIB.MENUS.FRAME.PMENU.
menu frame;

Design the menu bar. The ITEM statements specify the items in the menu bar. The value of
MENU= corresponds to a subsequent MENU statement.
item File menu=f;
item Help menu=h;

Design the File menu. The MENU statement equates to the MENU= option in a preceding
ITEM statement. The ITEM statements specify the selections that are available under File on
the menu bar.
menu f;
item Cancel;
item End;

Design the Help menu. The MENU statement equates to the MENU= option in a preceding
ITEM statement. The ITEM statements specify the selections that are available under Help on
the menu bar. The value of the SELECTION= option equates to a subsequent SELECTION
statement.
menu h;
item About the application selection=a;
item About the keys
selection=k;

Specify commands for the Help menu. The SETHELP command species a HELP entry
that contains user-written information for this application. The semicolon that appears after the
HELP entry name enables the HELP command to be included in the string. The HELP
command invokes the HELP entry.
selection a sethelp proclib.menucat.app.help;help;
selection k sethelp proclib.menucat.keys.help;help;
run;
quit;

702

Steps to Associate Menus with a FRAME

Chapter 33

Steps to Associate Menus with a FRAME

1 In the BUILD environment for the FRAME entry, from the menu bar, select

View

Properties Window

2 In the Properties window, select the Value eld for the pmenuEntry Attribute

Name. The Select An Entry window opens.

3 In the Select An Entry window, enter the name of the catalog entry that is

specied in the PROC PMENU step that creates the menus.

4 Test the FRAME as follows from the menu bar of the FRAME:

Build

Test

Notice that the menus are now associated with the FRAME.

Refer to Getting Started with the FRAME Entry: Developing Object-Oriented

Applications for more information on SAS programming with FRAME entries.

703

CHAPTER

34
The PRINT Procedure
Overview: PRINT Procedure 703
What Does the PRINT Procedure Do? 703
Simple Listing Report 704
Customized Report 704
Syntax: PRINT Procedure
705
PROC PRINT Statement 707
BY Statement 715
ID Statement 716
PAGEBY Statement 717
SUM Statement 718
SUMBY Statement 719
VAR Statement 719
Results: Print Procedure 720
Procedure Output 720
Page Layout 720
Observations 720
Column Headings 722
Column Width 723
Examples: PRINT Procedure
723
Example 1: Selecting Variables to Print 723
Example 2: Customizing Text in Column Headers 727
Example 3: Creating Separate Sections of a Report for Groups of Observations 731
Example 4: Summing Numeric Variables with One BY Group 737
Example 5: Summing Numeric Variables with Multiple BY Variables 742
Example 6: Limiting the Number of Sums in a Report 748
Example 7: Controlling the Layout of a Report with Many Variables 754
Example 8: Creating a Customized Layout with BY Groups and ID Variables 761
Example 9: Printing All the Data Sets in a SAS Library 767

Overview: PRINT Procedure

What Does the PRINT Procedure Do?
The PRINT procedure prints the observations in a SAS data set, using all or some of
the variables. You can create a variety of reports ranging from a simple listing to a

704

Simple Listing Report

Chapter 34

highly customized report that groups the data and calculates totals and subtotals for
numeric variables.

Simple Listing Report

Output 34.1 illustrates the simplest kind of report that you can produce. The
statements that produce the output follow. Example 1 on page 723 creates the data set
EXPREV.
options nodate pageno=1 linesize=64 pagesize=60;

proc print data=exprev;

run;

Output 34.1

Simple Listing Report Produced with PROC PRINT

The SAS System

Obs

Region

1
2
3
4
5
6
7

Southern
Southern
Southern
Northern
Northern
Southern
Northern

State
GA
GA
FL
NY
NY
FL
MA

Month

Expenses

JAN95
FEB95
FEB95
FEB95
MAR95
MAR95
MAR95

2000
1200
8500
3000
6000
9800
1500

Revenues
8000
6000
11000
4000
5000
13500
1000

Customized Report
The following HTML report is a customized report that is produced by PROC PRINT
using ODS. The statements that create this report

Chapter 34

ID variable(s) <option>;
SUM variable(s) <option>;
VAR variable(s) <option>;

To do this

Use this statement

Produce a separate section of the report for each BY

group

Identify observations by the formatted values of the

variables that you list instead of by observation
numbers

Control page ejects that occur before a page is full

PAGEBY

Limit the number of sums that appear in the report

SUMBY

Total values of numeric variables

SUM

Select variables that appear in the report and

determine their order

VAR

The PRINT Procedure

PROC PRINT Statement

PROC PRINT < option(s)>;

To do this

Use this option

Specify text for the HTML contents link to the output

CONTENTS=

Specify the input data set

DATA=

Control general format

Write a blank line between observations

DOUBLE

Print the number of observations in the data set,

in BY groups, or both, and specify explanatory
text to print with the number

Suppress the column in the output that

identies each observation by number

NOOBS

Specify a column header for the column that

identies each observation by number

OBS=

Round unformatted numeric values to two

decimal places

ROUND

Control page format

Format the rows on a page

ROWS=

Use each variables formatted width as its

column width on all pages

WIDTH=UNIFORM

Control column format

Control the orientation of the column headings

HEADING=

Use variables labels as column headings

LABEL or SPLIT=

Specify the split character, which controls line

breaks in column headings

SPLIT=

Specify one or more style elements for the

Output Delivery System to use for different
parts of the report

STYLE

Determine the column width for each variable

WIDTH=

Options
CONTENTS=link-text

species the text for the links in the HTML contents le to the output produced by
the PROC PRINT statement. For information on HTML output, see SAS Output
Delivery System: Users Guide.
Restriction: CONTENTS= does not affect the HTML body le. It affects only the
HTML contents le.

707

708

PROC PRINT Statement

Chapter 34

DATA=SAS-data-set

species the SAS data set to print.

Main discussion: Input Data Sets on page 19
DOUBLE

writes a blank line between observations.

Alias:

Restriction: This option has no effect on the HTML output.

Featured in:

Example 1 on page 723

HEADING=direction

controls the orientation of the column headings, where direction is one of the
following:
HORIZONTAL
prints all column headings horizontally.
Alias: H

VERTICAL
prints all column headings vertically.
Alias: V
Default: Headings are either all horizontal or all vertical. If you omit HEADING=,

PROC PRINT determines the direction of the column headings as follows:

3 If you do not use LABEL, spacing dictates whether column headings are
vertical or horizontal.

3 If you use LABEL and at least one variable has a label, all headings are
horizontal.
LABEL

uses variables labels as column headings.

Alias:

Default: If you omit LABEL, PROC PRINT uses the variables name as the column

heading even if the PROC PRINT step contains a LABEL statement. If a variable
does not have a label, PROC PRINT uses the variables name as the column
heading.
Interaction: By default, if you specify LABEL and at least one variable has a label,

PROC PRINT prints all column headings horizontally. Therefore, using LABEL
may increase the number of pages of output. (Use HEADING=VERTICAL in the
PROC PRINT statement to print vertical column headings.)
Interaction: PROC PRINT sometimes conserves space by splitting labels across

multiple lines. Use SPLIT= in the PROC PRINT statement to control where these
splits occur. You do not need to use LABEL if you use SPLIT=.
To create a blank column header for a variable, use this LABEL statement in
your PROC PRINT step:

Tip:

label variable-name=00x;

See also: For information on using the LABEL statement to create temporary

labels in procedures see Chapter 3, Statements with the Same Function in

Multiple Procedures, on page 57.
For information on using the LABEL statement in a DATA step to create
permanent labels, see the section on statements in SAS Language Reference:
Dictionary.
Featured in:

Example 3 on page 731

The PRINT Procedure

PROC PRINT Statement

709

Note: The SAS system option LABEL must be in effect in order for any procedure
to use labels. For more information see the section on system options in SAS
Language Reference: Dictionary 4
N<=string-1 <string-2>>

prints the number of observations in the data set, in BY groups, or both and species
explanatory text to print with the number.
If you use the N option

PROC PRINT

with neither a BY nor a SUM statement

prints the number of observations in the data set

at the end of the report and labels the number
with the value of string-1.

with a BY statement

prints the number of observations in the BY group

at the end of each BY group and labels the number
with the value of string-1.

with a BY statement and a SUM statement

prints the number of observations in the BY group

at the end of each BY group and prints the
number of observations in the data set at the end
of the report. The numbers for BY groups are
labeled with string-1; the number for the entire
data set is labeled with string-2.

Example 2 on page 727 (alone)

Example 3 on page 731 (with a BY statement)
Example 4 on page 737 (with a BY statement and a SUM statement)

Featured in:

NOOBS

suppresses the observation number in the output.

Featured in: Example 3 on page 731
OBS=column-header

species a column header for the column that identies each observation by number.
Tip: OBS= honors the split character (see the discussion of SPLIT= on page 710 ).
Featured in: Example 2 on page 727
ROUND

rounds unformatted numeric values to two decimal places. (Formatted values are
already rounded by the format to the specied number of decimal places.) For both
formatted and unformatted variables, PROC PRINT uses these rounded values to
calculate any sums in the report.
If you omit ROUND, PROC PRINT adds the actual values of the rows to obtain
the sum even though it displays the formatted (rounded) values. Any sums are also
rounded by the format, but they include only one rounding error, that of rounding the
sum of the actual values. The ROUND option, on the other hand, rounds values
before summing them, so there may be multiple rounding errors. The results without
ROUND are more accurate, but ROUND is useful for published reports where it is
important for the total to be the sum of the printed (rounded) values.
Be aware that the results from PROC PRINT with the ROUND option may differ
from the results of summing the same data with other methods such as PROC
MEANS or the DATA step. Consider a simple case in which
3 the data set contains three values for X: .003, .004, and .009.
3 X has a format of 5.2.

710

PROC PRINT Statement

Chapter 34

Depending on how you calculate the sum, you can get three different answers:
0.02, 0.01, and 0.016. The following gure shows the results of calculating the sum
with PROC PRINT (without and with the ROUND option) and PROC MEANS.

Figure 34.1

Three Methods of Summing Variables

Actual Values

PROC PRINT without

the ROUND option

PROC PRINT with

the ROUND option

PROC MEANS

===================================================================================
||
||
||
||
| | Analysis Variable : X
||
OBS
X
OBS
X
||
||
||
||
||
||
.003
Sum
1
0.00
1
0.00
||
| | -----------||
.004
2
0.00
2
0.00
0.0160000
||
||
||
.009
3
0.01
3
0.01
-----------=====
=====
=====
||
||
||
.016
0.02
0.01
||
||
||
||
||
||
===================================================================================

Notice that the sum produced without the ROUND option (.02) is closer to the
actual result (0.16) than the sum produced with ROUND (0.01). However, the sum
produced with ROUND reects the numbers displayed in the report.
Alias:

CAUTION:

Do not use ROUND with PICTURE formats. ROUND is for use with numeric values.
SAS procedures treat variables that have picture formats as character variables.
Using ROUND with such variables may lead to unexpected results. 4
ROWS=page-format

formats rows on a page. Currently, PAGE is the only value that you can use for
page-format:
PAGE
prints only one row of variables for each observation per page. When you use
ROWS=PAGE, PROC PRINT does not divide the page into sections; it prints as
many observations as possible on each page. If the observations do not ll the last
page of the output, PROC PRINT divides the last page into sections and prints all
the variables for the last few observations.
Restriction: Physical page size does not mean the same thing in HTML output as it
does in traditional procedure output. Therefore, HTML output from PROC PRINT
appears the same whether or not you use ROWS=.
Tip: The PAGE value can reduce the number of pages in the output if the data set
contains large numbers of variables and observations. However, if the data set
contains a large number of variables but few observations, the PAGE value can
increase the number of pages in the output.
See also: Page Layout on page 720 for discussion of the default layout.
Featured in:

Example 7 on page 754

SPLIT=split-character

species the split character, which controls line breaks in column headers. It also
uses labels as column headers. PROC PRINT breaks a column heading when it
reaches the split character and continues the header on the next line. The split

The PRINT Procedure

PROC PRINT Statement

711

character is not part of the column heading although each occurrence of the split
character counts toward the 256-character maximum for a label.
Alias:

Interaction: You do not need to use both LABEL and SPLIT= because SPLIT=

implies the use of labels.

Interaction: The OBS= option honors the split character. (See the discussion of

OBS= on page 709.)

Featured in:

Example 2 on page 727

Note: PROC PRINT does not split labels of BY variables in the heading preceding
each BY group even if you specify SPLIT=. Instead, PROC PRINT replaces the split
character with a blank. 4
STYLE

<(location(s))>=< style-element-name>< [style-attribute-specication(s)]>

species the style element to use for the specied locations in the report.
Note: You can use braces ({ and }) instead of square brackets ([ and ]).

location
identies the part of the report that the STYLE option affects. The following table
shows the available locations and the other statements in which you can specify
them.
Note: Style specications in a statement other than the PROC PRINT
statement override the same style specication in the PROC PRINT statement.
However, style attributes that you specify in the PROC PRINT statement are
inherited, provided that you do not override the style with style specications in
another statement. For instance, if you specify a blue background and a white
foreground for all column headers in the PROC PRINT statement, and you specify
a gray background for the column headers of a variable in the VAR statement, the
background for that particular column header is gray, and the foreground is white
(as specied in the PROC PRINT statement). 4
Table 34.1 Specifying Locations in the STYLE Option
This location

Affects this part of the report

And can also be specied for

individual items in this
statement

BYLABEL

the label for the BY variable

on the line containing the
SUM totals

none

DATA

the cells of all columns

VAR
ID
SUM

GRANDTOTAL

the SUM line containing the

grand totals for the whole
report

SUM

HEADER

all column headers

VAR
ID
SUM

712

PROC PRINT Statement

Chapter 34

This location

Affects this part of the report

And can also be specied for

individual items in this
statement

N= table and contents

none

OBS

the data in the OBS column

none

OBSHEADER

the header of the OBS

column

none

TABLE

the structural part of the

report - that is, the
underlying table used to set
things like the width of the
border and the space
between cells

none

TOTAL

the SUM line containing

totals for each BY group

SUM

For your convenience and for consistency with other procedures, the following
table shows aliases for the different locations.
Table 34.2

Aliases for Locations

Location
BYLABEL

Aliases
BYSUMLABEL
BYLBL
BYSUMLBL

DATA

COLUMN
COL

GRANDTOTAL

GRANDTOT
GRAND
GTOTAL
GTOT

HEADER

HEAD
HDR

none

OBS

OBSDATA
OBSCOLUMN
OBSCOL

OBSHEADER

OBSHEAD
OBSHDR

TABLE

REPORT

TOTAL

TOT
BYSUMLINE
BYLINE
BYSUM

The PRINT Procedure

PROC PRINT Statement

713

style-element-name
is the name of a style element that is part of a style denition that is registered
with the Output Delivery System. SAS provides some style denitions. Users can
create their own style denitions with PROC TEMPLATE.
When style elements are processed, more specic style elements override less
specic style elements.
Default: The following table shows the default style element for each location.
Table 34.3

The Default Style Element for Each Location in PROC PRINT

Location

Default style element

BYLABEL

Header

DATA

Data (for all but ID statement)

RowHeader (for ID statement)

GRANDTOTAL

Header

HEADER

Header

NoteContent

OBS

RowHeader

OBSHEADER

Header

TABLE

Table

TOTAL

Header

style-attribute-specication
describes the style attribute to change. Each style-attribute-specication has this
general form:
style-attribute-name=style-attribute-value
You can set these style attributes in the TABLE location:

BACKGROUND=

FONT_WIDTH=*

BACKGROUNDIMAGE=

FOREGROUND=*

BORDERCOLOR=

FRAME=

BORDERCOLORDARK=

HTMLCLASS=

BORDERCOLORLIGHT=

JUST=

BORDERWIDTH=

OUTPUTWIDTH=

CELLPADDING=

POSTHTML=

CELLSPACING=

POSTIMAGE=

FONT=*

POSTTEXT=

FONT_FACE=*

PREHTML=

FONT_SIZE=*

PREIMAGE=

FONT_STYLE=*

PRETEXT=

FONT_WEIGHT=*

RULES=

714

PROC PRINT Statement

Chapter 34

*When you use these attributes, they affect only the text that is specied with the
PRETEXT=, POSTTEXT=, PREHTML=, and POSTHTML= attributes. To alter the
foreground color or the font for the text that appears in the table, you must set the
corresponding attribute in a location that affects the cells rather than the table.
You can set these style attributes in all locations other than TABLE:

ASIS=

FONT_WIDTH=

BACKGROUND=

HREFTARGET=

BACKGROUNDIMAGE=

HTMLCLASS=

BORDERCOLOR=

JUST=

BORDERCOLORDARK=

NOBREAKSPACE=

BORDERCOLORLIGHT=

POSTHTML=

BORDERWIDTH=

POSTIMAGE=

CELLHEIGHT=

POSTTEXT=

CELLWIDTH=

PREHTML=

FLYOVER=

PREIMAGE=

FONT=

PRETEXT=

FONT_FACE=

PROTECTSPECIALCHARS=

FONT_SIZE=

TAGATTR=

FONT_STYLE=

URL=

FONT_WEIGHT=

VJUST=

For information about style attributes, see DEFINE STYLE statement in SAS
Output Delivery System: Users Guide.
Restriction: This option affects all destinations except Listing and Output.
UNIFORM

See WIDTH=UNIFORM on page 715.

WIDTH=column-width

determines the column width for each variable. The value of column-width must be
one of the following:
FULL
uses a variables formatted width as the column width. If the variable does not
have a format that explicitly species a eld width, PROC PRINT uses the default
width. For a character variable, the default width is the length of the variable.
For a numeric variable, the default width is 12. When you use WIDTH=FULL, the
column widths do not vary from page to page.
Tip: Using WIDTH=FULL can reduce execution time.

MINIMUM
uses for each variable the minimum column width that accommodates all values of
the variable.
Alias: MIN

The PRINT Procedure

BY Statement

715

UNIFORM
uses each variables formatted width as its column width on all pages. If the
variable does not have a format that explicitly species a eld width, PROC
PRINT uses the widest data value as the column width. When you specify
WIDTH=UNIFORM, PROC PRINT normally needs to read the data set twice.
However, if all the variables in the data set have formats that explicitly specify a
eld width (for example, BEST12. but not BEST.), PROC PRINT reads the data
set only once.
Alias: U
Tip: If the data set is large and you want a uniform report, you can save computer

resources by using formats that explicitly specify a eld width so that PROC
PRINT reads the data only once.
Tip: WIDTH=UNIFORM is the same as UNIFORM.
Restriction: When not all variables have formats that explicitly specify a width,

you cannot use WIDTH=UNIFORM with an engine that supports concurrent

access if another user is updating the data set at the same time.
UNIFORMBY
formats all columns uniformly within a BY group, using each variables formatted
width as its column width. If the variable does not have a format that explicitly
species a eld width, PROC PRINT uses the widest data value as the column
width.
Alias: UBY
Restriction: You cannot use UNIFORMBY with a sequential data set.
Default: If you omit WIDTH= and do not specify the UNIFORM option, PROC

PRINT individually constructs each page of output. The procedure analyzes the
data for a page and decides how best to display them. Therefore, column widths
may differ from one page to another.
Column width is affected not only by variable width but also by the length of
column headings. Long column headings may lessen the usefulness of WIDTH=.

Tip:

See also: For a discussion of default column widths, see Column Width on page

723.

BY Statement
Produces a separate section of the report for each BY group.
Main discussion:

BY on page 58

Example 3 on page 731, Example 4 on page 737, Example 5 on page 742,

Example 6 on page 748, and Example 8 on page 761

Featured in:

BY <DESCENDING> variable-1
< <DESCENDING> variable-n>
<NOTSORTED>;

716

ID Statement

Chapter 34

Required Arguments
variable

species the variable that the procedure uses to form BY groups. You can specify
more than one variable. If you do not use the NOTSORTED option in the BY
statement, the observations in the data set must either be sorted by all the variables
that you specify, or they must be indexed appropriately. Variables in a BY statement
are called BY variables.

Options
DESCENDING

species that the data set is sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data are grouped in another way, such as chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations with the same values for the BY variables
are not contiguous, the procedure treats each contiguous set as a separate BY group.

Using the BY Statement with an ID Statement

PROC PRINT uses a special layout if all BY variables appear in the same order at
the beginning of the ID statement. (See Example 8 on page 761.)

Using the BY Statement with the NOBYLINE Option

If you use the BY statement with the SAS system option NOBYLINE, which
suppresses the BY line that normally appears in output produced with BY-group
processing, PROC PRINT always starts a new page for each BY group. This behavior
ensures that if you create customized BY lines by putting BY-group information in the
title and suppressing the default BY lines with NOBYLINE, the information in the
titles matches the report on the pages.

ID Statement
Identies observations by using the formatted values of the variables that you list instead of by
using observation numbers.
Featured in:

Example 7 on page 754 and Example 8 on page 761

ID variable(s) </ STYLE <(location(s))>

=<style-element-name><[style-attribute-specication(s)]>>;

The PRINT Procedure

PAGEBY Statement

717

Required Arguments
variable(s)

species one or more variables to print instead of the observation number at the
beginning of each row of the report.
Restriction: If the ID variables occupy so much space that no room remains on the

line for at least one other variable, PROC PRINT writes a warning to the SAS log
and does not treat all ID variables as ID variables.
Interaction: If a variable in the ID statement also appears in the VAR statement,

the output contains two columns for that variable.

Options
STYLE <(location(s))>=<style-element-name><[style-attribute-specication(s)]>

species the style element to use for ID columns created with the ID statement. For
information about the arguments of this option and how it is used, see STYLE on
page 711 in the PROC PRINT statement.
To specify different style elements for different ID columns, use a separate ID
statement for each variable and add a different STYLE option to each ID
statement.

Tip:

Using the BY Statement with an ID Statement

PROC PRINT uses a special layout if all BY variables appear in the same order at
the beginning of the ID statement. (See Example 8 on page 761.)

PAGEBY Statement
Controls page ejects that occur before a page is full.
Requirements:

BY statement

Featured in: Example 3 on page 731

PAGEBY BY-variable;

Required Arguments

BY-variable

identies a variable appearing in the BY statement in the PROC PRINT step. If the
value of the BY variable changes, or if the value of any BY variable that precedes it
in the BY statement changes, PROC PRINT begins printing a new page.

718

SUM Statement

Chapter 34

Interaction: If you use the BY statement with the SAS system option NOBYLINE,

which suppresses the BY line that normally appears in output produced with
BY-group processing, PROC PRINT always starts a new page for each BY group.
This behavior ensures that if you create customized BY lines by putting BY-group
information in the title and suppressing the default BY lines with NOBYLINE, the
information in the titles matches the report on the pages. (See Creating Titles
That Contain BY-Group Information on page 20.)

SUM Statement
Totals values of numeric variables.
Example 4 on page 737, Example 5 on page 742, Example 6 on page 748,
and Example 8 on page 761

Featured in:

SUM variable(s) </ STYLE <(location(s))>

=<style-element-name><[style-attribute-specication(s)]>>;

Required Arguments

variable(s)

identies the numeric variables to total in the report.

Option
STYLE <(location(s))>=<style-element-name><[style-attribute-specication(s)]>

species the style element to use for cells containing sums that are created with the
SUM statement. For information about the arguments of this option and how it is
used, see STYLE on page 711 in the PROC PRINT statement.
To specify different style elements for different cells reporting sums, use a
separate SUM statement for each variable and add a different STYLE option to
each SUM statement.

Tip:

If the STYLE option is used in multiple SUM statements that affect the same
location, the STYLE option in the last SUM statement will be used.

Tip:

Using the SUM and BY Statements Together

When you use a SUM statement and a BY statement with one BY variable, PROC
PRINT sums the SUM variables for each BY group that contains more than one
observation and totals them over all BY groups (see Example 4 on page 737).
When you use a SUM statement and a BY statement with multiple BY variables,
PROC PRINT sums the SUM variables for each BY group that contains more than one
observation, just as it does if you use only one BY variable. However, it provides sums
only for those BY variables whose values change when the BY group changes. (See
Example 5 on page 742.)

The PRINT Procedure

VAR Statement

719

Note: When the value of a BY variable changes, the SAS System considers that the
values of all variables listed after it in the BY statement also change. 4

SUMBY Statement
Limits the number of sums that appear in the report.
BY statement
Featured in: Example 6 on page 748
Requirements:

SUMBY BY-variable;

Required Arguments
BY-variable

identies a variable that appears in the BY statement in the PROC PRINT step. If
the value of the BY variable changes, or if the value of any BY variable that precedes
it in the BY statement changes, PROC PRINT prints the sums of all variables listed
in the SUM statement.

What Variables Are Summed?

If you use a SUM statement, PROC PRINT subtotals only the SUM variables.
Otherwise, PROC PRINT subtotals all the numeric variables in the data set except
those listed in the ID and BY statements.

VAR Statement
Selects variables that appear in the report and determines their order.
If you omit the VAR statement, PROC PRINT prints all variables in the data set.
Featured in: Example 1 on page 723 and Example 8 on page 761
Tip:

VAR variable(s) </ STYLE <(location(s))>

=<style-element-name><[style-attribute-specication(s)]>>;

720

Results: Print Procedure

Chapter 34

Required Arguments
variable(s)

identies the variables to print. PROC PRINT prints the variables in the order that
you list them.
Interaction: In the PROC PRINT output, variables that are listed in the ID
statement precede variables that are listed in the VAR statement. If a variable in
the ID statement also appears in the VAR statement, the output contains two
columns for that variable.

Option
STYLE <(location(s))>=<style-element-name><[style-attribute-specication(s)]>

species the style element to use for all columns that are created by a VAR
statement. For information about the arguments of this option and how it is used,
see STYLE on page 711 in the PROC PRINT statement.
Tip: To specify different style elements for different columns, use a separate VAR
statement to create a column for each variable and add a different STYLE option
to each VAR statement.

Results: Print Procedure

Procedure Output
PROC PRINT always produces a printed report. You control the appearance of the
report with statements and options. See Examples: PRINT Procedure on page 723 for
a sampling of the types of reports that the procedure produces.

Page Layout

Observations
By default, PROC PRINT uses an identical layout for all observations on a page of
output. First, it attempts to print observations on a single line (see Figure 34.2 on page
721).

The PRINT Procedure

Figure 34.2

Page Layout

721

Printing Observations on a Single Line

1
Obs

Var_1

Var_2

Var_3

~~~~

If PROC PRINT cannot t all the variables on a single line, it splits the observations
into two or more sections and prints the observation number or the ID variables at the
beginning of each line. For example, in Figure 34.3 on page 721, PROC PRINT prints
the values for the rst three variables in the rst section of each page and the values
for the second three variables in the second section of each page.

Figure 34.3

Splitting Observations into Multiple Sections on One Page

Obs

Var_1

Var_2

Var_3

~~~~

Obs

Var_4

~~~~

Var_5
Var_6
Obs
Var_1
~~~~
~~~~
4
~~~~
~~~~
~~~~
5
~~~~
~~~~
~~~~
6
~~~~

2
Var_2

Var_3

~~~~

Obs

Var_4

Var_5

Var_6

~~~~

If PROC PRINT cannot t all the variables on one page, the procedure prints
subsequent pages with the same observations until it has printed all the variables. For
example, in Figure 34.4 on page 722, PROC PRINT uses the rst two pages to print
values for the rst three observations and the second two pages to print values for the
rest of the observations.

722

Page Layout

Chapter 34

Figure 34.4

Splitting Observations across Multiple Pages

1
2

Obs

Var_1

Var_2

Var_3

~~~~

Obs

Var_7

Var_8

Var_9

~~~~

Obs

Var_4

Var_5

Var_6

~~~~

Obs

Var_10

Var_11

Var_12

~~~~

3
4

Obs

Var_1

Var_2

Var_3

~~~~

Obs

Var_7

Var_8

Var_9

~~~~

Var_10

Var_11

Var_12

Obs

Var_4

Var_5

Var_6

~~~~

Obs

~~~~

Note: You can alter the page layout with the ROWS= option in the PROC PRINT
statement (see the discussion of ROWS= on page 710). 4
Note: PROC PRINT may produce slightly different output if the data set is not
RADIX addressable. Version 6 compressed les are not RADIX addressable, while,
beginning with Version 7, compressed les are RADIX addressable. (The integrity of the
data is not compromised; the procedure simply numbers the observations differently.) 4

Column Headings
By default, spacing dictates whether PROC PRINT prints column headings
horizontally or vertically. Figure 34.2 on page 721, Figure 34.3 on page 721, and Figure
34.4 on page 722 all illustrate horizontal headings. Figure 34.5 on page 722 illustrates
vertical headings.
Figure 34.5

Using Vertical Headings

O
b
s

V
a
r

~~~~

Note: If you use LABEL and at least one variable has a label, PROC PRINT prints
all column headings horizontally unless you specify HEADING=VERTICAL. 4

The PRINT Procedure

Example 1: Selecting Variables to Print

723

Column Width
By default, PROC PRINT uses a variables formatted width as the column width.
(The WIDTH= option overrides this default behavior.) If the variable does not have a
format that explicitly species a eld width, PROC PRINT uses the widest data value
for that variable on that page as the column width.
If the formatted value of a character variable or the data width of an unformatted
character variable exceeds the linesize minus the length of all the ID variables, PROC
PRINT may truncate the value. Consider the following situation:

3
3
3
3

The linesize is 80.

IdNumber is a character variable with a length of 10. It is used as an ID variable.
State is a character variable with a length of 2. It is used as an ID variable.
Comment is a character variable with a length of 200.

When PROC PRINT prints these three variables on a line, it uses 14 print positions
for the two ID variables and the space after each one. This leaves 8014, or 66, print
positions for COMMENT. Longer values of COMMENT are truncated.
WIDTH= controls the column width.
Note: Column width is affected not only by variable width but also by the length of
column headings. Long column headings may lessen the usefulness of WIDTH=. 4

Examples: PRINT Procedure

Example 1: Selecting Variables to Print

Procedure features:

PROC PRINT statement options:

DOUBLE
STYLE
VAR statement
Other Features:

ODS HTML statement

This example
3 selects three variables for the report

3 uses variable labels as column headings

3 double spaces between rows of the report.

724

Program: Creating a Listing Report

Chapter 34

Program: Creating a Listing Report

Set the SAS system options.

options nodate pageno=1 linesize=70 pagesize=60;

Create the input data set. EXPREV contains information about a companys monthly
expenses and revenues for two regions of the United States.
data exprev;
input Region $ State $ Month monyy5.
Expenses Revenues;
format month monyy5.;
datalines;
Southern GA JAN95 2000 8000
Southern GA FEB95 1200 6000
Southern FL FEB95 8500 11000
Northern NY FEB95 3000 4000
Northern NY MAR95 6000 5000
Southern FL MAR95 9800 13500
Northern MA MAR95 1500 1000
;

Print the data set EXPREV. DOUBLE inserts a blank line between observations. (This option
has no effect on the HTML output.)
proc print data=exprev double;

Select the variables to include in the report. The VAR statement creates columns for
Month, State, and Expenses, in that order.
var month state expenses;

Specify a title. The TITLE statement species a title for the report.
title Monthly Expenses for Offices in Each State;
run;

The PRINT Procedure

Program: Creating an HTML Report

725

Output: Listing

Output 34.2

Selecting Variables: Listing Output

By default, PROC PRINT identies each observation by number under the column heading Obs.

Monthly Expenses for Offices in Each State

Obs

Month

State

Expenses

JAN95

2000

FEB95

1200

FEB95

8500

FEB95

3000

MAR95

6000

MAR95

9800

MAR95

1500

Program: Creating an HTML Report

You can easily create HTML output by adding ODS statements. In the following
example, ODS statements were added to produce HTML output.
options nodate pageno=1 linesize=70 pagesize=60;

Create HTML output and specify the le to store the output in. The ODS HTML
statement opens the HTML destination. FILE= species the external le that you want to
contain the HTML output.
ods html file=your_file.html;
proc print data=exprev double;
var month state expenses;
title Monthly Expenses for Offices in Each State;
run;

Close the HTML destination. The ODS HTML CLOSE statement closes the HTML
destination.
ods html close;

726

Output: HTML

Chapter 34

Output: HTML
Display 34.2

Selecting Variables: Default HTML Output

Program: Creating an HTML Report with the STYLE Option

You can go a step further and add more formatting to your HTML output. The
following example uses the STYLE option to add shading to your HTML report.
options nodate pageno=1 linesize=70 pagesize=60;
ods html file=your_file.html;

Create stylized HTML output. The rst STYLE option species that the column headers be
written in white italic font.
The second STYLE option species that SAS change the color of the background of the
observations column to red.

Proc Print data=exprev double

style(HEADER) = {font_style=italic foreground = white}
style(OBS) = {background=red};
var month state expenses;
title Monthly Expenses for Offices in Each State;
run;

Close the HTML destination. The ODS HTML CLOSE statement closes the HTML
destination.
ods html close;

The PRINT Procedure

Program: Creating a Listing Report

Output: HTML Output with Styles

Display 34.3

Selecting Variables: HTML Output Using Styles

Example 2: Customizing Text in Column Headers

Procedure features:

PROC PRINT statement options:

N
OBS=
SPLIT=
STYLE
VAR statement option:
STYLE
Other features:

LABEL statement
ODS PDF statement
Data set:

EXPREV on page 724

This example

3 customizes and underlines the text in column headings for variables

3 customizes the column header for the column that identies observations by
number

3 shows the number of observations in the report

3 writes the values of Expenses with commas.

Program: Creating a Listing Report

options nodate pageno=1 linesize=70 pagesize=60;

727

728

Program: Creating a Listing Report

Chapter 34

Print the report and dene the column headings. SPLIT= identies the asterisk as the
character that starts a new line in column headers. The N option prints the number of
observations at the end of the report. OBS= species the column header for the column that
identies each observation by number. The split character (*) starts a new line in the column
heading. Therefore, the equal signs (=) in the value of OBS= underline the column header.
proc print data=exprev split=* n obs=Observation*Number*===========;

Select the variables to include in the report. The VAR statement creates columns for
Month, State, and Expenses, in that order.
var month state expenses;

Assign the variables labels as column headings. The LABEL statement associates a label
with each variable for the duration of the PROC PRINT step. When you use SPLIT= in the
PROC PRINT statement, the procedure uses labels for column headers. The split character (*)
starts a new line in the column heading. Therefore, the equal signs (=) in the labels underline
the column headers.
label month=Month**=====
state=State**=====
expenses=Expenses**========;

Specify a title for the report, and format any variable containing numbers. The
FORMAT statement assigns a format to use for Expenses in the report. The TITLE statement
species a title.
format expenses comma10.;
title Monthly Expenses for Offices in Each State;
run;

The PRINT Procedure

Program: Creating a PDF Report

729

Output: Listing

Output 34.3

Customizing Text in Column Headers: Listing Output

Monthly Expenses for Offices in Each State
Observation
Number

Month

State

Expenses

===========

=====

========

JAN95

2,000

2
3

FEB95
FEB95

GA
FL

1,200
8,500

4
5
6

FEB95
MAR95
MAR95

NY
NY
FL

3,000
6,000
9,800

MAR95

1,500

N = 7

Program: Creating a PDF Report

You can easily create PDF output by adding a few ODS statements. In the following
example, ODS statements were added to produce PDF output.
options nodate pageno=1 linesize=70 pagesize=60;

Create PDF output and specify the le to store the output in. The ODS PDF statement
opens the PDF destination and creates PDF output. The FILE= argument species your
external le that contains the PDF output.
ods pdf file=your_file.pdf;
proc print data=exprev split=* n obs=Observation*Number*===========;
var month state expenses;
label month=Month**=====
state=State**=====
expenses=Expenses**========;
format expenses comma10.;
title Monthly Expenses for Offices in Each State;
run;

Close the PDF destination. The ODS PDF CLOSE statement closes the PDF destination.
ods pdf close;

730

Output: PDF

Chapter 34

Output: PDF
Display 34.4

Customizing Text in Column Headers: Default PDF Output

Program: Creating a PDF Report with the STYLE Option

options nodate pageno=1 linesize=70 pagesize=60;
ods pdf file=your_file.pdf;

Create stylized PDF output. The rst STYLE option species that the background color of
the cell containing the value for N be changed to blue and that the font style be changed to
italic. The second STYLE option species that the background color of the observation column,
the observation header, and the other variables headers be changed to white.
proc print data=exprev split=* n obs=Observation*Number*===========
style(N) = {font_style=italic background= blue}
Style(HEADER OBS OBSHEADER) = {background=white};

Create stylized PDF output. The STYLE option changes the color of the cells containing
data to gray.
var month state expenses / style (DATA)= [ background = gray ] ;
label month=Month**=====
state=State**=====
expenses=Expenses**========;
format expenses comma10.;

The PRINT Procedure

Example 3: Creating Separate Sections of a Report for Groups of Observations

title Monthly Expenses for Offices in Each State;

run;

Close the PDF destination. The ODS PDF CLOSE statement closes the PDF destination.
ods pdf close;

Output: PDF Report with Styles

Display 34.5

Customizing Text in Column Headers: PDF Output Using Styles

Example 3: Creating Separate Sections of a Report for Groups of

Observations
Procedure features:

PROC PRINT statement options:

LABEL
N=
NOOBS
STYLE
BY statement
PAGEBY statement

731

732

Program: Creating a Listing Report

Chapter 34

Other features:

SORT procedure
LABEL statement
ODS RTF statement
Data set:

EXPREV on page 724

This example

3 suppresses the printing of observation numbers at the beginning of each row

3 presents the data for each state in a separate section of the report
3 begins a new page for each region.

Program: Creating a Listing Report

options pagesize=60 pageno=1 nodate linesize=70;

Sort the EXPREV data set. PROC SORT sorts the observations by Region, State, and Month.
proc sort data=exprev;
by region state month;
run;

Print the report, specify the total number of observations in each BY group, and
suppress the printing of observation numbers. N= prints the number of observations in a
BY group at the end of that BY group. The explanatory text that the N= option provides
precedes the number. NOOBS suppresses the printing of observation numbers at the beginning
of the rows. LABEL uses variables labels as column headings.
proc print data=exprev n=Number of observations for the state:
noobs label;

Specify the variables to include in the report. The VAR statement creates columns for
Month, Expenses, and Revenues, in that order.
var month expenses revenues;

Create a separate section for each region of the state and specify page breaks for each
BY group of Region. The BY statement produces a separate section of the report for each BY
group and prints a heading above each one. The PAGEBY statement starts a new page each
time the value of Region changes.
by region state;
pageby region;

The PRINT Procedure

Output: Listing

733

Establish the column headings. The LABEL statement associates a label with the variable
Region for the duration of the PROC PRINT step. When you use the LABEL option in the
PROC PRINT statement, the procedure uses labels for column headings.
label region=Sales Region;

Format the columns that contain numbers and specify a title. The FORMAT statement
assigns a format to Expenses and Revenues for this report. The TITLE statement species a
title.
format revenues expenses comma10.;
title Sales Figures Grouped by Region and State;
run;

Output: Listing

Output 34.4

Creating Separate Sections of a Report for Groups of Observations: Listing Output

Sales Figures Grouped by Region and State

------------------- Sales Region=Northern State=MA ------------------Month

Expenses

Revenues

MAR95

1,500

1,000

Number of observations for the state: 1

------------------- Sales Region=Northern State=NY ------------------Month

Expenses

Revenues

FEB95
MAR95

3,000
6,000

4,000
5,000

Number of observations for the state: 2

734

Program: Creating an RTF Report

Chapter 34

Sales Figures Grouped by Region and State

------------------- Sales Region=Southern State=FL ------------------Month

Expenses

Revenues

FEB95
MAR95

8,500
9,800

11,000
13,500

Number of observations for the state: 2

------------------- Sales Region=Southern State=GA ------------------Month

Expenses

Revenues

JAN95
FEB95

2,000
1,200

8,000
6,000

Number of observations for the state: 2

Program: Creating an RTF Report

options pagesize=60 pageno=1 nodate linesize=70;

Create output for Microsoft Word and specify the le to store the output in. The ODS
RTF statement opens the RTF destination and creates output formatted for Microsoft Word. The
FILE= option species your external le that contains the RTF output. The STARTPAGE=NO
option species that no new pages be inserted within the PRINT procedure, even if new pages
are requested by the procedure code.
ods rtf startpage=no file=your_file.rtf;
proc sort data=exprev;
by region state month;
run;
proc print data=exprev n=Number of observations for the state:
noobs label;
var month expenses revenues;
by region state;
pageby region;
label region=Sales Region;
format revenues expenses comma10.;
title Sales Figures Grouped by Region
and State;
run;

Close the RTF destination. The ODS RTF CLOSE statement closes the RTF destination.
ods rtf close;

The PRINT Procedure

Program: Creating an RTF Report with the STYLE Option

Output: RTF
Display 34.6

Creating Separate Sections of a Report for Groups of Observations: Default RTF Output

Program: Creating an RTF Report with the STYLE Option

options pagesize=60 pageno=1 nodate linesize=70;
ods rtf file=your_file.rtf;
proc sort data=exprev;
by region state month;
run;

735

736

Program: Creating an RTF Report with the STYLE Option

Chapter 34

Create a stylized RTF report. The rst STYLE option species that the background color
of the cell containing the number of observations be changed to gray.
The second STYLE option species that the background color of the column header for the
variable MONTH be changed to white.
The third STYLE option species that the background color of the column header for the
variable EXPENSES be changed to blue and the font color be changed to white.
The fourth STYLE option species that the background color of the column header for the
variable REVENUES be changed to gray.
proc print data=exprev n=Number of observations for the state:
noobs label style(N) = {background=gray};
var month / style(HEADER) = [background = white];
var expenses / style(HEADER) = [background = blue foreground=white];
var revenues / style(HEADER) = [background = gray];
by region state;
pageby region;
label region=Sales Region;
format revenues expenses comma10.;
title Sales Figures Grouped by Region
and State;
run;
ods rtf close;

The PRINT Procedure

Example 4: Summing Numeric Variables with One BY Group

Output: RTF with Styles

Display 34.7

Creating Separate Sections of a Report for Groups of Observations: RTF Output Using Styles

Example 4: Summing Numeric Variables with One BY Group

Procedure features:

PROC PRINT statement options:

N=
BY statement

737

738

Program: Creating a Listing Report

Chapter 34

SUM statement
Other features:

ODS MARKUP statement

SORT procedure
TITLE statement
#BYVAL specication
SAS system options:
BYLINE
NOBYLINE
Data set:

EXPREV on page 724

This example

3 sums expenses and revenues for each region and for all regions
3 shows the number of observations in each BY group and in the whole report
3 creates a customized title, containing the name of the region. This title replaces
the default BY line for each BY group.

Program: Creating a Listing Report

Start each BY group on a new page and suppress the printing of the default BY line.
The SAS system option NOBYLINE suppresses the printing of the default BY line. When you
use PROC PRINT with NOBYLINE, each BY group starts on a new page.
options nodate pageno=1 linesize=70 pagesize=60 nobyline;

Sort the data set. PROC SORT sorts the observations by Region.

proc sort data=exprev;

by region;
run;

Print the report, suppress the printing of observation numbers, and print the total
number of observations for the selected variables. NOOBS suppresses the printing of
observation numbers at the beginning of the rows. N= prints the number of observations in a
BY group at the end of that BY group and (because of the SUM statement) prints the number
of observations in the data set at the end of the report. The rst piece of explanatory text that
N= provides precedes the number for each BY group. The second piece of explanatory text that
N= provides precedes the number for the entire data set.
proc print data=exprev noobs
n=Number of observations for the state:
Number of observations for the data set: ;

The PRINT Procedure

Output: Listing

739

Sum the values for the selected variables. The SUM statement alone sums the values of
Expenses and Revenues for the entire data set. Because the PROC PRINT step contains a BY
statement, the SUM statement also sums the values of Expenses and Revenues for each region
that contains more than one observation.
sum expenses revenues;
by region;

Format the numeric values for a specied column. The FORMAT statement assigns the
COMMA10. format to Expenses and Revenues for this report.
format revenues expenses comma10.;

Specify and format a dynamic (or current) title. The TITLE statement species a title. The
#BYVAL specication places the current value of the BY variable Region in the title. Because
NOBYLINE is in effect, each BY group starts on a new page, and the title serves as a BY line.
title Revenue and Expense Totals for the
#byval(region) Region;
run;

Generate the default BY line. The SAS system option BYLINE resets the printing of the
default BY line.
options byline;

Output: Listing

Output 34.5

Summing Numeric Variables with One BY Group: Listing Output

Revenue and Expense Totals for the Northern Region
State

Month

Expenses

Revenues

NY
NY
MA
-----Region

FEB95
MAR95
MAR95

3,000
6,000
1,500
---------10,500

4,000
5,000
1,000
---------10,000

Number of observations for the state: 3

740

Program: Creating an XML File

Chapter 34

Revenue and Expense Totals for the Southern Region

State

Month

Expenses

Revenues

GA
GA
FL
FL
-----Region

JAN95
FEB95
FEB95
MAR95

2,000
1,200
8,500
9,800
---------21,500
==========
32,000

8,000
6,000
11,000
13,500
---------38,500
==========
48,500

Number of observations for the state: 4

Number of observations for the data set: 7

Program: Creating an XML File

The following example opens the MARKUP destination. The output le will contain
only XML tagging unless you have a browser that reads XML.
options nodate pageno=1 linesize=70 pagesize=60 nobyline;

Produce output that is tagged with Extensible Markup Language (XML) tags and
specify the le to store it in. The ODS MARKUP statement opens the MARKUP destination
and creates a le containing output that is tagged with XML tags. The FILE= argument
species your external le that contains the XML output.
ods markup file=your_file.xml;

proc sort data=exprev;

by region;
run;

proc print data=exprev noobs

n=Number of observations for the state:
Number of observations for the data set: ;

sum expenses revenues;

by region;

format revenues expenses comma10.;

title Revenue and Expense Totals for the

#byval(region) Region;
run;

options byline;

The PRINT Procedure

Output: XML le

Close the MARKUP destination. The ODS RTF CLOSE statement closes the MARKUP
destination.
ods markup close;

Output: XML le
Output 34.6

Summing Numeric Variables with One BY Group: Partial XML Output Viewed with a Text Editor

<?xml version="1.0" encoding="windows-1252"?>

<odsxml>
<head>
<meta operator="user"/>
</head>
<body>
<proc name="Print">
<label name="IDX"/>
<title class="SystemTitle" toc-level="1">Revenue and Expense Totals for the Northern Region</title>
<branch name="Print" label="The Print Procedure" class="ContentProcName" toc-level="1">
<bygroup>
<branch name="ByGroup1" label="ByGroup1" class="ByContentFolder" toc-level="2">
<leaf name="Print" label="Data Set WORK.EXPREV" class="ContentItem" toc-level="3">
<output name="Print" label="Data Set WORK.EXPREV" clabel="Data Set WORK.EXPREV">
<output-object type="table" class="Table">
<style>
<border spacing="1" padding="7" rules="groups" frame="box"/>
</style>
<colspecs columns="4">
<colgroup>
<colspec name="1" width="6" type="string"/>
<colspec name="2" width="5" type="string"/>
<colspec name="3" width="10" type="string"/>
<colspec name="4" width="10" type="string"/>
</colgroup>
... more lines of XML output ...
<row>
<data type="string" class="NoteContent" row="8" column="1" column-end="4">
<style>
<span columns="4"/>
</style>
<value>Number of observations for the state: 4<br/>Number of observations for the data set: 7</value>
</data>
</row>
</output-body>
</output-object>
</output>
</leaf>
</bygroup>
</branch>
</branch>
</proc>
</body>
</odsxml>

741

742

Example 5: Summing Numeric Variables with Multiple BY Variables

Chapter 34

Example 5: Summing Numeric Variables with Multiple BY Variables

Procedure features:

BY statement
SUM statement
Other features: SORT procedure
Data set:

EXPREV on page 724

This example
3 sums expenses and revenues for
3 each region
3 each state with more than one row in the report
3 all rows in the report.

3 shows the number of observations in each BY group and in the whole report.

Program: Creating a Listing Report

options nodate pageno=1 linesize=70 pagesize=60;

Sort the data set. PROC SORT sorts the observations by Region and State.

proc sort data=exprev;

by region state;
run;

Print the report, suppress the printing of observation numbers, and print the total
number of observations for the selected variables. The N option prints the number of
observations in a BY group at the end of that BY group and prints the total number of
observations used in the report at the bottom of the report. NOOBS suppresses the printing of
observation numbers at the beginning of the rows.
proc print data=exprev n noobs;

Create a separate section of the report for each BY group, and sum the values for the
selected variables. The BY statement produces a separate section of the report for each BY
group. The SUM statement alone sums the values of Expenses and Revenues for the entire data
set. Because the program contains a BY statement, the SUM statement also sums the values of
Expenses and Revenues for each BY group that contains more than one observation.
by region state;
sum expenses revenues;

The PRINT Procedure

Program: Creating a Listing Report

743

Establish a label for a selected variable, format the values of specied variables, and
create a title. The LABEL statement associates a label with the variable Region for the
duration of the PROC PRINT step. The BY line at the beginning of each BY group uses the
label. The FORMAT statement assigns a format to the variables Expenses and Revenues for
this report. The TITLE statement species a title.
label region=Sales Region;
format revenues expenses comma10.;
title Revenue and Expense Totals for Each State and Region;
run;

744

Output: Listing

Chapter 34

Output: Listing

Output 34.7

Summing Numeric Variables with Multiple BY Variables: Listing Output

The report uses default column headers (variable names) because neither the SPLIT= nor the
LABEL option is used. Nevertheless, the BY line at the top of each section of the report shows
the BY variables labels and their values. The name of a BY variable identies the subtotals in
the report.
PROC PRINT sums Expenses and Revenues for each BY group that contains more than one
observation. However, sums are shown only for the BY variables whose values change from one
BY group to the next. For example, in the third BY group, where the sales region is Southern
and the state is FL, Expenses and Revenues are summed only for the state because the next BY
group is for the same region.

Revenue and Expense Totals for Each State and Region

------------------- Sales Region=Northern State=MA ------------------Month

Expenses

Revenues

MAR95

1,500

1,000

N = 1
------------------- Sales Region=Northern State=NY ------------------Month

Expenses

Revenues

FEB95
MAR95
-----State
Region

3,000
6,000
---------9,000
10,500

4,000
5,000
---------9,000
10,000

N = 2
------------------- Sales Region=Southern State=FL ------------------Month

Expenses

Revenues

FEB95
MAR95
-----State

8,500
9,800
---------18,300

11,000
13,500
---------24,500

N = 2
------------------- Sales Region=Southern State=GA ------------------Month

Expenses

Revenues

JAN95
FEB95
-----State
Region

2,000
1,200
---------3,200
21,500
==========
32,000

8,000
6,000
---------14,000
38,500
==========
48,500

N = 2
Total N = 7

The PRINT Procedure

Program: Creating an HTML Report

options nodate pageno=1 linesize=70 pagesize=60;

Produce HTML output and specify the le to store the output in. The ODS HTML
statement opens the HTML destination and creates a le that contains HTML output. The
FILE= argument species your external le that contains the HTML output.
ods html file=your_file.html;
proc sort data=exprev;
by region state;
run;
proc print data=exprev n noobs;
by region state;
sum expenses revenues;
label region=Sales Region;
format revenues expenses comma10.;
title Revenue and Expense Totals for Each State and Region;
run;

Close the HTML destination. The ODS HTML CLOSE statement closes the HTML
destination.
ods html close;

745

746

Output: HTML

Chapter 34

Output: HTML
Display 34.8

Summing Numeric Variables with Multiple BY Variables: Default HTML Output

Program: Creating an HTML Report with the STYLE Option

options nodate pageno=1 linesize=70 pagesize=60;
ods html file=your_file.html;
proc sort data=exprev;
by region state;
run;
proc print data=exprev n noobs;

The PRINT Procedure

Output: HTML with Styles

747

Create stylized HTML output. The STYLE option in the rst SUM statement species that
the background color of the cell containing the grand total for the variable EXPENSES be
changed to white and the font color be changed to dark gray.
The STYLE option in the second SUM statement species that the background color of cells
containing totals for the variable REVENUES be changed to blue and the font color be changed
to white.
by region state;
sum expenses / style(GRANDTOTAL) = [background =white foreground=blue];
sum revenues / style(TOTAL) = [background =dark gray foreground=white];
label region=Sales Region;
format revenues expenses comma10.;
title Revenue and Expense Totals for Each State and Region;
run;
ods html close;

Output: HTML with Styles

Display 34.9

Summing Numeric Variables with Multiple BY Variables: HTML Output Using Styles

748

Example 6: Limiting the Number of Sums in a Report

Chapter 34

Example 6: Limiting the Number of Sums in a Report

Features:

BY statement
SUM statement
SUMBY statement
Other features:

SORT procedure
LABEL statement
Data set:

EXPREV on page 724

This example

3 creates a separate section of the report for each combination of state and region
3 sums expenses and revenues only for each region and for all regions, not for
individual states.

Program: Creating a Listing Report

options nodate pageno=1 linesize=70 pagesize=60;

Sort the data set. PROC SORT sorts the observations by Region and State.

proc sort data=exprev;

by region state;
run;

Print the report and remove the observation numbers. NOOBS suppresses the printing
of observation numbers at the beginning of the rows.
proc print data=exprev noobs;

Sum the values for each region. The SUM and BY statements work together to sum the
values of Revenues and Expenses for each BY group as well as for the whole report. The
SUMBY statement limits the subtotals to one for each region.
by region state;
sum revenues expenses;
sumby region;

The PRINT Procedure

Program: Creating a Listing Report

749

Assign labels to specic variables. The LABEL statement associates a label with the
variable Region for the duration of the PROC PRINT step. This label is used in the BY lines.
label region=Sales Region;

Assign a format to the necessary variables and specify a title. The FORMAT statement
assigns the COMMA10. format to Expenses and Revenues for this report.
format revenues expenses comma10.;
title Revenue and Expense Figures for Each Region;
run;

750

Output: Listing

Chapter 34

Output: Listing

Output 34.8

Limiting the Number of Sums in a Report: Listing Output

8,000
6,000
---------38,500
==========
48,500

The PRINT Procedure

Program: Creating a PostScript le

options nodate pageno=1 linesize=70 pagesize=60;

Produce PostScript output and specify the le to store the output in. The ODS PS
statement opens the PS destination and creates a le that contains PostScript output. The
FILE= argument species your external le that contains the PostScript output.
ods ps file=your_file.ps;

proc sort data=exprev;

by region state;
run;

proc print data=exprev noobs;

by region state;
sum revenues expenses;
sumby region;

label region=Sales Region;

format revenues expenses comma10.;

title Revenue and Expense Figures for Each Region;
run;

Close the PS destination. The ODS PS CLOSE statement closes the PS destination.
ods ps close;

751

752

Output: PostScript

Chapter 34

Output: PostScript
Display 34.10 Limiting the Number of Sums in a Report: PostScript Output

Program: Creating a PostScript Report with the STYLE Option

options nodate pageno=1 linesize=70 pagesize=60;

ods ps file=your_file.ps;

proc sort data=exprev;

by region state;
run;

proc print data=exprev noobs;

by region state;

The PRINT Procedure

Program: Creating a PostScript Report with the STYLE Option

753

Create stylized PostScript output. The STYLE option in the rst SUM statement species
that the background color of cells containing totals for the variable REVENUES be changed to
blue and the font color be changed to white.
The STYLE option in the second SUM statement species that the background color of the cell
containing the grand total for the EXPENSES variable be changed to white and the font color
be changed to dark gray.
sum revenues / style(TOTAL) = [background =blue foreground=white];
sum expenses / style(GRANDTOTAL) = [background =white foreground=dark gray];

label region=Sales Region;

format revenues expenses comma10.;

title Revenue and Expense Figures for Each Region;
run;

ods ps close;

754

Output: PostScript with Styles

Chapter 34

Output: PostScript with Styles

Display 34.11 Limiting the Number of Sums in a Report: PostScript Output Using Styles

Example 7: Controlling the Layout of a Report with Many Variables

Procedure features:

PROC PRINT statement options:

ROWS=
ID statement options:
STYLE
Other features:

ODS RTF statement

SAS data set options:
OBS=

The PRINT Procedure

Program: Creating a Listing Report

755

This example shows two ways of printing a data set with a large number of
variables: one is the default, and the other uses ROWS=. For detailed explanations of
the layouts of these two reports, see the ROWS= option on page 710 and see Page
Layout on page 720.
These reports use a pagesize of 24 and a linesize of 64 to help illustrate the different
layouts.
Note:

When the two reports are written as HTML output, they do not differ.

Program: Creating a Listing Report

options nodate pageno=1 linesize=64 pagesize=24 ;

Create the EMPDATA data set. The data set EMPDATA contains personal and job-related
information about a companys employees. A DATA step on page 1385 creates this data set.
data empdata;
input IdNumber $ 1-4 LastName $ 9-19 FirstName $ 20-29
City $ 30-42 State $ 43-44 /
Gender $ 1 JobCode $ 9-11 Salary 20-29 @30 Birth date9.
@43 Hired date9. HomePhone $ 54-65;
format birth hired date9.;
datalines;
1919
Adams
Gerald
Stamford
CT
M
TA2
34376
15SEP1948
07JUN1975
203/781-1255
1653
Alexander Susan
Bridgeport
CT
F
ME2
35108
18OCT1952
12AUG1978
203/675-7715
. . . more lines of data . . .
1407
M
1114
F
;

Grant
PT1
Green
TA2

Daniel
68096
Janice
32928

Mt. Vernon
26MAR1957
New York
21SEP1957

NY
21MAR1978

914/468-1616

30JUN1975

212/588-1092

Print only the rst 12 observations in a data set. The OBS= data set option uses only the
rst 12 observations to create the report. (This is just to conserve space here.) The ID statement
identies observations with the formatted value of IdNumber rather than with the observation
number. This report is shown in Example 7 on page 754.
proc print data=empdata(obs=12);
id idnumber;
title Personnel Data;
run;

756

Output: Listing

Chapter 34

Print a report that contains only one row of variables on each page. ROWS=PAGE
prints only one row of variables for each observation on a page. This report is shown in Example
7 on page 754.
proc print data=empdata(obs=12) rows=page;
id idnumber;
title Personnel Data;
run;

Output: Listing

Output 34.9

Default Layout for a Report with Many Variables: Listing Output

In the traditional procedure output, each page of this report contains

values for all variables in each observation. In the HTML output, this
report is identical to the report that uses ROWS=PAGE.
Note that PROC PRINT automatically splits the variable names that
are used as column headers at a change in capitalization if the entire
name does not t in the column. Compare, for example, the column
headers for LastName (which ts in the column) and FirstName (which
does not t in the column).

Personnel Data
Id
Number
1919
1653
1400
1350
1401
1499
1101
Id
Number
1919
1653
1400
1350
1401
1499
1101

LastName

First
Name

Adams
Alexander
Apple
Arthur
Avery
Barefoot
Baucom

Gerald
Susan
Troy
Barbara
Jerry
Joseph
Walter

Job
Code
TA2
ME2
ME1
FA3
TA3
ME3
SCP

City
Stamford
Bridgeport
New York
New York
Paterson
Princeton
New York

Salary

Birth

Hired

34376
35108
29769
32886
38822
43025
18723

15SEP48
18OCT52
08NOV55
03SEP53
16DEC38
29APR42
09JUN50

07JUN75
12AUG78
19OCT78
01AUG78
20NOV73
10JUN68
04OCT78

State
CT
CT
NY
NY
NJ
NJ
NY

Gender
M
F
M
F
M
M
M

HomePhone
203/781-1255
203/675-7715
212/586-0808
718/383-1549
201/732-8787
201/812-5665
212/586-8060

The PRINT Procedure

Personnel Data
Id
Number
1333
1402
1479
1403
1739
Id
Number
1333
1402
1479
1403
1739

Output 34.10

LastName

First
Name

Blair
Blalock
Bostic
Bowden
Boyce

Justin
Ralph
Marie
Earl
Jonathan

Job
Code

City

State

Stamford
New York
New York
Bridgeport
New York

CT
NY
NY
CT
NY

Salary

Birth

Hired

88606
32615
38785
28072
66517

02APR49
20JAN51
25DEC56
31JAN57
28DEC52

13FEB69
05DEC78
08OCT77
24DEC79
30JAN79

PT2
TA2
TA3
ME1
PT1

Gender
M
M
F
M
M

HomePhone
203/781-1777
718/384-2849
718/384-8816
203/675-3434
212/587-1247

Layout Produced by the ROWS=PAGE Option: Listing Output

Each page of this report contains values for only some of the variables
in each observation. However, each page contains values for more
observations than the default report does.

Personnel Data
Id
Number
1919
1653
1400
1350
1401
1499
1101
1333
1402
1479
1403
1739

LastName

First
Name

Adams
Alexander
Apple
Arthur
Avery
Barefoot
Baucom
Blair
Blalock
Bostic
Bowden
Boyce

Gerald
Susan
Troy
Barbara
Jerry
Joseph
Walter
Justin
Ralph
Marie
Earl
Jonathan

City

State

Stamford
Bridgeport
New York
New York
Paterson
Princeton
New York
Stamford
New York
New York
Bridgeport
New York

CT
CT
NY
NY
NJ
NJ
NY
CT
NY
NY
CT
NY

Gender
M
F
M
F
M
M
M
M
M
F
M
M

Output: Listing

757

758

Program: Creating an RTF Report

Chapter 34

Personnel Data
Id
Number
1919
1653
1400
1350
1401
1499
1101
1333
1402
1479
1403
1739

Job
Code
TA2
ME2
ME1
FA3
TA3
ME3
SCP
PT2
TA2
TA3
ME1
PT1

Salary

Birth

Hired

34376
35108
29769
32886
38822
43025
18723
88606
32615
38785
28072
66517

15SEP48
18OCT52
08NOV55
03SEP53
16DEC38
29APR42
09JUN50
02APR49
20JAN51
25DEC56
31JAN57
28DEC52

HomePhone

07JUN75
12AUG78
19OCT78
01AUG78
20NOV73
10JUN68
04OCT78
13FEB69
05DEC78
08OCT77
24DEC79
30JAN79

203/781-1255
203/675-7715
212/586-0808
718/383-1549
201/732-8787
201/812-5665
212/586-8060
203/781-1777
718/384-2849
718/384-8816
203/675-3434
212/587-1247

Program: Creating an RTF Report

options nodate pageno=1 linesize=64 pagesize=24;

data empdata;
input IdNumber $ 1-4 LastName $ 9-19 FirstName $ 20-29
City $ 30-42 State $ 43-44 /
Gender $ 1 JobCode $ 9-11 Salary 20-29 @30 Birth date9.
@43 Hired date9. HomePhone $ 54-65;
format birth hired date9.;
datalines;
1919
Adams
Gerald
Stamford
CT
M
TA2
34376
15SEP1948
07JUN1975
203/781-1255
1653
Alexander Susan
Bridgeport
CT
F
ME2
35108
18OCT1952
12AUG1978
203/675-7715
. . . more lines of data . . .
1407
M
1114
F
;

Grant
PT1
Green
TA2

Daniel
68096
Janice
32928

Mt. Vernon
26MAR1957
New York
21SEP1957

NY
21MAR1978

914/468-1616

30JUN1975

212/588-1092

The PRINT Procedure

Program: Creating an RTF Report with the STYLE Option

759

proc print data=empdata(obs=12);

id idnumber;
title Personnel Data;
run;

Close the RTF destination. The ODS RTF CLOSE statement closes the RTF destination.
ods rtf close;

Output: RTF
Display 34.12 Layout for a Report with Many Variables: RTF Output

Program: Creating an RTF Report with the STYLE Option

options nodate pageno=1 linesize=64 pagesize=24;

760

Output: RTF with Styles

1407
M
1114
F
;

Chapter 34

Grant
PT1
Green
TA2

Daniel
68096
Janice
32928

Mt. Vernon
26MAR1957
New York
21SEP1957

NY
21MAR1978

914/468-1616

30JUN1975

212/588-1092

ods rtf file=your_file.rtf;

proc print data=empdata(obs=12);

Create stylized output for Microsoft Word.

id idnumber / style(DATA)
{background =
style(HEADER)
{background =

=
red foreground = white}
=
blue foreground = white};

title Personnel Data;

run;

ods rtf close;

Output: RTF with Styles

Display 34.13 Layout for a Report with Many Variables: RTF Output Using Styles

The PRINT Procedure

Program: Creating a Listing Report

761

Example 8: Creating a Customized Layout with BY Groups and ID Variables

Procedure features:

BY statement
ID statement
SUM statement
VAR statement
Other features:

SORT procedure
Data set:

EMPDATA on page 755

This customized report

3 selects variables to include in the report and controls their order

3 selects observations to include in the report
3 groups the selected observations by JobCode
3 sums the salaries for each job code and for all job codes
3 displays numeric data with commas and dollar signs.

Program: Creating a Listing Report

Create and sort a temporary data set. PROC SORT creates a temporary data set in which
the observations are sorted by JobCode and Gender.
options nodate pageno=1 linesize=64 pagesize=60;
proc sort data=empdata out=tempemp;
by jobcode gender;
run;

Identify the character that starts a new line in column headers. SPLIT= identies the
asterisk as the character that starts a new line in column headers.
proc print data=tempemp split=*;

Specify the variables to include in the report. The VAR statement and the ID statement
together select the variables to include in the report. The ID statement and the BY statement
produce the special format.
id jobcode;
by jobcode;
var gender salary;

762

Program: Creating a Listing Report

Chapter 34

Calculate the total value for each BY group. The SUM statement totals the values of
Salary for each BY group and for the whole report.
sum salary;

Assign labels to the appropriate variables. The LABEL statement associates a label with
each variable for the duration of the PROC PRINT step. When you use SPLIT= in the PROC
PRINT statement, the procedure uses labels for column headings.
label jobcode=Job Code*========
gender=Gender*======
salary=Annual Salary*=============;

Create formatted columns. The FORMAT statement assigns a format to Salary for this
report. The WHERE statement selects for the report only the observations for job codes that
contain the letters FA or ME. The TITLE statements specify two titles.
format salary dollar11.2;
where jobcode contains FA or jobcode contains ME;
title Expenses Incurred for;
title2 Salaries for Flight Attendants and Mechanics;
run;

The PRINT Procedure

Output: Listing

Output 34.11 Creating a Customized Layout with BY Groups and ID Variables:

Listing Output
The ID and BY statements work together to produce this layout. The
ID variable is listed only once for each BY group. The BY lines are
suppressed. Instead, the value of the ID variable, JobCode, identies
each BY group.

Expenses Incurred for

Salaries for Flight Attendants and Mechanics
Job Code
========
FA1

Gender
======
F
F
M

$23,177.00
$22,454.00
$22,268.00
------------$67,899.00

F
F
M

$28,888.00
$27,787.00
$28,572.00
------------$85,247.00

F
F
M

$32,886.00
$33,419.00
$32,217.00
------------$98,522.00

M
M
M

$29,769.00
$28,072.00
$28,619.00
------------$86,460.00

F
F
M
M
M
M

$35,108.00
$34,929.00
$35,345.00
$36,925.00
$35,090.00
$35,185.00
------------$212,582.00

$43,025.00
=============
$593,735.00

-------FA1
FA2

-------FA2
FA3

-------FA3
ME1

-------ME1
ME2

-------ME2
ME3

Annual Salary
=============

Output: Listing

763

764

Program: Creating an HTML Report

Chapter 34

Program: Creating an HTML Report

options nodate pageno=1 linesize=64 pagesize=60 obs=15;
proc sort data=empdata out=tempemp;
by jobcode gender;
run;

proc print data=tempemp (obs=10) split=*;

id jobcode;
by jobcode;
var gender salary;

sum salary;

label jobcode=Job Code*========

gender=Gender*======
salary=Annual Salary*=============;

format salary dollar11.2;

where jobcode contains FA or jobcode contains ME;
title Expenses Incurred for;
title2 Salaries for Flight Attendants and Mechanics;
run;

Close the HTML destination. The ODS HTML CLOSE statement closes the HTML
destination.
ods html close;

The PRINT Procedure

Program: Creating an HTML Report with the STYLE Option

765

Output: HTML
Display 34.14 Creating a Customized Layout with BY Groups and ID Variables: Default HTML Output

Program: Creating an HTML Report with the STYLE Option

options nodate pageno=1 linesize=64 pagesize=60 obs=15;
proc sort data=empdata out=tempemp;
by jobcode gender;
run;

ods html file=your_file.html;

Create stylized HTML output. The rst STYLE option species that the font of the headers
be changed to italic. The second STYLE option species that the background of cells that
contain input data be changed to blue and the foreground of these cells be changed to white.
proc print data=tempemp (obs=10) split=* style(HEADER) =
{font_style=italic}
style(DATA) =
{background=blue foreground = white};

id jobcode;
by jobcode;
var gender salary;

766

Program: Creating an HTML Report with the STYLE Option

Chapter 34

Create total values that are written in red. The STYLE option species that the color of the
foreground of the cell that contain the totals be changed to red.
sum salary

/ style(total)= [foreground=red];

label jobcode=Job Code*========

gender=Gender*======
salary=Annual Salary*=============;

format salary dollar11.2;

where jobcode contains FA or jobcode contains ME;
title Expenses Incurred for;
title2 Salaries for Flight Attendants and Mechanics;
run;

ods html close;

The PRINT Procedure

Example 9: Printing All the Data Sets in a SAS Library

Output: HTML with Styles

Display 34.15 Creating a Customized Layout with BY Groups and ID Variables: HTML Output Using Styles

Example 9: Printing All the Data Sets in a SAS Library

Features:

Macro facility
DATASETS procedure
PRINT procedure
Data set:

EXPREV on page 724 and LIST

This example prints all the data sets in a SAS library. You can use the same
programming logic with any procedure. Just replace the PROC PRINT step near the
end of the example with whatever procedure step you want to execute. The example
uses the macro language. For details about the macro language, see SAS Guide to
Macro Processing, Version 6, Second Edition.

767

768

Program

Chapter 34

Program
libname printlib SAS-data-library
options nodate pageno=1 linesize=80 pagesize=60;

Copy the desired data sets from the WORK library to a permanent library. PROC
DATASETS copies two data sets from the WORK library to the PRINTLIB library in order to
limit the number of data sets available to the example.
proc datasets library=work memtype=data nolist;
copy out=printlib;
select list exprev;
run;

Create a macro and specify the parameters. The %MACRO statement creates the macro
PRINTALL. When you call the macro, you can pass one or two parameters to it. The rst
parameter is the name of the library whose data set you want to print. The second parameter is
a library used by the macro. If you do not specify this parameter, the WORK library is the
default.
%macro printall(libname,worklib=work);

Create the local macro variables. The %LOCAL statement creates two local macro variables,
NUM and I, to use in a loop.
%local num i;

The PRINTTO procedure denes destinations for SAS procedure output and for the
SAS log. By default, SAS procedure output and the SAS log are routed to the default
procedure output le and the default SAS log le for your method of operation. See
Table 35.1 on page 771. You can store the SAS log or procedure output in an external
le or in a SAS catalog entry. With additional programming, you can use SAS output as
input data within the same job.

Table 35.1

Default Destinations for SAS Log and Procedure Output

Method of running the SAS System

SAS log destination

Procedure output destination

windowing environment

the LOG window

the OUTPUT window

interactive line mode

the display monitor (as

statements are entered)

the display monitor (as each step

executes)

noninteractive mode or batch mode

depends on the host operating

system

depends on the operating

environment

Operating Environment Information: For information and examples specic to your

operating system or environment, see the appropriate SAS Companion or technical
report. 4

772

Syntax: PRINTTO Procedure

Chapter 35

Syntax: PRINTTO Procedure

See:

PRINTTO Procedure in the documentation for your operating environment.

PROC PRINTTO <option(s)>;

PROC PRINTTO Statement

Tip: To reset the destination for the SAS log and procedure output to the default, use
the PROC PRINTTO statement without options.
Tip: To route the SAS log and procedure output to the same le, specify the same le
with both the LOG= and PRINT= options.
Restriction: To route SAS log and procedure output directly to a printer, you must use a
FILENAME statement with the PROC PRINTTO statement. See Example 4 on page
785.

PROC PRINTTO <option(s)>;

To do this

Use this option

provide a description for a SAS log or procedure output

stored in a SAS catalog entry

LABEL=

route the SAS log to a permanent external le or SAS

catalog entry

LOG=

combine the SAS log and procedure output into a single le

LOG= and PRINT= with same

destination

replace the le instead of appending to it

NEW

route procedure output to a permanent external le or SAS

catalog entry or printer.

PRINT=

Without Options
Using a PROC PRINTTO statement with no options
3 closes any les opened by a PROC PRINTTO statement
3 points both the SAS log and SAS procedure output to their default destinations.

Interaction: To close the appropriate le and to return only the SAS log or

procedure output to its default destination, use LOG=LOG or PRINT=PRINT.

Featured in: Example 1 on page 776 and Example 2 on page 779

The PRINTTO Procedure

PROC PRINTTO Statement

773

Options
LABEL=description

provides a description for a catalog entry that contains a SAS log or procedure output.
Range: 1 to 256 characters
Interaction: Use the LABEL= option only when you specify a catalog entry as the

value for the LOG= or the PRINT= option.

Featured in:

Example 2 on page 779

LOG=LOG | le-specication | SAS-catalog-entry

routes the SAS log to one of three locations:

LOG
routes the SAS log to its default destination.
le-specication
routes the SAS log to an external le. le-specication can be one of the following:
external-le
the name of an external le specied in quotation marks.
log-lename
is an unquoted alphanumeric text string. SAS creates a log that uses
log-lename.log as the log lename.
Operating Environment Information: For more information about using
log-lename, see the documentation for your operating environment. 4
leref
a leref previously assigned to an external le.
SAS-catalog-entry
routes the SAS log to a SAS catalog entry. By default, libref is SASUSER, catalog
is PROFILE, and type is LOG. Express SAS-catalog-entry in one of the following
ways:
libref.catalog.entry<.LOG>
a SAS catalog entry stored in the SAS data library and SAS catalog specied.
catalog.entry<.LOG>
a SAS catalog entry stored in the specied SAS catalog in the default SAS data
library SASUSER.
entry.LOG
a SAS catalog entry stored in the default SAS library and catalog:
SASUSER.PROFILE.
leref
a leref previously assigned to a SAS catalog entry. Search for "FILENAME,
CATALOG Access Method" in the SAS online documentation.
Default: LOG.

After routing the log to an external le or a catalog entry, you can specify LOG
to route the SAS log back to its default destination.

Tip:

When routing the SAS log, include a RUN statement in the PROC PRINTTO
statement. If you omit the RUN statement, the rst line of the following DATA or
PROC step is not routed to the new le. (This occurs because a statement does not
execute until a step boundary is crossed.)

Tip:

774

PROC PRINTTO Statement

Chapter 35

Interaction: The SAS log and procedure output cannot be routed to the same

catalog entry at the same time.

Interaction: The NEW option replaces the existing contents of a le with the new

log. Otherwise, the new log is appended to the le.

Interaction: To route the SAS log and procedure output to the same le, specify the

same le with both the LOG= and PRINT= options.

Interaction: When routing the log to a SAS catalog entry, you can use the LABEL

option to provide a description for the entry in the catalog directory.

Featured in:

Example 1 on page 776, Example 2 on page 779, and Example 3 on

page 782
NEW

clears any information that exists in a le and prepares the le to receive the SAS
log or procedure output.
Default: If you omit NEW, the new information is appended to the existing le.
Interaction: If you specify both LOG= and PRINT=, NEW applies to both.
Featured in:

Example 1 on page 776, Example 2 on page 779, and Example 3 on

page 782
PRINT= PRINT | le-specication | SAS-catalog-entry

routes procedure output to one of three locations:

PRINT
routes procedure output to its default destination. After routing it to an external
le or a catalog entry, you can specify PRINT to route subsequent procedure
output to its default destination.
le-specication
routes procedure output to an external le. It is one of the following:
external-le
the name of an external le specied in quotation marks.
print-lename
is an unquoted alphanumeric text string. SAS creates a print le that uses
print-lename as the print lename.
Operating Environment Information: For more information about using
print-lename, see the documentation for your operating environment. 4
leref
a leref previously assigned to an external le.
Operating Environment Information: See your operating environment
documentation for additional information about le-specication for the PRINT
option. 4
SAS-catalog-entry
routes procedure output to a SAS catalog entry. By default, libref is SASUSER,
catalog is PROFILE, and type is OUTPUT. Express SAS-catalog-entry in one of the
following ways:
libref.catalog.entry<.OUTPUT>
a SAS catalog entry stored in the SAS data library and SAS catalog specied.
catalog.entry<.OUTPUT>
a SAS catalog entry stored in the specied SAS catalog in the default SAS data
library SASUSER.

The PRINTTO Procedure

Routing SAS Log or Procedure Output Directly to a Printer

775

entry.OUTPUT
a SAS catalog entry stored in the default SAS library and catalog:
SASUSER.PROFILE.
leref
a leref previously assigned to a SAS catalog entry. Search for "FILENAME,
CATALOG Access Method" in the SAS online documentation.
Aliases:

FILE=, NAME=

Default: PRINT
Interaction: The procedure output and the SAS log cannot be routed to the same

catalog entry at the same time.

Interaction: The NEW option replaces the existing contents of a le with the new

procedure output. If you omit NEW, the new output is appended to the le.
Interaction: To route the SAS log and procedure output to the same le, specify the

same le with both the LOG= and PRINT= options.

Interaction: When routing procedure output to a SAS catalog entry, you can use

the LABEL option to provide a description for the entry in the catalog directory.
Featured in:

Example 3 on page 782

UNIT=nn

routes the output to the le identied by the leref FTnnF001, where nn is an

integer between 1 and 99.
Range: 1 to 99, integer only.

You can dene this leref yourself; however, some operating systems predene
certain lerefs in this form.

Tip:

Concepts: PRINTTO Procedure

Page Numbering
3 When the SAS system option NUMBER is in effect, there is a single
page-numbering sequence for all output in the current job or session. When
NONUMBER is in effect, output pages are not numbered.

3 You can specify the beginning page number for the output you are currently
producing by using the PAGENO= in an OPTIONS statement.

Routing SAS Log or Procedure Output Directly to a Printer

To route SAS log or procedure output directly to a printer, use a FILENAME
statement to associate a leref with the printer name, and then use that leref in the
LOG= or PRINT= option. For an example, see Example 4 on page 785.
For more information see the FILENAME statement in SAS Language Reference:
Dictionary.
Operating Environment Information: For examples of printer names, see the
documentation for your operating system. 4

776

Examples: PRINTTO Procedure

Chapter 35

Examples: PRINTTO Procedure

Example 1: Routing to External Files

Procedure features:

PRINTTO statement:
Without options
Options:
LOG=
NEW
PRINT=

This example uses PROC PRINTTO to route the log and procedure output to an
external le and then reset both destinations to the default.

Program

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. The SOURCE
option writes lines of source code to the default destination for the SAS log.
options nodate pageno=1 linesize=80 pagesize=60 source;

Route the SAS log to an external le. PROC PRINTTO uses the LOG= option to route the
SAS log to an external le. By default, this log is appended to the current contents of log-file.
proc printto log=log-file;
run;

The PRINTTO Procedure

Log

777

Create the NUMBERS data set. The DATA step uses list input to create the NUMBERS data
set.
data numbers;
input x y z;
datalines;
14.2
25.2
96.8
10.8
51.6
96.8
9.5
34.2 138.2
8.8
27.6
83.2
11.5
49.4 287.0
6.3
42.0 170.7
;

Route the procedure output to an external le. PROC PRINTTO routes output to an
external le. Because NEW is specied, any output written to output-file will overwrite the
les current contents.
proc printto print=output-file new;
run;

Print the NUMBERS data set. The PROC PRINT output is written to the specied external
le.
proc print data=numbers;
title Listing of NUMBERS Data Set;
run;

Reset the SAS log and procedure output destinations to default. PROC PRINTTO routes
subsequent logs and procedure output to their default destinations and closes both of the
current les.
proc printto;
run;

Log

Output 35.1
1
2
3

Portion of Log Routed to the Default Destination

options nodate pageno=1 linesize=80 pagesize=60 source;
proc printto log=log-file;
run;

778

Output

Chapter 35

Output 35.2
5
6
7
8

Portion of Log Routed to an External File

data numbers;
input x y z;
datalines;

NOTE: The data set WORK.NUMBERS has 6 observations and 3 variables.

NOTE: DATA statement used:
real time
0.00 seconds
cpu time
0.00 seconds

15
16
16
17

;
proc printto print=output-file new;
run;

NOTE: PROCEDURE PRINTTO used:

real time
0.00 seconds
cpu time
0.00 seconds

18
19
20
21

proc print data=numbers;

title Listing of NUMBERS Data Set;
run;

NOTE: The PROCEDURE PRINT printed page 1.

NOTE: PROCEDURE PRINT used:
real time
0.00 seconds
cpu time
0.00 seconds

22
23
24

proc printto;
run;

Output

Output 35.3

Procedure Output Routed to an External File

Listing of NUMBERS Data Set
OBS
1
2
3
4
5
6

x
14.2
10.8
9.5
8.8
11.5
6.3

y
25.2
51.6
34.2
27.6
49.4
42.0

z
96.8
96.8
138.2
83.2
287.0
170.7

The PRINTTO Procedure

Program

779

Example 2: Routing to SAS Catalog Entries

Procedure features:

PRINTTO statement:
Without options
Options:
LABEL=
LOG=
NEW
PRINT=

This example uses PROC PRINTTO to route the SAS log and procedure output to a
SAS catalog entry and then to reset both destinations to the default.

Program

Assign a libname.
libname lib1 SAS-data-library;

Route the SAS log to a SAS catalog entry. PROC PRINTTO routes the SAS log to a SAS
catalog entry named SASUSER.PROFILE.TEST.LOG. The PRINTTO procedure uses the default
libref and catalog SASUSER.PROFILE because only the entry name and type are specied.
LABEL= assigns a description for the catalog entry.
proc printto log=test.log label=Inventory program new;
run;

Create the LIB1.INVENTORY data set. The DATA step creates a permanent SAS data set.
data lib1.inventry;
length Dept $ 4 Item $ 6 Season $ 6 Year 4;
input dept item season year @@;
datalines;
3070 20410 spring 1996 3070 20411 spring 1997
3070 20412 spring 1997 3070 20413 spring 1997
3070 20414 spring 1996 3070 20416 spring 1995

780

Program

Chapter 35

3071
3071
3071
3071
;

20500
20502
20505
20507

spring
spring
spring
spring

1994
1996
1994
1994

3071
3071
3071
3071

20501
20503
20506
20424

spring
spring
spring
spring

1995
1996
1994
1994

Route the procedure output to a SAS catalog entry. PROC PRINTTO routes procedure
output from the subsequent PROC REPORT step to the SAS catalog entry
LIB1.CAT1.INVENTRY.OUTPUT. LABEL= assigns a description for the catalog entry.
proc printto print=lib1.cat1.inventry.output
label=Inventory program new;
run;
proc report data=lib1.inventry nowindows headskip;
column dept item season year;
title Current Inventory Listing;
run;

Reset the SAS log and procedure output back to the default and close the le. PROC
PRINTTO closes the current les that were opened by the previous PROC PRINTTO step and
reroutes subsequent SAS logs and procedure output to their default destinations.
proc printto;
run;

The PRINTTO Procedure

Log

Output 35.4

SAS Log Routed to SAS Catalog Entry SASUSER.PROFILE.TEST.LOG.

You can view this catalog entry in the BUILD window of the SAS Explorer.

8
9
10
11
12

data lib1.inventry;
length Dept $ 4 Item $ 6 Season $ 6 Year 4;
input dept item season year @@;
datalines;

NOTE: SAS went to a new line when INPUT statement reached past the end of a
line.
NOTE: The data set LIB1.INVENTRY has 14 observations and 4 variables.
NOTE: DATA statement used:
real time
0.00 seconds
cpu time
0.00 seconds

20
21
22
23
24

;
proc printto print=lib1.cat1.inventry.output
label=Inventory program new;
run;

NOTE: PROCEDURE PRINTTO used:

real time
0.00 seconds
cpu time
0.00 seconds

25
26
27
28
29

proc report data=lib1.inventry nowindows headskip;

column dept item season year;
title Current Inventory Listing;
run;

NOTE: PROCEDURE REPORT used:

real time
0.00 seconds
cpu time
0.00 seconds

30
31
32

proc printto;
run;

Log

781

782

Output

Chapter 35

Output

Output 35.5

Procedure Output Routed to SAS Catalog Entry LIB1.CAT1.INVENTRY.OUTPUT.

You can view this catalog entry in the BUILD window of the SAS Explorer.

Current Inventory Listing

Dept

Item

Season

Year

3070
3070
3070
3070
3070
3070
3071
3071
3071
3071
3071
3071
3071
3071

20410
20411
20412
20413
20414
20416
20500
20501
20502
20503
20505
20506
20507
20424

spring
spring
spring
spring
spring
spring
spring
spring
spring
spring
spring
spring
spring
spring

1996
1997
1997
1997
1996
1995
1994
1995
1996
1996
1994
1994
1994
1994

Example 3: Using Procedure Output as an Input File

Procedure features:

PRINTTO statement:
Without options
Options:
LOG=
NEW
PRINT=

This example uses PROC PRINTTO to route procedure output to an external le and
then uses that le as input to a DATA step.

Generate random values for the variables. The DATA step uses the RANUNI function to
randomly generate values for the variables X and Y in the data set A.
data test;
do n=1 to 1000;
x=int(ranuni(77777)*7);
y=int(ranuni(77777)*5);
output;
end;
run;

The PRINTTO Procedure

Example 3: Using Procedure Output as an Input File

783

Assign a leref and route procedure output to the le that is referenced. The
FILENAME statement assigns a leref to an external le. PROC PRINTTO routes subsequent
procedure output to the le that is referenced by the leref ROUTED. See Output 35.6.
filename routed output-filename;

proc printto print=routed new;

run;

Produce the frequency counts. PROC FREQ computes frequency counts and a chi-square
analysis of the variables X and Y in the data set TEST. This output is routed to the le that is
referenced as ROUTED.
proc freq data=test;
tables x*y / chisq;
run;

Close the le. You must use another PROC PRINTTO to close the le that is referenced by
leref ROUTED so that the following DATA step can read it. The step also routes subsequent
procedure output to the default destination. PRINT= causes the step to affect only procedure
output, not the SAS log.
proc printto print=print;
run;

Create the data set PROBTEST. The DATA step uses ROUTED, the le containing PROC
FREQ output, as an input le and creates the data set PROBTEST. This DATA step reads all
records in ROUTED but creates an observation only from a record that begins with Chi-Squa.
data probtest;
infile routed;
input word1 $ @;
if word1=Chi-Squa then
do;
input df chisq prob;
keep chisq prob;
output;
end;
run;

Print the PROBTEST data set. PROC PRINT produces a simple listing of data set
PROBTEST. This output is routed to the default destination. See Output 35.7.
proc print data=probtest;
title Chi-Square Analysis for Table of X by Y;
run;

784

Example 3: Using Procedure Output as an Input File

Output 35.6

Chapter 35

PROC FREQ Output Routed to the External File Referenced as ROUTED

The FREQ Procedure
Table of x by y
x

Frequency|
Percent |
Row Pct |
Col Pct |
0|
1|
2|
3|
4| Total
---------+--------+--------+--------+--------+--------+
0 |
29 |
33 |
12 |
25 |
27 |
126
|
2.90 |
3.30 |
1.20 |
2.50 |
2.70 | 12.60
| 23.02 | 26.19 |
9.52 | 19.84 | 21.43 |
| 15.18 | 16.18 |
6.25 | 11.74 | 13.50 |
---------+--------+--------+--------+--------+--------+
1 |
23 |
26 |
29 |
20 |
19 |
117
|
2.30 |
2.60 |
2.90 |
2.00 |
1.90 | 11.70
| 19.66 | 22.22 | 24.79 | 17.09 | 16.24 |
| 12.04 | 12.75 | 15.10 |
9.39 |
9.50 |
---------+--------+--------+--------+--------+--------+
2 |
28 |
26 |
32 |
30 |
25 |
141
|
2.80 |
2.60 |
3.20 |
3.00 |
2.50 | 14.10
| 19.86 | 18.44 | 22.70 | 21.28 | 17.73 |
| 14.66 | 12.75 | 16.67 | 14.08 | 12.50 |
---------+--------+--------+--------+--------+--------+
3 |
26 |
24 |
36 |
32 |
45 |
163
|
2.60 |
2.40 |
3.60 |
3.20 |
4.50 | 16.30
| 15.95 | 14.72 | 22.09 | 19.63 | 27.61 |
| 13.61 | 11.76 | 18.75 | 15.02 | 22.50 |
---------+--------+--------+--------+--------+--------+
4 |
25 |
31 |
28 |
36 |
29 |
149
|
2.50 |
3.10 |
2.80 |
3.60 |
2.90 | 14.90
| 16.78 | 20.81 | 18.79 | 24.16 | 19.46 |
| 13.09 | 15.20 | 14.58 | 16.90 | 14.50 |
---------+--------+--------+--------+--------+--------+
5 |
32 |
29 |
26 |
33 |
27 |
147
|
3.20 |
2.90 |
2.60 |
3.30 |
2.70 | 14.70
| 21.77 | 19.73 | 17.69 | 22.45 | 18.37 |
| 16.75 | 14.22 | 13.54 | 15.49 | 13.50 |
---------+--------+--------+--------+--------+--------+
6 |
28 |
35 |
29 |
37 |
28 |
157
|
2.80 |
3.50 |
2.90 |
3.70 |
2.80 | 15.70
| 17.83 | 22.29 | 18.47 | 23.57 | 17.83 |
| 14.66 | 17.16 | 15.10 | 17.37 | 14.00 |
---------+--------+--------+--------+--------+--------+
Total
191
204
192
213
200
1000
19.10
20.40
19.20
21.30
20.00
100.00
2
The FREQ Procedure
Statistics for Table of x by y
Statistic
DF
Value
Prob
-----------------------------------------------------Chi-Square
24
27.2971
0.2908
Likelihood Ratio Chi-Square
24
28.1830
0.2524
Mantel-Haenszel Chi-Square
1
0.6149
0.4330
Phi Coefficient
0.1652
Contingency Coefficient
0.1630
Cramers V
0.0826
Sample Size = 1000

The PRINTTO Procedure

Output 35.7

Program

785

PROC PRINT Output of Data Set PROBTEST, Routed to Default Destination

Chi-Square Analysis for Table of X by Y
Obs
1

chisq

prob

27.297

0.291

Example 4: Routing to a Printer

Procedure features:

PRINTTO statement:
Option:
PRINT= option

This example uses PROC PRINTTO to route procedure output directly to a printer.

Program
Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page.
options nodate pageno=1 linesize=80 pagesize=60;

Associate a leref with the printer name. The FILENAME statement associates a leref
with the printer name that you specify. If you want to associate a leref with the default printer,
omit printer-name.
filename your_fileref printer printer-name;

Specify the le to route to the printer. The PRINT= option species the le that PROC
PRINTTO routes to the printer.
proc printto print=your_fileref;
run;

786

787

CHAPTER

36
The PROTO Procedure
Information about the PROTO Procedure

787

Information about the PROTO Procedure

See: For documentation of the PROTO procedure, go to http://support.sas.com/
documentation/onlinedoc. Select Base SAS from the Product-Specic Documentation

list.

788

789

CHAPTER

37
The PRTDEF Procedure
Overview: PRTDEF Procedure 789
Syntax: PRTDEF Procedure
789
PROC PRTDEF Statement 789
Input Data Set: PRTDEF Procedure 791
Summary of Valid Variables 791
Required Variables 792
Optional Variables 793
Examples: PRTDEF Procedure 796
Example 1: Dening Multiple Printer Denitions 796
Example 2: Creating a Ghostview Printer in SASUSER to Preview PostScript Printer Output in
SASUSER 797
Example 3: Creating a Single Printer Denition That Is Available to All Users 798
Example 4: Adding, Modifying, and Deleting Printer Denitions 799
Example 5: Deleting a Single Printer Denition 801

Overview: PRTDEF Procedure

The PRTDEF procedure creates printer denitions in batch mode either for an
individual user or for all SAS users at your site. Your system administrator can create
printer denitions in the SAS registry and make these printers available to all SAS
users at your site by using PROC PRTDEF with the USESASHELP option. An
individual user can create personal printer denitions in the SAS registry by using
PROC PRTDEF.

Syntax: PRTDEF Procedure

PROC PRTDEF < option(s)>;

PROC PRTDEF Statement

PROC PRTDEF < option(s)>;

790

PROC PRTDEF Statement

Chapter 37

To do this

Use this option

Specify the input data set that contains the printer

attributes

DATA=

Specify that the default operation is to delete the printer

denitions from the registry

DELETE

Specify that the registry entries are being created for export
to a different host

FOREIGN

Specify that a list of printers that are created or replaced

will be written to the log

LIST

Specify that any printer name that already exists will be

modied by using the information in the printer attributes
data set

REPLACE

Specify whether the printer denitions are available to all

users or just the users running PROC PRTDEF

USESASHELP

Options
DATA=SAS-data-set

species the SAS input data set that contains the printer attributes.
Requirements: Printer attributes variables that must be specied are DEST,
DEVICE, MODEL, and NAME, except when the value of the variable OPCODE is
DELETE, in which case only the NAME variable is required.
DELETE

species that the default operation is to delete the printer denitions from the
registry.
Interaction: If both DELETE and REPLACE are specied, then DELETE is the
default operation.
Tip: If the user-dened printer denition is deleted, then the administrator-dened
printer may still appear if it exists in the SASHELP catalog.
FOREIGN

species that the registry entries are being created for export to a different host. As a
consequence, tests of any host-dependent items, such as the TRANTAB, are skipped.
LIST

species that a list of printers that are created or replaced will be written to the log.
REPLACE

species that the default operation is to modify existing printer denitions. Any
printer name that already exists will be modied by using the information in the
printer attributes data set. Any printer name that does not exist will be added.
Interaction: If both REPLACE and DELETE are specied, then a DELETE will be
performed.
USESASHELP

species that the printer denitions that are to be placed in the SASHELP library,
where they are available to all users.
If the USESASHELP option is not specied, then the printer denitions that are
placed in the current SASUSER library, where they are available to the local user
only.

The PRTDEF Procedure

Summary of Valid Variables

791

Restriction: To use the USESASHELP option, you must have permission to write

to the SASHELP catalog.

Operating Environment Information: You can create printer denitions with PROC
PRTDEF in the Windows operating environment. However, because Universal
Printing is turned off by default in Windows, these printer denitions do not appear
in the Print window.
If you want to use your printer denitions when Universal Printing is turned off,
then do one of the following:
3 specify the printer denition as part of the PRINTERPATH system option
3 from the Output Delivery System (ODS), issue the following code:
ODS PRINTER SAS PRINTER=myprinter;

where myprinter is the name of your printer denition.

Input Data Set: PRTDEF Procedure

Summary of Valid Variables

To create your printer denitions, you must create a SAS data set whose variables
contain the appropriate printer attributes. The following table lists and describes both
the required and the optional variables for this data set.
Variable Name

Variable Description

Required
DEST

Destination

DEVICE

Device

MODEL

Prototype

NAME

Printer name

BOTTOM

Default bottom margin

CHARSET

Default font character set

DESC

Description

FONTSIZE

Point size of the default font

HOSTOPT

Host options

LEFT

Default left margin

LRECL

Output buffer size

OPCODE

Operation code

PAPERIN

Paper source or input tray

PAPEROUT

Paper destination or output tray

PAPERSIZ

Paper size

Optional

792

Required Variables

Chapter 37

Variable Name

Variable Description

PAPERTYP

Paper type

PREVIEW

Preview

PROTOCOL

Protocol

RES

Default printer resolution

RIGHT

Default right margin

STYLE

Default font style

TOP

Default top margin

TRANTAB

Translation table

TYPEFACE

Default font

UNITS

CM or IN units

VIEWER

Viewer

WEIGHT

Default font weight

Required Variables
To create or modify a printer, you must supply the NAME, MODEL, DEVICE, and
DEST variables. All the other variables use default values from the printer prototype
that is specied by the MODEL variable.
To delete a printer, specify only the required NAME variable.
The following variables are required in the input data set:
DEST

species the output destination for the printer.

Operating Environment Information:
some devices. 4

DEST is case sensitive for

Restriction: DEST is limited to 1023 characters.

DEVICE

species the type of I/O device to use when sending output to the
printer. Valid devices are listed in the Printer Denition wizard and
in the SAS Registry Editor.
Restriction: DEVICE is limited to 31 characters.

MODEL

species the printer prototype to use when dening the printer.

For a valid list of prototypes or model descriptions, you can look in
the SAS Registry Editor under CORE\PRINTING\PROTOTYPES.
Tip: While in interactive mode, you can invoke the registry with
the REGEDIT command.
Tip: While in interactive mode, you can invoke the Print Setup
dialog (DMPRTSETUP) and press New to view the list that is
specied in the second window of the Printer Denition wizard.
Restriction: MODEL is limited to 127 characters.

NAME

species the printer denition name that will be associated with the
rest of the attributes in the printer denition.
The name is unique within a given registry. If a new printer
denition contains a name that already exists, then the record will

The PRTDEF Procedure

Optional Variables

793

not be processed unless the REPLACE option has been specied or

unless the value of the OPCODE variable is Modify.
Restriction: NAME must have the following features:

3 It is limited to 127 characters.

3 It must have at least one nonblank character.
3 It cannot contain a backslash.
Note: Leading and trailing blanks will be stripped from the
name. 4

Optional Variables
The following variables are optional in the input data set:
BOTTOM
species the default bottom margin in the units that are specied by the UNITS
variable.
CHARSET
species the default font character set.
Restriction: The value must be one of the character set names in the typeface
that is specied by the TYPEFACE variable.
Restriction: CHARSET is limited to 31 characters.
DESC
species the description of the printer.
Restriction: The description can have a maximum of 1023 characters.
Default: DESC defaults to the prototype that is used to create the printer.
FONTSIZE
species the point size of the default font.
HOSTOPT
species any host options for the output destination. The host options are not case
sensitive.
Restriction: The host options can have a maximum of 1023 characters.
LEFT
species the default left margin in the units that are specied by the UNITS
variable.
LRECL
species the buffer size or record length to use when sending output to the printer.
Default: If LRECL is less than zero when modifying an existing printer, the
printers buffer size will be reset to that specied by the printer prototype.
OPCODE
is a character variable that species what action (Add, Delete, or Modify) to
perform on the printer denition.
Add
creates a new printer denition in the registry. If the REPLACE option has
been specied, then this operation will also modify an existing printer
denition.
Delete
removes an existing printer denition from the registry.

794

Optional Variables

Chapter 37

Restriction: This operation requires only the NAME variable to be dened.

The other variables are ignored.

Modify
changes an existing printer denition in the registry or adds a new one.
If a user modies and saves new attributes on a printer in the SASHELP
library, then these modications are stored in the SASUSER library. Values
that are specied by the user will override values that are set by the
administrator, but they will not replace them.

Tip:

Restriction: OPTCODE is limited to 8 characters.

PAPERIN
species the default paper source or input tray.
Restriction: The value of PAPERIN must be one of the paper source names in
the printer prototype that is specied by the MODEL variable.
Restriction: PAPERIN is limited to 31 characters.

PAPEROUT
species the default paper destination or output tray.
Restriction: The value of PAPEROUT must be one of the paper destination
names in the printer prototype that is specied by the MODEL variable.
Restriction: PAPEROUT is limited to 31 characters.

PAPERSIZ
species the default paper source or input tray.
Restriction: The value of PAPERSIZ must be one of the paper size names listed
in the printer prototype that is specied by the MODEL variable.
Restriction: PAPERSIZ is limited to 31 characters.

PAPERTYP
species the default paper type.
Restriction: The value of PAPERTYP must be one of the paper source names
listed in the printer prototype that is specied by the MODEL variable.
Restriction: PAPERTYP is limited to 31 characters.

PREVIEW
species the printer application to use for print preview.
Restriction: PREVIEW is limited to 127 characters.

PROTOCOL
species the I/O protocol to use when sending output to the printer.
Operating Environment Information: On mainframe systems, the protocol
describes how to convert the output to a format that can be processed by a protocol
converter that connects the mainframe to an ASCII device. 4
Restriction: PROTOCOL is limited to 31 characters.

RES
species the default printer resolution.
Restriction: The value of RES must be one of the resolution values available to
the printer prototype that is specied by the MODEL variable.
Restriction: RES is limited to 31 characters.

RIGHT
species the default right margin in the units that are specied by the UNITS
variable.

The PRTDEF Procedure

Optional Variables

795

STYLE
species the default font style.
Restriction: The value of STYLE must be one of the styles available to the

typeface that is specied by the TYPEFACE variable.

Restriction: STYLE is limited to 31 characters.

TOP
species the default top margin in the units that are specied by the UNITS
variable.
TRANTAB
species which translation table to use when sending output to the printer.
Operating Environment Information: The translation table is needed when an
EBCDIC host sends data to an ASCII device. 4
Restriction: TRANTAB is limited to 8 characters.

TYPEFACE
species the typeface of the default font.
Restriction: The typeface must be one of the typeface names available to the

printer prototype that is specied by the MODEL variable.

Restriction: TYPEFACE is limited to 63 characters.

UNITS
species the units CM or IN that are used by margin variables.
VIEWER
species the host system command that is to be used during print previews. As a
result, PROC PRTDEF causes a preview printer to be created.
Preview printers are specialized printers that are used to display printer output
on the screen before printing.
The values of the PREVIEW, PROTOCOL, DEST, and HOSTOPT variables
are ignored when a value for VIEWER has been specied. Place %s where the
input lename would normally be in the viewer command. The %s can be used
as many times as needed.

Tip:

Restriction: VIEWER is limited to 127 characters.

WEIGHT
species the default font weight.
Restriction: The value must be one of the valid weights for the typeface that is

specied by the TYPEFACE variable.

796

Examples: PRTDEF Procedure

Chapter 37

Examples: PRTDEF Procedure

Example 1: Dening Multiple Printer Denitions

Procedure features:

PROC PRTDEF statement options:

DATA=
USESASHELP

This example shows you how to set up various printers.

Program

Create the PRINTERS data set. The INPUT statement contains the names of the four
required variables. Each data line contains the information that is needed to produce a single
printer denition.
data printers;
input name $ 1-14 model $ 16-42 device $ 46-53 dest $ 57-70;
datalines;
Myprinter
PostScript Level 1 (Color)
PRINTER
printer1
Laserjet
PCL 5
PIPE
lp -dprinter5
Color LaserJet PostScript Level 2 (Color)
PIPE
lp -dprinter2
;

Specify the input data set that contains the printer attributes, create the printer
denitions, and make the denitions available to all users. The DATA= option species
PRINTERS as the input data set that contains the printer attributes.
PROC PRTDEF creates the printer denitions for the SAS registry, and the USESASHELP
option species that the printer denitions will be available to all users.
proc prtdef data=printers usesashelp;
run;

The PRTDEF Procedure

Program

797

Example 2: Creating a Ghostview Printer in SASUSER to Preview PostScript

Printer Output in SASUSER
Procedure features:

PROC PRTDEF statement options:

DATA=
LIST
REPLACE

This example creates a Ghostview printer denition in the SASUSER library for
previewing PostScript output.

Program

Create the GSVIEW data set, and specify the printer name, printer description,
printer prototype, and commands to be used for print preview. The GSVIEW data set
contains the variables whose values contain the information that is needed to produce the
printer denitions.
The NAME variable species the printer name that will be associated with the rest of the
attributes in the printer denition data record.
The DESC variable species the description of the printer.
The MODEL variable species the printer prototype to use when dening this printer.
The VIEWER variable species the host system commands to be used for print preview.
GSVIEW must be installed on your system and the value for VIEWER must include the path
to nd it. You must enclose the value in single quotation marks because of the %s. If you use
double quotation marks, SAS will assume that %s is a macro variable.
DEVICE and DEST are required variables, but no value is needed in this example. Therefore,
a dummy or blank value should be assigned.
data gsview;
name = "Ghostview";
desc = "Print Preview with Ghostview";
model= "PostScript Level 2 (Color)";
viewer = ghostview %s;
device = "Dummy";
dest = " ";

798

Example 3: Creating a Single Printer Denition That Is Available to All Users

Chapter 37

Specify the input data set that contains the printer attributes, create the printer
denitions, write the printer denitions to the SAS log, and replace a printer
denition in the SAS registry. The DATA= option species GSVIEW as the input data set
that contains the printer attributes.
PROC PRTDEF creates the printer denitions.
The LIST option species that a list of printers that are created or replaced will be written to
the SAS log.
The REPLACE option species that a printer denition will replace a printer denition in the
registry if the name of the printer denition matches a name already in the registry. If the
printer denition names do not match, then the new printer denition is added to the registry.
proc prtdef data=gsview list replace;
run;

Example 3: Creating a Single Printer Denition That Is Available to All Users

Procedure features:

PROC PRTDEF statement option:

DATA=

This example creates a denition for a Tektronix Phaser 780 printer with a
Ghostview print previewer with the following specications:

3 bottom margin set to 1 inch

3 font size set to 14 point
3 paper size set to A4.

The PRTDEF Procedure

Example 4: Adding, Modifying, and Deleting Printer Denitions

799

Program
Create the TEK780 data set and supply appropriate information for the printer
destination. The TEK780 data set contains the variables whose values contain the information
that is needed to produce the printer denitions.
In the example, assignment statements are used to assign these variables.
The NAME variable species the printer name that will be associated with the rest of the
attributes in the printer denition data record.
The DESC variable species the description of the printer.
The MODEL variable species the printer prototype to use when dening this printer.
The DEVICE variable species the type of I/O device to use when sending output to the printer.
The DEST variable species the output destination for the printer.
The PREVIEW variable species which printer will be used for print preview.
The UNITS variable species whether the margin variables are measured in centimeters or
inches.
The BOTTOM variable species the default bottom margin in the units that are specied by the
UNITS variable.
The FONTSIZE variable species the point size of the default font.
The PAPERSIZ variable species the default paper size.
data tek780;
name = "Tek780";
desc = "Test Lab Phaser 780P";
model = "Tek Phaser 780 Plus";
device = "PRINTER";
dest = "testlab3";
preview = "Ghostview";
units = "cm";
bottom = 2.5;
fontsize = 14;
papersiz = "ISO A4";
run;

Create the TEK780 printer denition. The DATA= option species TEK780 as the input
data set.
proc prtdef data=tek780;
run;

Example 4: Adding, Modifying, and Deleting Printer Denitions

Procedure features:

PROC PRTDEF statement options:

DATA=
LIST

800

Program

Chapter 37

This example
3 adds two printer denitions

3 modies a printer denition

3 deletes two printer denitions.

Program
Create the PRINTERS data set and specify which actions to perform on the printer
denitions. The PRINTERS data set contains the variables whose values contain the
information that is needed to produce the printer denitions.
The MODEL variable species the printer prototype to use when dening this printer.
The DEVICE variable species the type of I/O device to use when sending output to the printer.
The DEST variable species the output destination for the printer.
The OPCODE variable species which action (add, delete, or modify) to perform on the printer
denition.
The rst Add operation creates a new printer denition for Color PostScript in the SAS registry,
and the second Add operation creates a new printer denition for ColorPS in the SAS registry.
The Mod operation modies the existing printer denition for LaserJet 5 in the registry.
The Del operation deletes the printer denitions for Gray PostScript and test from the registry.
The & species that two or more blanks separate character values. This allows the name and
model value to contain blanks.
data printers;
length name
$ 80
model $ 80
device $ 8
dest
$ 80
opcode $ 3
;
input opcode $& name $&
datalines;
add Color PostScript
mod LaserJet 5
del Gray PostScript
del test
add ColorPS
;

model $& device $& dest $&;

PostScript
PCL 5
PostScript
PostScript
PostScript

Level 2 (Color)
Level 2 (Gray Scale)
Level 2 (Color)
Level 2 (Color)

DISK
DISK
DISK
DISK
DISK

sasprt.ps
sasprt.pcl
sasprt.ps
sasprt.ps
sasprt.ps

Create multiple printer denitions and write them to the SAS log. The DATA= option
species the input data set PRINTERS that contains the printer attributes. PROC PRTDEF
creates ve printer denitions, two of which have been deleted. The LIST option species that a
list of printers that are created or replaced will be written to the log.
proc prtdef data=printers library=sasuser list;
run;

The PRTDEF Procedure

Program

801

Example 5: Deleting a Single Printer Denition

Procedure features:

PROC PRTDEF statement option:

DELETE

This example shows you how to delete a printer from the registry.

Program
Create the DELETEPRT data set. The NAME variable contains the name of the printer to
delete.
data deleteprt;
name=printer1;
run;

Delete the printer denition from the registry and write the deleted printer to the log.
The DATA= option species DELETEPRT as the input data set.
PROC PRTDEF creates printer denitions for the SAS registry.
DELETE species that the printer is to be deleted.
LIST species to write the deleted printer to the log.
proc prtdef data=deleteprt delete list;
run;

802

Overview: PRTEXP Procedure

The PRTEXP procedure enables you to extract printer attributes from the SAS
registry for replication and modication. PROC PRTEXP then writes these attributes to
the SAS log or to a SAS data set. You can specify that PROC PRTEXP search for these
attributes in the SASHELP portion of the registry or the entire SAS registry.

Syntax: PRTEXP Procedure

Note: If neither the SELECT nor the EXCLUDE statement is used, then all of the

printers will be included in the output.

PROC PRTEXP< option(s)>;

<SELECT printer_1 <printer_n>>;
<EXCLUDE printer_1 <printer_n>>;

804

PROC PRTEXP Statement

Chapter 38

PROC PRTEXP Statement

PROC PRTEXP<option(s)>;

Options
USESASHELP

species that SAS search only the SASHELP portion of the registry for printer
denitions.
Default: The default is to search both the SASUSER and SASHELP portions of the
registry for printer denitions.
OUT=SAS-data-set

species the SAS data set to create that contains the printer denitions.
The data set that is specied by the OUT=SAS-data-set option is the same type of
data set that is specied by the DATA=SAS-data-set option in PROC PRTDEF to
dene each printer.
Default: If OUT=SAS-data-set is not specied, then the data that is needed to
dene each printer is written to the SAS log.

EXCLUDE Statement
The EXCLUDE statement will cause the output to contain information from all those printers that
are not listed.
EXCLUDE printer_1 <printer_n>;

Required Arguments
printer_1 printer_n

species the printer(s) that you do not want the output to contain information about.

SELECT Statement
The SELECT statement will cause the output to contain information from only those printers that
are listed.
SELECT printer_1 <printer_n>;

Required Arguments
printer_1 printer_n

species the printer(s) that you would like the output to contain information about.

The PRTEXP Procedure

Program

805

Concepts: PRTEXP Procedure

The PRTEXP procedure, along with the PRTDEF procedure, can replicate, modify,
and create printer denitions either for an individual user or for all SAS users at your
site. PROC PRTEXP can extract only the attributes that are used to create printer
denitions from the registry. If you write them to a SAS data set, then you can later
replicate and modify them. You can then use PROC PRTDEF to create the printer
denitions in the SAS registry from your input data set. For a complete discussion of
PROC PRTDEF and the variables and attributes that are used to create the printer
denitions, see Input Data Set: PRTDEF Procedure on page 791.

Examples: PRTEXP Procedure

Example 1: Writing Attributes to the SAS Log

Procedure Features:

PROC PRTEXP statement option:

SELECT statement
USESASHELP option

This example shows you how to write the attributes that are used to dene a printer
to the SAS log.

Program

Specify the printer that you want information about, specify that only the SASHELP
portion of the registry be searched, and write the information to the SAS log. The
SELECT statement species that you want the attribute information that is used to dene the
printer Postscript to be included in the output. The USESASHELP option species that only the
SASHELP registry is to be searched for Postscripts printer denitions. The data that is needed
to dene each printer is written to the SAS log because the OUT= option was not used to specify
a SAS data set.
proc prtexp usesashelp;
select postscript;
run;

806

Example 2: Writing Attributes to a SAS Data Set

Chapter 38

Example 2: Writing Attributes to a SAS Data Set

Procedure Features:

PROC PRTEXP statement option:

OUT= option
SELECT statement

This example shows you how to create a SAS data set that contains the data that
PROC PRTDEF would use to dene the printers PCL4, PCL5, PCL5E, and PCLC.

Program

Specify the printers that you want information about and create the PRDVTER data
set. The SELECT statement species the printers PCL4, PCL5, PCL5E, and PCLC. The OUT=
option creates the SAS data set PRDVTER, which contains the same attributes that are used by
PROC PRTDEF to dene the printers PCL4, PCL5, PCL5E, and PCLC. SAS will search both
the SASUSER and SASHELP registries, because USESASHELP was not specied.
proc prtexp out=PRDVTER;
select pcl4 pcl5 pcl5e pcl5c;
run;

See Also
Procedures
Chapter 37, The PRTDEF Procedure, on page 789

807

CHAPTER

39
The PWENCODE Procedure
Overview: PWENCODE Procedure 807
Syntax: PWENCODE Procedure 807
PROC PWENCODE Statement 807
Concepts: PWENCODE Procedure 808
Using Encoded Passwords in SAS Programs 808
Encoding versus Encryption 808
Examples: PWENCODE Procedure 809
Example 1: Encoding a Password 809
Example 2: Using an Encoded Password in a SAS Program 809
Example 3: Saving an Encoded Password to the Paste Buffer 811

Overview: PWENCODE Procedure

The PWENCODE procedure enables you to encode passwords. Encoded passwords
can be used in place of plain-text passwords in SAS programs that access relational
database management systems (RDBMSs), SAS/SHARE servers, and SAS Integrated
Object Model (IOM) servers (such as the SAS Metadata Server).

Syntax: PWENCODE Procedure

PROC PWENCODE IN=password <OUT=leref> <METHOD=encoding-method>;

PROC PWENCODE Statement

PROC PWENCODE IN=password <OUT=leref> <METHOD=encoding-method>;

808

Concepts: PWENCODE Procedure

Chapter 39

Required Argument
IN=password

species the password to encode. password can have no more than 512 characters.
password can contain letters, numerals, spaces, and special characters. If password
contains embedded single or double quotation marks, then use the standard SAS
rules for quoting character constants (see SAS Constants in Expressions in SAS
Language Reference: Concepts for details).
Featured in:

Example 1 on page 809, Example 2 on page 809, and Example 3 on

page 811

Options
OUT=leref

species a leref to which the output string is to be written. If the OUT= option is
not specied, then the output string is written to the SAS log.
Featured in:

Example 2 on page 809 and Example 3 on page 811

METHOD=encoding-method

species the encoding method to use. Currently, sas001 is the only supported
encoding method and is the default if the METHOD= option is omitted.

Concepts: PWENCODE Procedure

Using Encoded Passwords in SAS Programs

When a password is encoded with PROC PWENCODE, the output string includes a
tag that identies the string as having been encoded. An example of a tag is {sas001}.
The tag indicates the encoding method. SAS servers and SAS/ACCESS engines
recognize the tag and decode the string before using it. Encoding a password enables
you to write SAS programs without having to specify a password in plain text.
Note: SAS does not currently support encoded read, write, or alter passwords for
SAS data sets. 4

Encoding versus Encryption

PROC PWENCODE uses encoding to disguise passwords. With encoding, one
character set is translated to another character set through some form of table lookup.
Encryption, by contrast, involves the transformation of data from one form to another
through the use of mathematical operations and, usually, a key value. Encryption is
generally more difcult to break than encoding. PROC PWENCODE is intended to
prevent casual, non-malicious viewing of passwords. You should not depend on PROC
PWENCODE for all your data security needs; a determined and knowledgeable attacker
can decode the encoded passwords.

The PWENCODE Procedure

Example 2: Using an Encoded Password in a SAS Program

809

Examples: PWENCODE Procedure

Example 1: Encoding a Password

Procedure features:

IN= argument

This example shows a simple case of encoding a password and writing the encoded
password to the SAS log.

Program

Encode the password.

proc pwencode in=my password;
run;

Log
Output 39.1
6
7

proc pwencode in=my password;

run;

{sas001}bXkgcGFzc3dvcmQ=
NOTE: PROCEDURE PWENCODE used (Total process time):
real time
0.31 seconds
cpu time
0.08 seconds

Example 2: Using an Encoded Password in a SAS Program

Procedure features:

IN= argument
OUT= option

This example
3 encodes a password and saves it to an external le
3 reads the encoded password with a DATA step, stores it in a macro variable, and
uses it in a SAS/ACCESS LIBNAME statement.

810

Program 1: Encoding the Password

Chapter 39

Program 1: Encoding the Password

Declare a leref.
filename pwfile external-filename

Encode the password and write it to the external le. The OUT= option species which
external leref the encoded password will be written to.
proc pwencode in=mypass1 out=pwfile;
run;

Program 2: Using the Encoded Password

Declare a leref for the encoded-password le.

filename pwfile external-filename;

Set the SYMBOLGEN SAS system option. The purpose of this step is to show that the
actual password cannot be revealed, even when the macro variable that contains the encoded
password is resolved in the SAS log. This step is not required in order for the program to work
properly. For more information about the SYMBOLGEN SAS system option, see SAS Macro
Language: Reference.
options symbolgen;

Read the le and store the encoded password in a macro variable. The DATA step stores
the encoded password in the macro variable DBPASS. For details about the INFILE and INPUT
statements, the $VARYING. informat, and the CALL SYMPUT routine, see SAS Language
Reference: Dictionary.
data _null_;
infile pwfile obs=1 length=l;
input @;
input @1 line $varying1024. l;
call symput(dbpass,substr(line,1,l));
run;

Use the encoded password to access a DBMS. You must use double quotation marks ( ) so
that the macro variable resolves properly.
libname x odbc dsn=SQLServer user=testuser password="&dbpass";

The PWENCODE Procedure

Program

811

Log

28
29
30
31
32
33

data _null_;
infile pwfile obs=1 length=l;
input @;
input @1 line $varying1024. l;
call symput(dbpass,substr(line,1,l));
run;

NOTE: The infile PWFILE is:

File Name=external-filename,
RECFM=V,LRECL=256
NOTE: 1 record was read from the infile PWFILE.
The minimum record length was 20.
The maximum record length was 20.
NOTE: DATA statement used (Total process time):
real time
3.94 seconds
cpu time
0.03 seconds

34
libname x odbc
SYMBOLGEN: Macro variable DBPASS resolves to {sas001}bXlwYXNzMQ==
34 !
dsn=SQLServer user=testuser password="&dbpass";
NOTE: Libref X was successfully assigned as follows:
Engine:
ODBC
Physical Name: SQLServer

Example 3: Saving an Encoded Password to the Paste Buffer

Procedure features:

IN= argument
OUT= option
Other features:

FILENAME statement with CLIPBRD access method

This example saves an encoded password to the paste buffer. You can then paste the
encoded password into another SAS program or into the password eld of an
authentication dialog box.

Program

Declare a leref with the CLIPBRD access method. For more information about the
FILENAME statement with the CLIPBRD accedd method, see SAS Language Reference:
Dictionary.
filename clip clipbrd;

812

Program

Chapter 39

Encode the password and save it to the paste buffer. The OUT= option saves the encoded
password to the leref that was declared in the previous statement.
proc pwencode in=my password out=clip;
run;

813

CHAPTER

40
The RANK Procedure
Overview: RANK Procedure 813
What Does the RANK Procedure Do? 813
Ranking Data 814
Syntax: RANK Procedure 815
PROC RANK Statement 816
BY Statement 818
RANKS Statement 819
VAR Statement 820
Concepts: RANK Procedure 820
Computer Resources 820
Statistical Applications 820
Results: RANK Procedure 821
Missing Values 821
Output Data Set 821
Examples: RANK Procedure 822
Example 1: Ranking Values of Multiple Variables 822
Example 2: Ranking Values within BY Groups 823
Example 3: Partitioning Observations into Groups Based on Ranks
References 829

826

Overview: RANK Procedure

What Does the RANK Procedure Do?
The RANK procedure computes ranks for one or more numeric variables across the
observations of a SAS data set and outputs the ranks to a new SAS data set. PROC
RANK by itself produces no printed output.

814

Ranking Data

Chapter 40

Ranking Data
Output 40.1 shows the results of ranking the values of one variable with a simple
PROC RANK step. In this example, the new ranking variable shows the order of nish
of ve golfers over a four-day competition. The player with the lowest number of
strokes nishes in rst place. The following statements produce the output:
proc rank data=golf out=rankings;
var strokes;
ranks Finish;
run;
proc print data=rankings;
run;

Output 40.1

Assignment of the Lowest Rank Value to the Lowest Variable Value

The SAS System
Obs
1
2
3
4
5

Player
Jack
Jerry
Mike
Randy
Tito

Strokes

Finish

279
283
274
296
302

2
3
1
4
5

In Output 40.2, the candidates for city council are ranked by district according to
the number of votes that they received in the election and according to the number of
years that they have served in ofce.
This example shows how PROC RANK can
3 reverse the order of the rankings so that the highest value receives the rank of 1,
the next highest value receives the rank of 2, and so on
3 rank the observations separately by values of multiple variables

3 rank the observations within BY groups

3 handle tied values.
For an explanation of the program that produces this report, see Example 2 on page 823.

The RANK Procedure

Output 40.2

Syntax: RANK Procedure

815

Assignment of the Lowest Rank Value to the Highest Variable Value within Each BY Group
Results of City Council Election

---------------------------------- District=1 ----------------------------------

Obs
1
2
3
4

Candidate

Vote

Years

Cardella
Latham
Smith
Walker

1689
1005
1406
846

Vote
Rank

8
2
0
0

1
3
2
4

Years
Rank
1
2
3
3

N = 4

---------------------------------- District=2 ----------------------------------

Obs
5
6
7
8

Candidate

Vote

Years

Hinkley
Kreitemeyer
Lundell
Thrash

912
1198
2447
912

Vote
Rank

0
0
6
2

3
2
1
3

Years
Rank
3
3
1
2

N = 4

Syntax: RANK Procedure

Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See

Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.

PROC RANK <option(s)>;

BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
VAR data-set-variables(s);
RANKS new-variables(s);

To do this

Use this statement

Calculate a separate set of ranks for each BY group

Identify a variables that contain the ranks

RANKS

Specify the variables to rank

VAR

816

PROC RANK Statement

Chapter 40

PROC RANK Statement

PROC RANK <option(s)>;

To do this

Use this option

Specify the input data set

DATA=

Create an output data set

OUT=

Specify the ranking method

Compute fractional ranks

FRACTION or NPLUS1

Partition observations into groups

GROUPS=

Compute normal scores

NORMAL=

Compute percentages

PERCENT

Compute Savage scores

SAVAGE

Reverse the order of the rankings

DESCENDING

Specify how to rank tied values

TIES=

Note:

You can specify only one ranking method in a single PROC RANK step.

Options
DATA=SAS-data-set

species the input SAS data set.

Main discussion: Input Data Sets on page 19
Restriction: You cannot use PROC RANK with an engine that supports concurrent

access if another user is updating the data set at the same time.
DESCENDING

reverses the direction of the ranks. With DESCENDING, the largest value receives a
rank of 1, the next largest value receives a rank of 2, and so on. Otherwise, values
are ranked from smallest to largest.
Featured in:

Example 1 on page 822 and Example 2 on page 823

FRACTION

computes fractional ranks by dividing each rank by the number of observations

having nonmissing values of the ranking variable.
Alias:

Interaction: TIES=HIGH is the default with the FRACTION option. With

TIES=HIGH, fractional ranks are considered values of a right-continuous

empirical cumulative distribution function.

The RANK Procedure

PROC RANK Statement

817

assigns group values ranging from 0 to number-of-groups minus 1. Common

specications are GROUPS=100 for percentiles, GROUPS=10 for deciles, and
GROUPS=4 for quartiles. For example, GROUPS=4 partitions the original values
into four groups, with the smallest values receiving, by default, a quartile value of 0
and the largest values receiving a quartile value of 3.
The formula for calculating group values is

FLOOR (rank 3 k= (n + 1))

where FLOOR is the FLOOR function, rank is the values order rank, k is the value
of GROUPS=, and n is the number of observations having nonmissing values of the
ranking variable.
If the number of observations is evenly divisible by the number of groups, each
group has the same number of observations, provided there are no tied values at the
boundaries of the groups. Grouping observations by a variable that has many tied
values can result in unbalanced groups because PROC RANK always assigns
observations with the same value to the same group.
Tip:

Use DESCENDING to reverse the order of the group values.

Featured in:

Example 3 on page 826

NORMAL=BLOM | TUKEY | VW

computes normal scores from the ranks. The resulting variables appear normally
distributed. The formulas are
BLOM

yi=8 (ri3/8)/(n+1/4)

TUKEY

yi=8 (ri1/3)/(n+1/3)

yi=8 (ri)/(n+1)

where 8 is the inverse cumulative normal (PROBIT) function, ri is the rank of the
ith observation, and n is the number of nonmissing observations for the ranking
variable.
VW stands for van der Waerden. With NORMAL=VW, you can use the scores for a
nonparametric location test. All three normal scores are approximations to the exact
expected order statistics for the normal distribution, also called normal scores. The
BLOM version appears to t slightly better than the others (Blom 1958; Tukey 1962).
1

Interaction: If you specify the TIES= option, then PROC RANK computes the

normal score from the ranks based on non-tied values and applies the TIES=
specication to the resulting normal score.
NPLUS1

computes fractional ranks by dividing each rank by the denominator n+1, where n is
the number of observations having nonmissing values of the ranking variable.
Aliases:

FN1, N1

Interaction: TIES=HIGH is the default with the NPLUS1 option.

See also: FRACTION option
OUT=SAS-data-set

names the output data set. If SAS-data-set does not exist, PROC RANK creates it. If
you omit OUT=, the data set is named using the DATAn naming convention.

818

BY Statement

Chapter 40

PERCENT

divides each rank by the number of observations that have nonmissing values of the
variable and multiplies the result by 100 to get a percentage.
Alias:

Interaction: TIES=HIGH is the default with the PERCENT option.

You can use PERCENT to calculate cumulative percentages, but use

GROUPS=100 to compute percentiles.

Tip:

SAVAGE

computes Savage (or exponential) scores from the ranks by the following formula
(Lehman 1998):

3
2
X 1
501
=4
j

=n0r +1 j
i

TIES=HIGH | LOW | MEAN

species how to compute normal scores or ranks for tied data values.
HIGH
assigns the largest of the corresponding ranks (or largest of the normal scores
when NORMAL= is specied).
LOW
assigns the smallest of the corresponding ranks (or smallest of the normal scores
when NORMAL= is specied).
MEAN
assigns the mean of the corresponding rank (or mean of the normal scores when
NORMAL= is specied).
Default: MEAN (unless the FRACTION option or PERCENT option is in effect).
Interaction: If you specify the NORMAL= option, then the TIES= specication

applies to the normal score, not to the rank that is used to compute the normal
score.
Featured in:

Example 1 on page 822 and Example 2 on page 823

BY Statement
Produces a separate set of ranks for each BY group.
Main discussion: BY on page 58
Featured in:

Example 2 on page 823 and Example 3 on page 826

BY < DESCENDING> variable-1

<< DESCENDING> variable-n>
<NOTSORTED>;

The RANK Procedure

RANKS Statement

819

Required Arguments
variable

Options
DESCENDING

species that the observations are sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The observations are grouped in another way, such as chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the
NOTSORTED option. In fact, the procedure does not use an index if you specify
NOTSORTED. The procedure denes a BY group as a set of contiguous observations
that have the same values for all BY variables. If observations with the same values
for the BY variables are not contiguous, the procedure treats each contiguous set as a
separate BY group.

RANKS Statement
Creates new variables for the rank values.
Requirement:

If you use the RANKS statement, you must also use the VAR statement.

If you omit the RANKS statement, the rank values replace the original variable
values in the output data set.

Default:

Featured in: Example 1 on page 822 and Example 2 on page 823

RANKS new-variables(s);

Required Arguments
new-variable(s)

species one or more new variables that contain the ranks for the variable(s) listed in
the VAR statement. The rst variable listed in the RANKS statement contains the
ranks for the rst variable listed in the VAR statement, the second variable listed in
the RANKS statement contains the ranks for the second variable listed in the VAR
statement, and so forth.

820

VAR Statement

Chapter 40

VAR Statement
Species the input variables.
If you omit the VAR statement, PROC RANK computes ranks for all numeric
variables in the input data set.

Default:

Featured in:

Example 1 on page 822, Example 2 on page 823, and Example 3 on page 826

VAR data-set-variables(s);

Required Arguments
data-set-variable(s)

species one or more variables for which ranks are computed.

Using the VAR Statement with the RANKS Statement

The VAR statement is required when you use the RANKS statement. Using these
statements together creates the ranking variables named in the RANKS statement that
correspond to the input variables specied in the VAR statement. If you omit the
RANKS statement, the rank values replace the original values in the output data set.

Concepts: RANK Procedure

Computer Resources
PROC RANK stores all values in memory of the variables for which it computes
ranks.

Statistical Applications
Ranks are useful for investigating the distribution of values for a variable. The ranks
divided by n or n+1 form values in the range 0 to 1, and these values estimate the
cumulative distribution function. You can apply inverse cumulative distribution
functions to these fractional ranks to obtain probability quantile scores, which you can
compare to the original values to judge the t to the distribution. For example, if a set
of data has a normal distribution, the normal scores should be a linear function of the
original values, and a plot of scores versus original values should be a straight line.
Many nonparametric methods are based on analyzing ranks of a variable:

3 A two-sample t-test applied to the ranks is equivalent to a Wilcoxon rank sum test
using the t approximation for the signicance level. If you apply the t-test to the
normal scores rather than to the ranks, the test is equivalent to the van der
Waerden test. If you apply the t-test to median scores (GROUPS=2), the test is
equivalent to the median test.

The RANK Procedure

Output Data Set

821

3 A one-way analysis of variance applied to ranks is equivalent to the

Kruskal-Wallis k-sample test; the F-test generated by the parametric procedure
applied to the ranks is often better than the 2 approximation used by
Kruskal-Wallis. This test can be extended to other rank scores (Quade 1966).

3 You can obtain a Friedmans two-way analysis for block designs by ranking within
BY groups and then performing a main-effects analysis of variance on these ranks
(Conover 1998).

3 You can investigate regression relationships by using rank transformations with a

method described by Iman and Conover (1979).

Results: RANK Procedure

Missing Values
Missing values are not ranked and are left missing when ranks or rank scores
replace the original values in the output data set.

Output Data Set

The RANK procedure creates a SAS data set containing the ranks or rank scores but
does not create any printed output. You can use PROC PRINT, PROC REPORT, or
another SAS reporting tool to print the output data set.
The output data set contains all the variables from the input data set plus the
variables named in the RANKS statement. If you omit the RANKS statement, the rank
values replace the original variable values in the output data set.

822

Examples: RANK Procedure

Chapter 40

Examples: RANK Procedure

Example 1: Ranking Values of Multiple Variables

Procedure features:

PROC RANK statement options:

DESCENDING
TIES=
RANKS statement
VAR statement
Other features:

PRINT procedure

This example
3 reverses the order of the ranks so that the highest value receives the rank of 1
3 assigns tied values the best possible rank
3 creates ranking variables and prints them with the original variables.

Program

Set the SAS system options. The NODATE option species to omit the date and time when
the SAS job began. The PAGENO= option species the page number for the next page of output
that SAS produces. The LINESIZE= option species the line size. The PAGESIZE= option
species the number of lines for a page of SAS output.
options nodate pageno=1 linesize=80 pagesize=60;

Create the CAKE data set. This data set contains each participants last name, score for
presentation, and score for taste in a cake-baking contest.
data cake;
input Name
datalines;
Davis
77
Orlando
93
Ramey
68
Roe
68
Sanders
56
Simms
68
Strickland 82
;

$ 1-10 Present 12-13 Taste 15-16;

84
80
72
75
79
77
79

The RANK Procedure

Example 2: Ranking Values within BY Groups

823

Generate the ranks for the numeric variables in descending order and create the
output data set ORDER. DESCENDING reverses the order of the ranks so that the high
score receives the rank of 1. TIES=LOW gives tied values the best possible rank. OUT= creates
the output data set ORDER.
proc rank data=cake out=order descending ties=low;

Create two new variables that contain ranks. The VAR statement species the variables to
rank. The RANKS statement creates two new variables, PresentRank and TasteRank, that
contain the ranks for the variables Present and Taste, respectively.
var present taste;
ranks PresentRank TasteRank;
run;

Print the data set. PROC PRINT prints the ORDER data set. The TITLE statement species
a title.
proc print data=order;
title "Rankings of Participants Scores";
run;

Output
Rankings of Participants Scores

Obs
1
2
3
4
5
6
7

Name

Present

Davis
Orlando
Ramey
Roe
Sanders
Simms
Strickland

Taste

77
93
68
68
56
68
82

Example 2: Ranking Values within BY Groups

Procedure features:

PROC RANK statement options:

DESCENDING
TIES=
BY statement
RANKS statement
VAR statement

84
80
72
75
79
77
79

Present
Rank

Taste
Rank

3
1
4
4
7
4
2

1
2
7
6
3
5
3

824

Program

Chapter 40

Other features:

PRINT procedure

This example
3 ranks observations separately within BY groups
3 reverses the order of the ranks so that the highest value receives the rank of 1
3 assigns tied values the best possible rank
3 creates ranking variables and prints them with the original variables.

Program

Create the ELECT data set. This data set contains each candidates last name, district
number, vote total, and number of years experience on the city council.
data elect;
input Candidate
datalines;
Cardella
1 1689
Latham
1 1005
Smith
1 1406
Walker
1 846
Hinkley
2 912
Kreitemeyer 2 1198
Lundell
2 2447
Thrash
2 912
;

$ 1-11 District 13 Vote 15-18 Years 20;

8
2
0
0
0
0
6
2

Generate the ranks for the numeric variables in descending order and create the
output data set RESULTS. DESCENDING reverses the order of the ranks so that the highest
vote total receives the rank of 1. TIES=LOW gives tied values the best possible rank. OUT=
creates the output data set RESULTS.
proc rank data=elect out=results ties=low descending;

Create a separate set of ranks for each BY group. The BY statement separates the
rankings by values of District.
by district;

The RANK Procedure

Output

825

Create two new variables that contain ranks. The VAR statement species the variables to
rank. The RANKS statement creates the new variables, VoteRank and YearsRank, that contain
the ranks for the variables Vote and Years, respectively.
var vote years;
ranks VoteRank YearsRank;
run;

Print the data set. PROC PRINT prints the RESULTS data set. The N option prints the
number of observations in each BY group. The TITLE statement species a title.
proc print data=results n;
by district;
title Results of City Council Election;
run;

Output

In the second district, Hinkley and Thrash tied with 912 votes. They both receive a rank of 3
because TIES=LOW.

Results of City Council Election

---------------------------------- District=1 ----------------------------------

Obs
1
2
3
4

Candidate

Vote

Years

Cardella
Latham
Smith
Walker

1689
1005
1406
846

8
2
0
0

Vote
Rank
1
3
2
4

Years
Rank
1
2
3
3

N = 4

---------------------------------- District=2 ----------------------------------

Obs
5
6
7
8

Candidate

Vote

Years

Hinkley
Kreitemeyer
Lundell
Thrash

912
1198
2447
912

0
0
6
2

N = 4

Vote
Rank
3
2
1
3

Years
Rank
3
3
1
2

826

Example 3: Partitioning Observations into Groups Based on Ranks

Chapter 40

Example 3: Partitioning Observations into Groups Based on Ranks

Procedure features:

PROC RANK statement option:

GROUPS=
BY statement
VAR statement
Other features:

PRINT procedure
SORT procedure

This example

3 partitions observations into groups on the basis of values of two input variables
3 groups observations separately within BY groups
3 replaces the original variable values with the group values.

Program

Create the SWIM data set. This data set contains swimmers rst names and their times, in
seconds, for the backstroke and the freestyle. This example groups the swimmers into pairs,
within male and female classes, based on times for both strokes so that every swimmer is paired
with someone who has a similar time for each stroke.
data swim;
input Name $ 1-7 Gender $ 9 Back 11-14 Free 16-19;
datalines;
Andrea F 28.6 30.3
Carole F 32.9 24.0
Clayton M 27.0 21.9
Curtis M 29.0 22.6
Doug
M 27.3 22.4
Ellen
F 27.8 27.0
Jan
F 31.3 31.2
Jimmy
M 26.3 22.5
Karin
F 34.6 26.2
Mick
M 29.0 25.4
Richard M 29.7 30.2
Sam
M 27.2 24.1

The RANK Procedure

Susan
;

Program

827

F 35.1 36.1

Sort the SWIM data set and create the output data set PAIRS. PROC SORT sorts the
data set by Gender. This is required to obtain a separate set of ranks for each group. OUT=
creates the output data set PAIRS.
proc sort data=swim out=pairs;
by gender;
run;

Generate the ranks that are partitioned into three groups and create an output data
set. GROUPS=3 assigns one of three possible group values (0,1,2) to each swimmer for each
stroke. OUT= creates the output data set RANKPAIR.
proc rank data=pairs out=rankpair groups=3;

Create a separate set of ranks for each BY group. The BY statement separates the
rankings by Gender.
by gender;

Replace the original values of the variables with the rank values. The VAR statement
species that Back and Free are the variables to rank. With no RANKS statement, PROC
RANK replaces the original variable values with the group values in the output data set.
var back free;
run;

Print the data set. PROC PRINT prints the RANKPAIR data set. The N option prints the
number of observations in each BY group. The TITLE statement species a title.
proc print data=rankpair n;
by gender;
title Pairings of Swimmers for Backstroke and Freestyle;
run;

828

Output

Chapter 40

Output

The group values pair up swimmers with similar times to work on each stroke. For example,
Andrea and Ellen work together on the backstroke because they have the fastest times in the
female class. The groups of male swimmers are unbalanced because there are seven male
swimmers; for each stroke, one group has three swimmers.

Pairings of Swimmers for Backstroke and Freestyle

----------------------------------- Gender=F ----------------------------------Obs

1
2
3
4
5
6

Name

Back

Andrea
Carole
Ellen
Jan
Karin
Susan

0
1
0
1
2
2

Free
1
0
1
2
0
2

N = 6

----------------------------------- Gender=M ----------------------------------Obs

7
8
9
10
11
12
13

Name

Back

Clayton
Curtis
Doug
Jimmy
Mick
Richard
Sam
N = 7

0
2
1
0
2
2
1

Free
0
1
0
1
2
2
1

The RANK Procedure

References

829

References
Blom, G. (1958), Statistical Estimates and Transformed Beta Variables, New York:
John Wiley & Sons, Inc.
Conover, W.J. (1998), Practical Nonparametric Statistics, Third Edition, New York:
John Wiley & Sons, Inc.
Conover, W.J. and Iman, R.L. (1976), On Some Alternative Procedures Using Ranks
for the Analysis of Experimental Designs, Communications in Statistics, A5, 14,
13481368.
Conover, W.J. and Iman, R.L. (1981), Rank Transformations as a Bridge between
Parametric and Nonparametric Statistics, The American Statistician, 35, 124129.
Iman, R.L. and Conover, W.J. (1979), The Use of the Rank Transform in Regression,
Technometrics, 21, 499509.
Lehman, E.L. (1998), Nonparametrics: Statistical Methods Based on Ranks, New
Jersey: Prentice Hall .
Quade, D. (1966), On Analysis of Variance for the k-Sample Problem, Annals of
Mathematical Statistics, 37, 17471758.
Tukey, John W. (1962), The Future of Data Analysis, Annals of Mathematical
Statistics, 33, 22.

830

831

CHAPTER

41
The REGISTRY Procedure
Overview: REGISTRY Procedure 831
Syntax: REGISTRY Procedure 831
PROC REGISTRY Statement 832
Creating Registry Files with the REGISTRY Procedure 836
Structure of a Registry File 836
Specifying Key Names 836
Specifying Values for Keys 836
Sample Registry Entries 837
Examples: REGISTRY Procedure 839
Example 1: Importing a File to the Registry 839
Example 2: Listing and Exporting the Registry 840
Example 3: Comparing the Registry to an External File
Example 4: Comparing Registry Files 842

841

Overview: REGISTRY Procedure

The REGISTRY procedure maintains the SAS registry. The registry consists of two
parts. One part is stored in the SASHELP library, and the other part is stored in the
SASUSER library.
The REGISTRY procedure enables you to

3
3
3
3
3
3

import registry les to populate the SASHELP and SASUSER registries

export all or part of the registry to another le
list the contents of the registry in the SAS log
compare the contents of the registry to a le
uninstall a registry le
deliver detailed status information when a key or value will be overwritten or
uninstalled

3 clear out entries in the SASUSER registry

3 validate that the registry exists
3 list diagnostic information.

Syntax: REGISTRY Procedure

PROC REGISTRY < option(s)>;

832

PROC REGISTRY Statement

Chapter 41

PROC REGISTRY Statement

PROC REGISTRY < option(s)>;

To do this

Use this option

Erase the contents of the SASUSER registry

CLEARSASUSER

Compare two registry les

COMPAREREG1 and
COMPAREREG2

Compare the contents of a registry to a le

COMPARETO

Enable registry debugging

DEBUGON

Disable registry debugging

DEBUGOFF

Write the contents of a registry to the specied le

EXPORT=

Provide additional information in the SAS log about the

results of the IMPORT= and the UNINSTALL options

FULLSTATUS

Import the specied le to a registry

IMPORT=

Write the contents of the registry to the SAS log. Used with
the STARTAT= option to list specic keys.

LIST

Write the contents of the SASHELP portion of the registry

to the SAS log

LISTHELP

Send the contents of a registry to the log

LISTREG

Write the contents of the SASUSER portion of the registry

to the SAS log

LISTUSER

Start exporting or writing or comparing the contents of a

registry at the specied key

STARTAT=

Delete from the specied registry all the keys and values
that are in the specied le

UNINSTALL

Uppercase all incoming key names

UPCASE

Perform the specied operation on the SASHELP portion of

the SAS registry

USESASHELP

Options
CLEARSASUSER

erases the content of the SASUSER portion of the SAS registry.

COMPAREREG1=libname.registry-name-1

species one of two registries to compare. The results appear in the SAS log.

The REGISTRY Procedure

PROC REGISTRY Statement

833

libname
is the name of the library in which the registry le resides.
registry-name-1
is the name of the rst registry.
Requirement:

Must be used with COMPAREREG2.

Interaction: To specify a single key and all of its subkeys, specify the STARTAT=

option.
Featured in:

Example 4 on page 842

COMPAREREG2=libname.registry-name-2

species the second of two registries to compare. The results appear in the SAS log.
libname
is the name of the library in which the registry le resides.
registry-name-2
is the name of the second registry.
Requirement:
Featured in:

Must be used with COMPAREREG1.

Example 4 on page 842

COMPARETO=le-specication

compares the contents of a le that contains registry information to a registry. It

returns information about keys and values that it nds in the le that are not in the
registry. It reports as differences

3 keys that are dened in the external le but not in the registry
3 value names for a given key that are in the external le but not in the registry
3 differences in the content of like-named values in like-named keys.
COMPARETO= does not report as differences any keys and values that are in the
registry but not in the le because the registry could easily be composed of pieces
from many different les.
le-specication is one of the following:
external-le
is the path and name of an external le that contains the registry information.
leref
is a leref that has been assigned to an external le.
Requirement: You must have previously associated the leref with an external

le in a FILENAME statement, a FILENAME function, the Explorer window, or

an appropriate operating environment command.
Interaction: By default, PROC REGISTRY compares le-specication to the

SASUSER portion of the registry. To compare le-specication to the SASHELP

portion of the registry, specify the option USESASHELP.
Featured in:

Example 3 on page 841

See also: For information about how to structure a le that contains registry

information, see Creating Registry Files with the REGISTRY Procedure on page
836.
DEBUGON

enables registry debugging by providing more descriptive log entries.

DEBUGOFF

disables registry debugging.

834

PROC REGISTRY Statement

Chapter 41

EXPORT=le-specication

writes the contents of a registry to the specied le, where

le-specication is one of the following:
external-le
is the name of an external le that contains the registry information.
leref
is a leref that has been assigned to an external le.
Requirement: You must have previously associated the leref with an external

le in a FILENAME statement, a FILENAME function, the Explorer window, or

an appropriate operating environment command.
If le-specication already exists, then PROC REGISTRY overwrites it. Otherwise,
PROC REGISTRY creates the le.
Interaction: By default, EXPORT= writes the SASUSER portion of the registry to

the specied le. To write the SASHELP portion of the registry, specify the
USESASHELP option. You must have write permission to the SASHELP library
to use USESASHELP.
Interaction: To export a single key and all of its subkeys, specify the STARTAT=

option.
Featured in:

Example 2 on page 840

FULLSTATUS

lists the keys, subkeys, and values that were added or deleted as a result of running
the IMPORT= and the UNINSTALL options.
IMPORT=le-specication

species the le to import into the SAS registry. PROC REGISTRY does not
overwrite the existing registry. Instead, it updates the existing registry with the
contents of the specied le.
Note: .sasxreg le extension is not required.
le-specication is one of the following:

external-le
is the path and name of an external le that contains the registry information.
leref
is a leref that has been assigned to an external le.
Requirement: You must have previously associated the leref with an external

le in a FILENAME statement, a FILENAME function, the Explorer window, or

an appropriate operating environment command.
Interaction: By default, IMPORT= imports the le to the SASUSER portion of the

SAS registry. To import the le to the SASHELP portion of the registry, specify
the USESASHELP option. You must have write permission to SASHELP to use
USESASHELP.
Interaction: To obtain additional information in the SAS log as you import a le,

use FULLSTATUS.
Featured in:

Example 1 on page 839

See also: For information about how to structure a le that contains registry

information, see Creating Registry Files with the REGISTRY Procedure on page
836.
LIST

writes the contents of the entire SAS registry to the SAS log.

The REGISTRY Procedure

PROC REGISTRY Statement

835

Interaction: To write a single key and all of its subkeys, use the STARTAT= option.
LISTHELP

writes the contents of the SASHELP portion of the registry to the SAS log.
Interaction: To write a single key and all of its subkeys, use the STARTAT= option.
LISTREG=libname.registry-name

lists the contents of the specied registry in the log.

libname
is the name of the library in which the registry le resides.
registry-name
is the name of the registry.
Example:
proc registry listreg=sashelp.regstry;
run;

Interaction: To list a single key and all of its subkeys, use the STARTAT= option.
LISTUSER

writes the contents of the SASUSER portion of the registry to the SAS log.
Interaction: To write a single key and all of its subkeys, use the STARTAT= option.
Featured in:

Example 2 on page 840

STARTAT=key-name

exports or writes the contents of a single key and all of its subkeys.
Interaction: USE STARTAT= with the EXPORT=, LIST, LISTHELP, LISTUSER,

COMPAREREG1=, COMPAREREG2= and the LISTREG option.

Featured in:

Example 4 on page 842

UNINSTALL=le-specication

deletes from the specied registry all the keys and values that are in the specied le.
le-specication is one of the following:
external-le
is the name of an external le that contains the keys and values to delete.
leref
is a leref that has been assigned to an external le. To assign a leref you can

3 use the Explorer Window

3 use the FILENAME statement. (For information about the FILENAME
statement, see the section on statements in SAS Language Reference:
Dictionary.)
Interaction: By default, UNINSTALL deletes the keys and values from the

SASUSER portion of the SAS registry. To delete the keys and values from the
SASHELP portion of the registry, specify the USESASHELP option. You must
have write permission to SASHELP to use this option.
Interaction: Use FULLSTATUS to obtain additional information in the SAS log as

you uninstall a registry.

See also: For information about how to structure a le that contains registry

information, see Creating Registry Files with the REGISTRY Procedure on page
836.
UPCASE

uppercases all incoming key names.

836

Creating Registry Files with the REGISTRY Procedure

Chapter 41

USESASHELP

performs the specied operation on the SASHELP portion of the SAS registry.
Interaction: Use USESASHELP with the IMPORT=, EXPORT=, COMPARETO, or

UNINSTALL option. To use USESASHELP with IMPORT= or UNINSTALL, you

must have write permission to SASHELP.

Creating Registry Files with the REGISTRY Procedure

Structure of a Registry File

You can create registry les with the SAS Registry Editor or with any text editor.
A registry le must have a particular structure. Each entry in the registry le
consists of a key name, followed on the next line by one or more values. The key name
identies the key or subkey that you are dening. Any values that follow specify the
names or data to associate with the key.

Specifying Key Names

Key names are entered on a single line between square brackets ([ and ]). To specify
a subkey, enter multiple key names between the brackets, starting with the root key.
Separate the names in a sequence of key names with a backslash (\). The length of a
single key name or a sequence of key names cannot exceed 255 characters (including
the square brackets and the backslashes). Key names can contain any character except
the backslash.
Examples of valid key name sequences follow. These sequences are typical of the SAS
registry:
[CORE\EXPLORER\MENUS\ENTRIES\CLASS]
[CORE\EXPLORER\NEWMEMBER\CATALOG]
[CORE\EXPLORER\NEWENTRY\CLASS]
[CORE\EXPLORER\ICONS\ENTRIES\LOG]

Specifying Values for Keys

Enter each value on the line that follows the key name that it is associated with. You
can specify multiple values for each key, but each value must be on a separate line.
The general form of a value is
value-name=value-content
A value-name can be an at sign (@), which indicates the default value name, or it can
be any text string in double quotation marks. If the text string contains an ampersand
(&), then the character (either uppercase or lowercase) that follows the ampersand is a
shortcut for the value name. See Sample Registry Entries on page 837.
The entire text string cannot contain more than 255 characters (including quotation
marks and ampersands). It can contain any character except a backslash (\).

The REGISTRY Procedure

Sample Registry Entries

837

Value-content can be any of the following:

3 the string double: followed by a numeric value.
3 a string. You can put anything inside the quotes, including nothing ("").
Note: To include a backslash in the quoted string, use two adjacent backslashes.
To include a double quotation mark, use two adjacent double quotation marks.

3 the string hex: followed by any number of hexadecimal characters, up to the

255-character limit, separated by commas. If you extend the hexadecimal
characters beyond a single line, then end the line with a backslash to indicate that
the data continues on the next line. Hex values may also be referred to as binary
values in the Registry Editor.
3 the string dword: followed by an unsigned long hexadecimal value.
3 the string int: followed by a signed long integer value.
3 the string uint: followed by an unsigned long integer value.
The following display shows how the different types of values that are described above
appear in the Registry Editor:
Display 41.1

Types of Registry Values, Displayed in the Registry Editor

The following list contains a sample of valid registry values:

3 A double value=double:2.4E-44

3
3
3
3
3

A string="my data"
Binary data=hex: 01,00,76,63,62,6B
Dword=dword:00010203
Signed integer value=int:-123
Unsigned integer value (decimal)=dword:0001E240

Sample Registry Entries

Registry entries can vary in content and appearance, depending on their purpose.
The following display shows a registry entry that contains default PostScript printer
settings.

838

Sample Registry Entries

Display 41.2

Chapter 41

Portion of a Registry Editor Showing Settings for a PostScript Printer

To see what the actual registry text le looks like, you can use PROC REGISTRY to
write the contents of the registry key to the SAS log, using the LISTUSER and
STARTAT= options:
Example Code 41.1

SAS code for sending a SASUSER registry entry to the log

proc registry
listuser
startat=sasuser-registry-key-name;
run;

Example Code 41.2

SAS code for sending a SASUSER registry entry to the log

proc registry
listuser
startat=HKEY_SYSTEM_ROOT\CORE\PRINTING\PRINTERS\PostScript\DEFAULT SETTINGS;
run;

For example, the list below begins at the

CORE\PRINTING\PRINTERS\PostScript\DEFAULT SETTINGS key.

The REGISTRY Procedure

Output 41.1

Source File

Log Output of a Registry Entry for a PostScript Printer

NOTE: Contents of SASUSER REGISTRY starting at subkey [CORE\

PRINTING\PRINTERS\PostScript\DEFAULT SETTINGS key]
Font Character Set="Western"
Font Size=double:12
Font Style="Regular"
Font Typeface="Courier"
Font Weight="Normal"
Margin Bottom=double:0.5
Margin Left=double:0.5
Margin Right=double:0.5
Margin Top=double:0.5
Margin Units="IN"
Paper Destination=""
Paper Size="Letter"
Paper Source=""
Paper Type=""
Resolution="300 DPI"
NOTE: PROCEDURE REGISTRY used (Total process time):
real time
0.03 seconds
cpu time
0.03 seconds

Examples: REGISTRY Procedure

Example 1: Importing a File to the Registry

Procedure features:

IMPORT=

Other features: FILENAME statement

This example imports a le into the SASUSER portion of the SAS registry.

Source File
The following le contains examples of valid key name sequences in a registry le:
[HKEY_USER_ROOT\AllGoodPeopleComeToTheAidOfTheirCountry]
@="This is a string value"
"Value2"=""
"Value3"="C:\\This\\Is\\Another\\String\\Value"

839

840

Program

Chapter 41

Program

Assign a leref to a le that contains valid text for the registry. The FILENAME
statement assigns the leref SOURCE to the external le that contains the text to read into the
registry.
filename source external-file;

Invoke PROC REGISTRY to import the le that contains input for the registry. PROC
REGISTRY reads the input le that is identied by the leref SOURCE. IMPORT= writes to
the SASUSER portion of the SAS registry by default.
proc registry
run;

import=source;

SAS Log
1
filename source external-file;
2
proc registry
3
import=source;
4
run;
Parsing REG file and loading the registry please wait....
Registry IMPORT is now complete.

Example 2: Listing and Exporting the Registry

Procedure features:

EXPORT=
LISTUSER

This example lists the SASUSER portion of the SAS registry and exports it to an
external le.
Note: This is usually a very large le. To export a portion of the registry, use the
STARTAT= option. 4

Program

Write the contents of the SASUSER portion of the registry to the SAS log. The
LISTUSER option causes PROC REGISTRY to write the entire SASUSER portion of the
registry to the log.
proc registry
listuser

The REGISTRY Procedure

Program

841

Export the registry to the specied le. The EXPORT= option writes a copy of the
SASUSER portion of the SAS registry to the external le.
export=external-file;
run;

SAS Log
1 proc registry listuser export=external-file;
2 run;
Starting to write out the registry file, please wait...
The export to file external-file is now complete.
Contents of SASUSER REGISTRY.
[ HKEY_USER_ROOT]
[
CORE]
[
EXPLORER]
[
CONFIGURATION]
Initialized= "True"
[
FOLDERS]
[
UNXHOST1]
Closed= "658"
Icon= "658"
Name= "Home Directory"
Open= "658"
Path= "~"

Example 3: Comparing the Registry to an External File

COMPARETO= option
Other features: FILENAME statement
Procedure features:

This example compares the SASUSER portion of the SAS registry to an external le.
Comparisons such as this are useful if you want to know the difference between a
backup le that was saved with a .txt le extension and the current registry le.
Note: To compare the SASHELP portion of the registry with an external le, specify
the USESASHELP option. 4

Program

Assign a leref to the external le that contains the text to compare to the registry.
The FILENAME statement assigns the leref TESTREG to the external le.

filename testreg external-file;

842

SAS Log

Chapter 41

Compare the specied le to the SASUSER portion of the SAS registry. The
COMPARETO option compares the contents of a le to a registry. It returns information about
keys and values that it nds in the le that are not in the registry.
proc registry
compareto=testreg;
run;

SAS Log
This SAS log shows two differences between the SASUSER portion of the registry
and the specied external le. In the registry, the value of Initialized is True; in the
external le, it is False. In the registry, the value of Icon is 658; in the external le
it is 343.

1
filename testreg external-file;
2
proc registry
3
compareto=testreg;
4 run;
Parsing REG file and comparing the registry please wait....
COMPARE DIFF: Value "Initialized" in
[HKEY_USER_ROOT\CORE\EXPLORER\CONFIGURATION]: REGISTRY TYPE=STRING, CURRENT
VALUE="True"
COMPARE DIFF: Value "Initialized" in
[HKEY_USER_ROOT\CORE\EXPLORER\CONFIGURATION]: FILE TYPE=STRING, FILE
VALUE="False"
COMPARE DIFF: Value "Icon" in
[HKEY_USER_ROOT\CORE\EXPLORER\FOLDERS\UNXHOST1]: REGISTRY TYPE=STRING,
CURRENT VALUE="658"
COMPARE DIFF: Value "Icon" in
[HKEY_USER_ROOT\CORE\EXPLORER\FOLDERS\UNXHOST1]: FILE TYPE=STRING, FILE
VALUE="343"
Registry COMPARE is now complete.
COMPARE: There were differences between the registry and the file.

Example 4: Comparing Registry Files

Procedure features

COMPAREREG1= and COMPAREREG2= options

STARTAT= option

This example uses the REGISTRY procedure options COMPAREREG1= and

COMPAREREG2= to specify two registry les for comparison.

Program

Declare the PROCLIB library. The PROCLIB library contains a registry le.
libname proclib SAS-data-library;

The REGISTRY Procedure

SAS Log

843

Start PROC REGISTRY and specify the rst registry le to be used in the comparison.
proc registry comparereg1=sasuser.regstry

Limit the comparison to the registry keys including and following the specied
registry key. The STARTAT= option limits the scope of the comparison to the EXPLORER
subkey under the CORE key. By default the comparison includes the entire contents of both
registries.
startat=CORE\EXPLORER

Specify the second registry le to be used in the comparison.

comparereg2=proclib.regstry;
run;

SAS Log
8
proc registry comparereg1=sasuser.regstry
9
10
startat=CORE\EXPLORER
11
comparereg2=proclib.regstry;
12
run;
NOTE: Comparing registry SASUSER.REGSTRY to registry PROCLIB.REGSTRY
NOTE: Diff in Key (CORE\EXPLORER\MENUS\FILES\SAS) Item (1;&Open)
SASUSER.REGSTRY Type: String len 17 data PGM;INCLUDE %s;
PROCLIB.REGSTRY Type: String len 15 data WHOSTEDIT %s;
NOTE: Diff in Key (CORE\EXPLORER\MENUS\FILES\SAS) Item (3;&Submit)
SASUSER.REGSTRY Type: String len 23 data PGM;INCLUDE %s;SUBMIT
PROCLIB.REGSTRY Type: String len 21 data WHOSTEDIT %s;SUBMIT
NOTE: Diff in Key (CORE\EXPLORER\MENUS\FILES\SAS) Item (4;&Remote Submit)
SASUSER.REGSTRY Type: String len 35 data SIGNCHECK;PGM;INCLUDE %s;RSUBMIT;
PROCLIB.REGSTRY Type: String len 33 data SIGNCHECK;WHOSTEDIT %s;RSUBMIT;
NOTE: Diff in Key (CORE\EXPLORER\MENUS\FILES\SAS) Item (@)
SASUSER.REGSTRY Type: String len 17 data PGM;INCLUDE %s;
PROCLIB.REGSTRY Type: String len 15 data WHOSTEDIT %s;
NOTE: Item (2;Open with &Program Editor) in key
(CORE\EXPLORER\MENUS\FILES\TXT) not found in registry PROCLIB.REGSTRY
NOTE: Diff in Key (CORE\EXPLORER\MENUS\FILES\TXT) Item (4;&Submit)
SASUSER.REGSTRY Type: String len 24 data PGM;INCLUDE %s;SUBMIT;
PROCLIB.REGSTRY Type: String len 22 data WHOSTEDIT %s;SUBMIT;
NOTE: Diff in Key (CORE\EXPLORER\MENUS\FILES\TXT) Item (5;&Remote Submit)
SASUSER.REGSTRY Type: String len 35 data SIGNCHECK;PGM;INCLUDE %s;RSUBMIT;
PROCLIB.REGSTRY Type: String len 33 data SIGNCHECK;WHOSTEDIT %s;RSUBMIT;
NOTE: PROCEDURE REGISTRY used (Total process time):
real time
0.07 seconds
cpu time
0.02 seconds

844

Syntax: REPORT Procedure 869

PROC REPORT Statement 870
BREAK Statement 885
BY Statement 889
CALL DEFINE Statement 890
COLUMN Statement 893
COMPUTE Statement 895
DEFINE Statement 897
ENDCOMP Statement 906
FREQ Statement 906
LINE Statement 907
RBREAK Statement 908
WEIGHT Statement 912
REPORT Procedure Windows 912
BREAK 913
COMPUTE 916
COMPUTED VAR 916
DATA COLUMNS 917
DATA SELECTION 917
DEFINITION 918
DISPLAY PAGE 923
EXPLORE 924
FORMATS 925
LOAD REPORT 925
MESSAGES 926
PROFILE 926
PROMPTER 927
REPORT 928
ROPTIONS 928
SAVE DATA SET 933
SAVE DEFINITION 933
SOURCE 934
STATISTICS 934
WHERE 935
WHERE ALSO 935
How PROC REPORT Builds a Report 936
Sequence of Events 936
Construction of Summary Lines 937
Report-Building Examples 937
Building a Report That Uses Groups and a Report Summary 937
Building a Report That Uses Temporary Variables 941
Examples: REPORT Procedure 948
Example 1: Selecting Variables for a Report 948
Example 2: Ordering the Rows in a Report 951
Example 3: Using Aliases to Obtain Multiple Statistics for the Same Variable 954
Example 4: Consolidating Multiple Observations into One Row of a Report 957
Example 5: Creating a Column for Each Value of a Variable 960
Example 6: Displaying Multiple Statistics for One Variable 964
Example 7: Storing and Reusing a Report Denition 966
Example 8: Condensing a Report into Multiple Panels 968
Example 9: Writing a Customized Summary on Each Page 971
Example 10: Calculating Percentages 975
Example 11: How PROC REPORT Handles Missing Values 977
Example 12: Creating and Processing an Output Data Set 980

The REPORT Procedure

Example
Example
Example
Example

13:
14:
15:
16:

What Do the Various Types of Reports Look Like?

Storing Computed Variables as Part of a Data Set 983

Using a Format to Create Groups 986
Specifying Style Elements for ODS Output in the PROC REPORT Statement
Specifying Style Elements for ODS Output in Multiple Statements 994

847

989

Overview: REPORT Procedure

What Does the REPORT Procedure Do?
The REPORT procedure combines features of the PRINT, MEANS, and TABULATE
procedures with features of the DATA step in a single report-writing tool that can
produce a variety of reports. You can use PROC REPORT in three ways:
3 in a windowing environment with a prompting facility that guides you as you
build a report.
3 in a windowing environment without the prompting facility.
3 in a nonwindowing environment. In this case, you submit a series of statements
with the PROC REPORT statement, just as you do in other SAS procedures. You
can submit these statements from the Program Editor with the NOWINDOWS
option in the PROC REPORT statement, or you can run SAS in batch,
noninteractive, or interactive line mode (see the information about running SAS in
SAS Language Reference: Concepts).
This documentation provides reference information about using PROC REPORT in a
windowing or nonwindowing environment. For task-oriented documentation for the
nonwindowing environment, see SAS Technical Report P-258, Using the REPORT
Procedure in a Nonwindowing Environment, Release 6.07.

What Types of Reports Can PROC REPORT Produce?

A detail report contains one row for every observation selected for the report. Each of
these rows is a detail row. A summary report consolidates data so that each row
represents multiple observations. Each of these rows is also called a detail row.
Both detail and summary reports can contain summary lines as well as detail rows.
A summary line summarizes numerical data for a set of detail rows or for all detail
rows. PROC REPORT provides both default and customized summaries (see Using
Break Lines on page 861).
This overview illustrates the kinds of reports that PROC REPORT can produce. The
statements that create the data sets and formats used in these reports are in Example
1 on page 948. The formats are stored in a permanent SAS data library. See
Examples: REPORT Procedure on page 948 for more reports and for the statements
that create them.

What Do the Various Types of Reports Look Like?

The data set that these reports use contains one days sales gures for eight stores in
a chain of grocery stores.
A simple PROC REPORT step produces a report similar to one produced by a simple
PROC PRINT step. Figure 42.1 on page 848 illustrates the simplest kind of report that
you can produce with PROC REPORT. The statements that produce the report follow.
The data set and formats that the program uses are created in Example 1 on page 948.

848

What Do the Various Types of Reports Look Like?

Chapter 42

Although the WHERE and FORMAT statements are not essential, here they limit the
amount of output and make the values easier to understand.
libname proclib SAS-data-library;

options nodate pageno=1 linesize=64 pagesize=60

fmtsearch=(proclib);
proc report data=grocery nowd;
where sector=se;
format sector $sctrfmt. manager $mgrfmt.
dept $deptfmt. sales dollar10.2;
run;

Figure 42.1

Simple Detail Report with a Detail Row for Each Observation

Detail row
The SAS System

Sector
Southeast
Southeast
Southeast
Southeast
Southeast
Southeast
Southeast
Southeast

Manager
Smith
Smith
Smith
Smith
Jones
Jones
Jones
Jones

Department
Paper
Meat/Dairy
Canned
Produce
Paper
Meat/Dairy
Canned
Produce

1
Sales
$50.00
$100.00
$120.00
$80.00
$40.00
$300.00
$220.00
$70.00

The report in Figure 42.2 on page 849 uses the same observations as those in Figure
42.1 on page 848. However, the statements that produce this report

3 order the rows by the values of Manager and Department

3 create a default summary line for each value of Manager
3 create a customized summary line for the whole report. A customized summary
lets you control the content and appearance of the summary information, but you
must write additional PROC REPORT statements to create one.
For an explanation of the program that produces this report, see Example 2 on page
951.

The REPORT Procedure

Figure 42.2

What Do the Various Types of Reports Look Like?

849

Ordered Detail Report with Default and Customized Summaries

Detail row
Sales for the Southeast Sector

Manager
Department
Sales
----------------------------------Jones

Paper
Canned
Meat/Dairy
Produce

$40.00
$220.00
$300.00
$70.00
------$630.00

Paper
Canned
Meat/Dairy
Produce

$50.00
$120.00
$100.00
$80.00
------$350.00

------Jones
Smith

------Smith

Total sales for these stores were: $980.00

Customized summary
line for the whole report

Default summary
line for Manager

The summary report in Figure 42.3 on page 849 contains one row for each store in
the northern sector. Each detail row represents four observations in the input data set,
one observation for each department. Information about individual departments does
not appear in this report. Instead, the value of Sales in each detail row is the sum of
the values of Sales in all four departments. In addition to consolidating multiple
observations into one row of the report, the statements that create this report

3 customize the text of the column headers

3 create default summary lines that total the sales for each sector of the city
3 create a customized summary line that totals the sales for both sectors.
For an explanation of the program that produces this report, see Example 4 on page
957.

Figure 42.3

Summary Report with Default and Customized Summaries

Default summary
line for Sector

Detail row

Sales Figures for Northern Sectors

Sector
---------

Manager
-------

Sales
----------

Northeast

Alomar
Andrews

786.00
1,045.00
---------$1,831.00

Northwest

Brown
Pelfrey
Reveiz

598.00
746.00
1,110.00
---------$2,454.00

Combined sales for the northern sectors were $4,285.00.

Customized summary
line for the whole report

850

What Do the Various Types of Reports Look Like?

Chapter 42

The summary report in Figure 42.4 on page 850 is similar to Figure 42.3 on page
849. The major difference is that it also includes information for individual
departments. Each selected value of Department forms a column in the report. In
addition, the statements that create this report
3 compute and display a variable that is not in the input data set

3 double-space the report

3 put blank lines in some of the column headers.
For an explanation of the program that produces this report, see Example 5 on page 960.

Figure 42.4

Summary Report with a Column for Each Value of a Variable

Computed variable

Sales Figures for Perishables in Northern Sectors

Sector

Manager

______Department_______
Meat/Dairy
Produce

Perishable
Total
-------------------------------------------------------Northeast Alomar
Andrews
Northwest Brown

$190.00

$86.00

$276.00

$300.00

$125.00

$425.00

$250.00

$73.00

$323.00

Pelfrey

$205.00

$76.00

$281.00

Reveiz

$600.00

$30.00

$630.00

--------------------------------------------------| Combined sales for meat and dairy : $1,545.00 |

| Combined sales for produce :
$390.00 |
|
|
| Combined sales for all perishables: $1,935.00 |
---------------------------------------------------

Customized summary lines

for the whole report

The customized report in Figure 42.5 on page 851 shows each managers store on a
separate page. Only the rst two pages appear here. The statements that create this
report create

3 a customized header for each page of the report

3 a computed variable (Prot) that is not in the input data set
3 a customized summary with text that is dependent on the total sales for that
managers store.
For an explanation of the program that produces this report, see Example 9 on page
971.

The REPORT Procedure

Figure 42.5

What Do the Various Types of Reports Look Like?

851

Customized Summary Report

Detail row

Computed variable
Sales for Individual Stores

Northeast Sector
Store managed by Alomar

Department
Sales
Profit
----------------------------------Canned
Meat/Dairy
Paper
Produce

$420.00
$190.00
$90.00
$86.00
--------$786.00

$168.00
$47.50
$36.00
$21.50
--------$196.50

Sales are in the target region.

Customized summary line
for Manager

Default summary line

for Manager

Detail row

Computed variable
Sales for Individual Stores

Northeast Sector
Store managed by Andrews

Department
Sales
Profit
----------------------------------Canned
Meat/Dairy
Paper
Produce

$420.00
$300.00
$200.00
$125.00
--------$1,045.00

$168.00
$75.00
$80.00
$31.25
--------$261.25

Sales exceeded goal!

Customized summary line
for Manager

Default summary line

for Manager

The report in Figure 42.6 on page 852 uses customized style elements to control
things like font faces, font sizes, and justication, as well as the width of the border of
the table and the width of the spacing between cells. This report was created by using
the HTML destination of the Output Delivery System (ODS) and the STYLE= option in
several statements in the procedure.
For an explanation of the program that produces this report, see Example 16 on page
994. For information on ODS, see Output Delivery System on page 32.

852

Concepts: REPORT Procedure

Figure 42.6

Chapter 42

HTML Output

Concepts: REPORT Procedure

Laying Out a Report

Planning the Layout

Report writing is simplied if you approach it with a clear understanding of what
you want the report to look like. The most important thing to determine is the layout of
the report. To design the layout, ask yourself the following kinds of questions:

3
3
3
3

What do I want to display in each column of the report?

In what order do I want the columns to appear?
Do I want to display a column for each value of a particular variable?
Do I want a row for every observation in the report, or do I want to consolidate
information for multiple observations into one row?

3 In what order do I want the rows to appear?

The REPORT Procedure

Laying Out a Report

853

When you understand the layout of the report, use the COLUMN and DEFINE
statements in PROC REPORT to construct the layout.
The COLUMN statement lists the items that appear in the columns of the report,
describes the arrangement of the columns, and denes headers that span multiple
columns. A report item can be

3 a data set variable

3 a statistic calculated by the procedure
3 a variable that you compute from other items in the report.
Omit the COLUMN statement if you want to include all variables in the input data
set in the same order as they occur in the data set.
Note: If you start PROC REPORT in the windowing environment without the
COLUMN statement, then the initial report includes only as many variables as will t
on one page. 4
The DEFINE statement (or, in the windowing environment, the DEFINITION
window) denes the characteristics of an item in the report. These characteristics
include how PROC REPORT uses the item in the report, the text of the column header,
and the format to use to display values.

Usage of Variables in a Report

Much of a reports layout is determined by the usages that you specify for variables
in the DEFINE statements or DEFINITION windows. For data set variables, these
usages are
DISPLAY
ORDER
ACROSS
GROUP
ANALYSIS
A report can contain variables that are not in the input data set. These variables
must have a usage of COMPUTED.

Display Variables
A report that contains one or more display variables has a row for every observation
in the input data set. Display variables do not affect the order of the rows in the report.
If no order variables appear to the left of a display variable, then the order of the rows
in the report reects the order of the observations in the data set. By default, PROC
REPORT treats all character variables as display variables.
Featured in: Example 1 on page 948

Order Variables
A report that contains one or more order variables has a row for every observation in
the input data set. If no display variable appears to the left of an order variable, then
PROC REPORT orders the detail rows according to the ascending, formatted values of
the order variable. You can change the default order with ORDER= and DESCENDING
in the DEFINE statement or with the DEFINITION window.
If the report contains multiple order variables, then PROC REPORT establishes the
order of the detail rows by sorting these variables from left to right in the report. PROC

854

Laying Out a Report

Chapter 42

REPORT does not repeat the value of an order variable from one row to the next if the
value does not change, unless an order variable to its left changes values.
Featured in:

Example 2 on page 951

Across Variables
PROC REPORT creates a column for each value of an across variable. PROC
REPORT orders the columns by the ascending, formatted values of the across variable.
You can change the default order with ORDER= and DESCENDING in the DEFINE
statement or with the DEFINITION window. If no other variable helps dene the
column (see COLUMN Statement on page 893), then PROC REPORT displays the N
statistic (the number of observations in the input data set that belong to that cell of the
report).
If you are familiar with procedures that use class variables, then you will see that
across variables are class variables that are used in the column dimension.
Featured in:

Example 5 on page 960

Group Variables
If a report contains one or more group variables, then PROC REPORT tries to
consolidate into one row all observations from the data set that have a unique
combination of formatted values for all group variables.
When PROC REPORT creates groups, it orders the detail rows by the ascending,
formatted values of the group variable. You can change the default order with ORDER=
and DESCENDING in the DEFINE statement or with the DEFINITION window.
If the report contains multiple group variables, then the REPORT procedure
establishes the order of the detail rows by sorting these variables from left to right in the
report. PROC REPORT does not repeat the values of a group variable from one row to
the next if the value does not change, unless a group variable to its left changes values.
If you are familiar with procedures that use class variables, then you will see that
group variables are class variables that are used in the row dimension.
Note: You cannot always create groups. PROC REPORT cannot consolidate
observations into groups if the report contains any order variables or any display
variables that do not have one or more statistics associated with them (see COLUMN
Statement on page 893). In the windowing environment, if PROC REPORT cannot
immediately create groups, then the procedure changes all display and order variables
to group variables so that it can create the group variable that you requested. In the
nonwindowing environment, it returns to the SAS log a message that explains why it
could not create groups. Instead, it creates a detail report that displays group variables
the same way as it displays order variables. Even when PROC REPORT creates a
detail report, the variables that you dene as group variables retain that usage in their
denitions. 4
Featured in:

Example 4 on page 957

The REPORT Procedure

Laying Out a Report

855

Analysis Variables
An analysis variable is a numeric variable that is used to calculate a statistic for all
the observations represented by a cell of the report. (Across variables, in combination
with group variables or order variables, determine which observations a cell
represents.) You associate a statistic with an analysis variable in the variables
denition or in the COLUMN statement. By default, PROC REPORT uses numeric
variables as analysis variables that are used to calculate the Sum statistic.
The value of an analysis variable depends on where it appears in the report:
3 In a detail report, the value of an analysis variable in a detail row is the value of
the statistic associated with that variable calculated for a single observation.
Calculating a statistic for a single observation is not practical; however, using the
variable as an analysis variable enables you to create summary lines for sets of
observations or for all observations.
3 In a summary report, the value displayed for an analysis variable is the value of
the statistic that you specify calculated for the set of observations represented by
that cell of the report.
3 In a summary line for any report, the value of an analysis variable is the value of
the statistic that you specify calculated for all observations represented by that
cell of the summary line.

See also:

BREAK Statement on page 885 and RBREAK Statement on page

908
Featured in: Example 2 on page 951, Example 3 on page 954, Example 4 on

page 957, and Example 5 on page 960

Note: Be careful when you use SAS dates in reports that contain summary lines.
SAS dates are numeric variables. Unless you explicitly dene dates as some other kind
of variable, PROC REPORT summarizes them. 4

Computed Variables
Computed variables are variables that you dene for the report. They are not in the
input data set, and PROC REPORT does not add them to the input data set. However,
computed variables are included in an output data set if you create one.
In the windowing environment, you add a computed variable to a report from the
COMPUTED VAR window.
In the nonwindowing environment, you add a computed variable by
3 including the computed variable in the COLUMN statement
3 dening the variables usage as COMPUTED in the DEFINE statement
3 computing the value of the variable in a compute block associated with the
variable.

Featured in: Example 5 on page 960, Example 10 on page 975, and Example 13

on page 983

Interactions of Position and Usage

The position and usage of each variable in the report determine the reports structure
and content. PROC REPORT orders the detail rows of the report according to the

856

Laying Out a Report

Chapter 42

values of order and group variables, considered from left to right in the report.
Similarly, PROC REPORT orders columns for an across variable from left to right,
according to the values of the variable.
Several items can collectively dene the contents of a column in a report. For
instance, in Figure 42.7 on page 856, the values that appear in the third and fourth
columns are collectively determined by Sales, an analysis variable, and by Department,
an across variable. You create this kind of report with the COLUMN statement or, in
the windowing environment, by placing report items above or below each other. This is
called stacking items in the report because each item generates a header, and the
headers are stacked one above the other.

Figure 42.7

Stacking Department and Sales

Sales Figures for Perishables in Northern Sectors

Sector

Manager

______Department_______
Meat/Dairy
Produce

Perishable
Total
-------------------------------------------------------Northeast Alomar
Andrews
Northwest Brown

$190.00

$86.00

$276.00

$300.00

$125.00

$425.00

$250.00

$73.00

$323.00

Pelfrey

$205.00

$76.00

$281.00

Reveiz

$600.00

$30.00

$630.00

When you use multiple items to dene the contents of a column, at most one of the
following can be in a column:

3
3
3
3
3

a display variable with or without a statistic above or below it

an analysis variable with or without a statistic above or below it
an order variable
a group variable
a computed variable.

More than one of these items in a column creates a conict for PROC REPORT about
which values to display.
Table 42.1 on page 857 shows which report items can share a column.
Note:

You cannot stack order variables with other report items.

The REPORT Procedure

Table 42.1

Laying Out a Report

857

Report Items That Can Share Columns

Display

Analysis

Order

Group

Computed

Across

Statistic

Display

Analysis

Order
Group

Computed
variable

Across

Statistic

When a display variable and an across variable share a column, the report must also contain another variable that is
not in the same column.

When a column is dened by stacked report items, PROC REPORT formats the
values in the column by using the format that is specied for the lowest report item in
the stack that does not have an ACROSS usage.
The following items can stand alone in a column:
3 display variable
3 analysis variable
3 order variable
3 group variable
3 computed variable
3 across variable
3 N statistic.
Note: The values in a column that is occupied only by an across variable are
frequency counts. 4

Statistics That Are Available in PROC REPORT

Descriptive statistic keywords
CSS

PCTSUM

RANGE

MAX

STDDEV|STD

MEAN

STDERR

MIN

SUM

SUMWGT

NMISS

USS

PCTN

VAR

Quantile statistic keywords

MEDIAN|P50

Q3|P75

P90

858

Using Compute Blocks

Chapter 42

P95

P10

P99

Q1|P25

QRANGE

Hypothesis testing keyword

PROBT

These statistics, the formulas that are used to calculate them, and their data
requirements are discussed in Keywords and Formulas on page 1340.
To compute standard error and the Students t-test you must use the default value of
VARDEF=, which is DF.
Every statistic except N must be associated with a variable. You associate a statistic
with a variable either by placing the statistic above or below a numeric display variable
or by specifying the statistic as a usage option in the DEFINE statement or in the
DEFINITION window for an analysis variable.
You can place N anywhere because it is the number of observations in the input data
set that contribute to the value in a cell of the report. The value of N does not depend
on a particular variable.
Note: If you use the MISSING option in the PROC REPORT statement, then N
includes observations with missing group, order, or across variables. 4

Using Compute Blocks

What Is a Compute Block?
A compute block is one or more programming statements that appear either between
a COMPUTE and an ENDCOMP statement or in a COMPUTE window. PROC
REPORT executes these statements as it builds the report. A compute block can be
associated with a report item (a data set variable, a statistic, or a computed variable) or
with a location (at the top or bottom of the report; before or after a set of observations).
You create a compute block with the COMPUTE window or with the COMPUTE
statement. One form of the COMPUTE statement associates the compute block with a
report item. Another form associates the compute block with a location in the report
(see Using Break Lines on page 861).
Note: When you use the COMPUTE statement, you do not have to use a
corresponding BREAK or RBREAK statement. (See Example 2 on page 951, which uses
COMPUTE AFTER but does not use the RBREAK statement). Use these statements
only when you want to implement one or more BREAK statement or RBREAK
statement options (see Example 9 on page 971, which uses both COMPUTE AFTER
MANAGER and BREAK AFTER MANAGER. 4

The Purpose of Compute Blocks

A compute block that is associated with a report item can
3 dene a variable that appears in a column of the report but is not in the input
data set
3 dene display attributes for a report item (see CALL DEFINE Statement on
page 890).
A compute block that is associated with a location can write a customized summary.

The REPORT Procedure

Using Compute Blocks

859

In addition, all compute blocks can use most SAS language elements to perform
calculations (see The Contents of Compute Blocks on page 859). A PROC REPORT
step can contain multiple compute blocks, but they cannot be nested.

The Contents of Compute Blocks

In the windowing environment, a compute block is in a COMPUTE window. In the
nonwindowing environment, a compute block begins with a COMPUTE statement and
ends with an ENDCOMP statement. Within a compute block, you can use these SAS
language elements:
3 %INCLUDE statement

3 these DATA step statements:

ARRAY
assignment

LENGTH

CALL

RETURN

DO (all forms)

SELECT

END

3
3
3
3

IF-THEN/ELSE

sum

comments
null statements
macro variables and macro invocations
all DATA step functions.

For information about SAS language elements see the appropriate section in SAS
Language Reference: Dictionary.
Within a compute block, you can also use these PROC REPORT features:
3 Compute blocks for a customized summary can contain one or more LINE
statements, which place customized text and formatted values in the summary.
(See LINE Statement on page 907.)

3 Compute blocks for a report item can contain one or more CALL DEFINE
statements, which set attributes like color and format each time a value for the
item is placed in the report. (See CALL DEFINE Statement on page 890.)

3 Any compute block can contain the automatic variable _BREAK_ (see The
Automatic Variable _BREAK_ on page 862.

Four Ways to Reference Report Items in a Compute Block

A compute block can reference any report item that forms a column in the report
(whether or not the column is visible). You reference report items in a compute block in
one of four ways:

3 by name.
3 by a compound name that identies both the variable and the name of the statistic
that you calculate with it. A compound name has this form
variable-name.statistic

3 by an alias that you create in the COLUMN statement or in the DEFINITION

window.

860

Using Compute Blocks

Chapter 42

3 by column number, in the form

_Cn_
where n is the number of the column (from left to right) in the report.
Note: Even though the columns that you dene with NOPRINT and NOZERO do
not appear in the report, you must count them when you are referencing columns
by number. See the discussion of NOPRINT on page 902 and NOZERO on page
903.

Note: Referencing variables that have missing values leads to missing values. If a
compute block references a variable that has a missing value, then PROC REPORT
displays that variable as a blank (for character variables) or as a period (for numeric
variables). 4
The following table shows how to use each type of reference in a compute block.
If the variable that you
reference is this type

Then refer to it by

For example

group

name*

Department

order

name

computed

name

display

name

Department

display sharing a column with a

statistic

a compound name*

Sales.sum

analysis

a compound name*

Sales.mean

any type sharing a column with an

across variable

column number

_c3_

If the variable has an alias, then you must reference it with the alias.

Even if the variable has an alias, you must reference it by column number.

Example 3 on page 954, which references analysis variables by

their aliases; Example 5 on page 960, which references variables by column
number; and Example 10 on page 975, which references group variables and
computed variables by name.

Featured in:

Compute Block Processing

PROC REPORT processes compute blocks in two different ways.
3 If a compute block is associated with a location, then PROC REPORT executes the
compute block only at that location. Because PROC REPORT calculates statistics
for groups before it actually constructs the rows of the report, statistics for sets of
detail rows are available before or after the rows are displayed, as are values for
any variables based on these statistics.
3 If a compute block is associated with a report item, then PROC REPORT executes
the compute block on every row of the report when it comes to the column for that

The REPORT Procedure

Using Break Lines

861

item. The value of a computed variable in any row of a report is the last value
assigned to that variable during that execution of the DATA step statements in the
compute block. PROC REPORT assigns values to the columns in a row of a report
from left to right. Consequently, you cannot base the calculation of a computed
variable on any variable that appears to its right in the report.
Note: PROC REPORT recalculates computed variables at breaks. For details on
compute block processing see How PROC REPORT Builds a Report on page 936. 4

Using Break Lines

What Are Break Lines?
Break lines are lines of text (including blanks) that appear at particular locations,
called breaks, in a report. A report can contain multiple breaks. Generally, break lines
are used to visually separate parts of a report, to summarize information, or both. They
can occur

3 at the beginning or end of a report

3 at the top or bottom of each page
3 between sets of observations (whenever the value of a group or order variable
changes).
Break lines can contain

3 text
3 values calculated for either a set of rows or for the whole report.

Creating Break Lines

There are two ways to create break lines. The rst way is simpler. It produces a
default summary. The second way is more exible. It produces a customized summary
and provides a way to slightly modify a default summary. Default summaries and
customized summaries can appear at the same location in a report.
Default summaries are produced with the BREAK statement, the RBREAK
statement, or the BREAK window. You can use default summaries to visually separate
parts of the report, to summarize information for numeric variables, or both. Options
provide some control over the appearance of the break lines, but if you choose to
summarize numeric variables, then you have no control over the content and the
placement of the summary information. (A break line that summarizes information is a
summary line.)
Customized summaries are produced in a compute block. You can control both the
appearance and content of a customized summary, but you must write the code to do so.

Order of Break Lines

You control the order of the lines in a customized summary. However, PROC
REPORT controls the order of lines in a default summary and the placement of a
customized summary relative to a default summary. When a default summary contains
multiple break lines, the order in which the break lines appear is
1 overlining or double overlining (in traditional SAS monospace output only)
2 summary line
3 underlining or double underlining (in traditional SAS monospace output only)

862

Using Compound Names

Chapter 42

4 blank line
5 page break.

In traditional SAS monospace output only, if you dene a customized summary for
the same location, then customized break lines appear after underlining or double
underlining.

The Automatic Variable _BREAK_

PROC REPORT automatically creates a variable called _BREAK_. This variable
contains

3 a blank if the current line is not part of a break

3 the value of the break variable if the current line is part of a break between sets of
observations

3 the value _RBREAK_ if the current line is part of a break at the beginning or end of
the report
3 the value _PAGE_ if the current line is part of a break at the beginning or end of a
page.

Using Compound Names

When you use a statistic in a report, you generally refer to it in compute blocks by a
compound name like Sales.sum. However, in different parts of the report, that same
name has different meanings. Consider the report in Output 42.1. The statements that
create the output follow. The user-dened formats that are used are created by a PROC
FORMAT step on page 949.
libname proclib SAS-data-library;

options nodate pageno=1 linesize=64

pagesize=60 fmtsearch=(proclib);
proc report data=grocery nowindows;
column sector manager sales;
define sector / group format=$sctrfmt.;
define sales / analysis sum
format=dollar9.2;
define manager / group format=$mgrfmt.;
break after sector / summarize skip ol;
rbreak after / summarize dol dul;
compute after;
sector=Total:;
endcomp;
run;

The REPORT Procedure

Output 42.1

Using Style Elements in PROC REPORT

863

Three Different Meanings of Sales.sum

The SAS System
Sector
Northeast

Manager
Alomar
Andrews

--------Northeast
Northwest

Sales
$786.00
$1,045.00
--------$1,831.00

Brown
Pelfrey
Reveiz

--------Southwest

$695.00
$353.00
--------$1,048.00

=========
Total:
=========

=========
$6,313.00
=========

$630.00
$350.00
--------$980.00

Adams
Taylor

$598.00
$746.00
$1,110.00
--------$2,454.00

Jones
Smith

--------Northwest
Southeast
--------Southeast
Southwest

Here Sales.sum has three different meanings:

u In detail rows, the value is the sales for one managers store in a sector of the city.
For example, the rst detail row of the report shows that the sales for the store
that Alomar manages were $786.00.
v In the group summary lines, the value is the sales for all the stores in one sector.
For example, the rst group summary line shows that sales for the Northeast
sector were $1,831.00.
w In the report summary line, the value $6,313.00 is the sales for all stores in the
city.

Note: Unless you use the NOALIAS option in the PROC REPORT statement, when
you refer in a compute block to a statistic that has an alias, you do not use a compound
name. Generally, you must use the alias. However, if the statistic shares a column with
an across variable, then you must reference it by column number (see Four Ways to
Reference Report Items in a Compute Block on page 859). 4

Using Style Elements in PROC REPORT

Using the STYLE= Option

If you use the Output Delivery System to create HTML, RTF, or Printer output from
PROC REPORT, then you can use the STYLE= option to specify style elements for the
procedure to use in various parts of the report. Style elements determine presentation
attributes like font type, font weight, color, and so forth. For information about the
attributes that you can set for a style, see SAS Output Delivery System: Users Guide.

864

Using Style Elements in PROC REPORT

Chapter 42

The general form of the STYLE= option is

STYLE<(location(s))>=< style-element-name><[style-attribute-specication(s)]>
Note:

You can use braces ({ and }) instead of square brackets ([ and ]).

location(s)
identies the part of the report that the STYLE= option affects. The following
table shows what parts of a report are affected by values of location.
Table 42.2

Location Values

Location Value

Part of Report Affected

CALLDEF

Cells identied by a CALL DEFINE

statement

COLUMN

Column cells

HEADER|HDR

Column headers

LINES

Lines generated by LINE statements

REPORT

Report as a whole

SUMMARY

Summary lines

The valid and default values for location vary by what statement the STYLE=
option appears in. Table 42.3 on page 864 shows valid and default values for
location for each statement. To specify more than one value of location in the same
STYLE= option, separate each value with a space.
style-element-name
is the name of a style element that is part of a style denition that is registered
with the Output Delivery System. SAS provides some style denitions. Users can
create their own style denitions with the TEMPLATE procedure (see SAS Output
Delivery System: Users Guide for information about PROC TEMPLATE). The
following table shows the default style elements for each statement.
Table 42.3

Locations and Default Style Elements for Each Statement in PROC REPORT

Statement

Valid Location Values

Default Location
Value

Default Style
Element

PROC REPORT

REPORT, COLUMN, HEADER|HDR,

SUMMARY, LINES, CALLDEF

REPORT

Table

BREAK

SUMMARY, LINES

SUMMARY

DataEmphasis

CALL DEFINE

CALLDEF

Data

COMPUTE

LINES

NoteContent

DEFINE

COLUMN, HEADER|HDR

COLUMN and
HEADER

COLUMN: Data

SUMMARY

DataEmphasis

RBREAK

SUMMARY, LINES

HEADER: Header

style-attribute-specication(s)
describes the style attribute to change. Each style-attribute-specication has this
general form:
style-attribute-name=style-attribute-value

The REPORT Procedure

Using Style Elements in PROC REPORT

865

To specify more than one style-attribute-specication, separate each one with a

space.
The following table shows valid values of style-attribute-name for the REPORT
location. Note that not all style attributes are valid in all destinations. See SAS
Output Delivery System: Users Guide for more information on these style
attributes, their valid values, and their applicable destinations.
BACKGROUND=

FONT_WIDTH=*

BACKGROUNDIMAGE=

FOREGROUND=*

BORDERCOLOR=

FRAME=

BORDERCOLORDARK=

HTMLCLASS=

BORDERCOLORLIGHT=

JUST=

BORDERWIDTH=

OUTPUTWIDTH=

CELLPADDING=

POSTHTML=

CELLSPACING=

POSTIMAGE=

FONT=*

POSTTEXT=
*

FONT_FACE=

PREHTML=

FONT_SIZE=

PREIMAGE=
*

FONT_STYLE=

PRETEXT=

FONT_WEIGHT=*

RULES=

* When you use these attributes in this location, they affect only the text that is specied
with the PRETEXT=, POSTTEXT=, PREHTML=, and POSTHTML= attributes. To alter
the foreground color or the font for the text that appears in the table, you must set the
corresponding attribute in a location that affects the cells rather than the table.

The following table shows valid values of style-attribute-name for the CALLDEF,
COLUMN, HEADER, LINES, and SUMMARY locations. Note that not all style
attributes are valid in all destinations. See SAS Output Delivery System: Users
Guide for more information on these style attributes, their valid values, and their
applicable destinations.
ASIS=

FONT_WIDTH=

BACKGROUND=

HREFTARGET=

BACKGROUNDIMAGE=

HTMLCLASS=

BORDERCOLOR=

JUST=

BORDERCOLORDARK=

NOBREAKSPACE=

BORDERCOLORLIGHT=

POSTHTML=

BORDERWIDTH=

POSTIMAGE=

CELLHEIGHT=

POSTTEXT=

CELLWIDTH=

PREHTML=

FLYOVER=

PREIMAGE=

FONT=

PRETEXT=

866

Using Style Elements in PROC REPORT

Chapter 42

FONT_FACE=

PROTECTSPECIALCHARS=

FONT_SIZE=

TAGATTR=

FONT_STYLE=

URL=

FONT_WEIGHT=

VJUST=

Specications in a statement other than the PROC REPORT statement override the
same specication in the PROC REPORT statement. However, any style attributes that
you specify in the PROC REPORT statement and do not override in another statement
are inherited. For instance, if you specify a blue background and a white foreground for
all column headings in the PROC REPORT statement, and you specify a gray
background for the column headings of a variable in the DEFINE statement, then the
background for that particular column heading is gray, and the foreground is white (as
specied in the PROC REPORT statement).

Using a Format to Assign a Style Attribute Value

You can use a format to assign a style attribute value. For example, the following
code assigns a red background color to cells in the Prot column for which the value is
negative, and a green background color where the values are positive:
proc format;
value proffmt low-<0=red
0-high=green;
run;
ods html body=external-HTML-file;
proc report data=profits nowd;
title Profits for Individual Stores;
column Store Profit;
define Store / display Store;
define Profit / display Profit style=[background=proffmt.];
run;
ods html close;

Controlling the Spacing between Rows

Users frequently need to shrink a report to t more rows on a page. Shrinking a
report involves changing both the font size and the spacing between the rows. In order
to give maximum exibility to the user, ODS uses the font size that is specied for the
REPORT location to calculate the spacing between the rows. Therefore, to shrink a
table, change the font size for both the REPORT location and the COLUMN location.
Here is an example:
proc report nowindows data=libref.data---set-name
style(report)=[font_size=8pt]
style(column)=[font=(Arial, 8pt)];

The REPORT Procedure

Printing a Report

867

Printing a Report

Printing with ODS

Printing reports with the Output Delivery System is much simpler and provides
more attractive output than the older methods of printing that are documented here.
For best results, use an output destination such as Printer or RTF. For details on these
destinations and on using the ODS statement, see SAS Output Delivery System: Users
Guide.

Printing from the REPORT Window

By default, if you print from the REPORT window, then the report is routed directly
to your printer. If you want, you can specify a form to use for printing (see Printing
with a Form on page 867). Forms specify things like the type of printer that you are
using, text format, and page orientation.
Note:

Forms are available only when you run SAS from a windowing environment.

Operating Environment Information: Printing is implemented differently in different

operating environments. For information related to printing, consult SAS Language
Reference: Concepts. Additional information may be available in the SAS
documentation for your operating environment. 4

Printing with a Form

To print with a form from the REPORT window:
1 Specify a form. You can specify a form with the FORMNAME command or, in
some cases, through the File menu.
2 Specify a print le if you want the output to go to a le instead of directly to the
printer. You can specify a print le with the PRTFILE command or, in some cases,
through the File menu.
3 Issue the PRINT or PRINT PAGE command from the command line or from the
File menu.
4 If you specied a print le, then do the following:
a Free the print le. You can free a le with the FREE command or, in some

cases, through Print utilities in the File menu. You cannot view or print
the le until you free it.
b Use operating environment commands to send the le to the printer.

Printing from the Output Window

If you are running PROC REPORT with the NOWINDOWS option, then the default
destination for the output is the Output window. Use the commands in the File menu
to print the report.

Printing from Noninteractive or Batch Mode

If you use noninteractive or batch mode, then SAS writes the output either to the
display or to external les, depending on the operating environment and on the SAS

868

Storing and Reusing a Report Denition

Chapter 42

options that you use. Refer to the SAS documentation for your operating environment
for information about how these les are named and where they are stored.
You can print the output le directly or use PROC PRINTTO to redirect the output to
another le. In either case, no form is used, but carriage control characters are written
if the destination is a print le.
Use operating environment commands to send the le to the printer.

Printing from Interactive Line Mode

If you use interactive line mode, then by default the output and log are displayed on
the screen immediately following the programming statements. Use PROC PRINTTO
to redirect the output to an external le. Then use operating environment commands to
send the le to the printer.

Using PROC PRINTTO

PROC PRINTTO denes destinations for the SAS output and the SAS log (see
Chapter 35, The PRINTTO Procedure, on page 771).
PROC PRINTTO does not use a form, but it does write carriage control characters if
you are writing to a print le.
Note: You need two PROC PRINTTO steps. The rst PROC PRINTTO step
precedes the PROC REPORT step. It redirects the output to a le. The second PROC
PRINTTO step follows the PROC REPORT step. It reestablishes the default destination
and frees the output le. You cannot print the le until PROC PRINTTO frees it. 4

Storing and Reusing a Report Denition

The OUTREPT= option in the PROC REPORT statement stores a report denition in
the specied catalog entry. If you are working in the nonwindowing environment, then
the denition is based on the PROC REPORT step that you submit. If you are in the
windowing environment, then the denition is based on the report that is in the
REPORT window when you end the procedure. SAS assigns an entry type of REPT to
the entry.
In the windowing environment, you can save the denition of the current report by
selecting
File

Save Report

A report denition may differ from the SAS program that creates the report (see the
discussion of OUTREPT= on page 879).
You can use a report denition to create an identically structured report for any SAS
data set that contains variables with the same names as the ones that are used in the
report denition. Use the REPORT= option in the PROC REPORT statement to load a
report denition when you start PROC REPORT. In the windowing environment, load a
report denition from the LOAD REPORT window by selecting
File

Open Report

The REPORT Procedure

Syntax: REPORT Procedure

869

Syntax: REPORT Procedure

Supports the Output Delivery System. See Output Delivery System on page 32
for details.

Tip:

ODS Table Name:

Report

Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See

Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.

PROC REPORT < option(s)>;

BREAK location break-variable</ option(s)>;
BY <DESCENDING> variable-1
<< DESCENDING> variable-n> <NOTSORTED>;
COLUMN column-specication(s);
COMPUTE location < target>
< / STYLE=<style-element-name>
< [style-attribute-specication(s)]>>;
LINE specication(s);
. . . select SAS language elements . . .
ENDCOMP;
COMPUTE report-item </ type-specication>;
CALL DEFINE (column-id, attribute-name, value);
. . . select SAS language elements . . .
ENDCOMP;
DEFINE report-item / <usage>
<attribute(s)>
<option(s)>
<justication>
<COLOR=color>
<column-header-1 <column-header-n>>
<style>;
FREQ variable;
RBREAK location </ option(s)>;
WEIGHT variable;

To do this

Use this statement

Produce a default summary at a change in the

value of a group or order variable

BREAK

Create a separate report for each BY group

Set the value of an attribute for a particular

column in the current row

CALL DEFINE

Describe the arrangement of all columns and of

headers that span more than one column

COLUMN

870

PROC REPORT Statement

Chapter 42

To do this

Use this statement

Specify one or more programming statements

that PROC REPORT executes as it builds the
report

COMPUTE and ENDCOMP

Describe how to use and display a report item

DEFINE

Treat observations as if they appear multiple

times in the input data set

FREQ

Provide a subset of features of the PUT

statement for writing customized summaries

LINE

Produce a default summary at the beginning or

end of a report or at the beginning and end of
each BY group

RBREAK

Specify weights for analysis variables in the

statistical calculations

WEIGHT

PROC REPORT Statement

PROC REPORT < option(s)>;

To do this

Use this option

Specify the input data set

DATA=

Specify the output data set

OUT=

Override the SAS system option THREADS

| NOTHREADS

THREADS | NOTHREADS

Select the windowing or the nonwindowing

environment

WINDOWS|NOWINDOWS

Use a report that was created before

compute blocks required aliases (before
Release 6.11)

NOALIAS

Control the statistical analysis

Specify the divisor to use in the
calculation of variances

VARDEF=

Specify the sample size to use for the

P2 quantile estimation method

QMARKERS=

Specify the quantile estimation method

QMETHOD=

Specify the mathematical denition to

calculate quantiles

QNTLDEF=

Exclude observations with nonpositive

weight values from the analysis.

EXCLNPWGT

The REPORT Procedure

To do this

PROC REPORT Statement

Use this option

Control classication levels

Create all possible combinations of the
across variable values

COMPLETECOLS|NOCOMPLETECOLS

Create all possible combinations of the

group variable values

COMPLETEROWS|NOCOMPLETEROWS

Control the layout of the report

Use formatting characters to add
line-drawing characters to the report

BOX*

Specify whether to center or left-justify

the report and summary text

CENTER|NOCENTER

Specify the default number of

characters for columns containing
computed variables or numeric data
set variables

COLWIDTH=*

Dene the characters to use as

line-drawing characters in the report

FORMCHAR=*

Specify the length of a line of the report

LS=*

Consider missing values as valid values

for group, order, or across variables

MISSING

Specify the number of panels on each

page of the report

PANELS=*

Specify the number of lines in a page

of the report

PS=

Specify the number of blank characters

between panels

PSPACE=*

Override options in the DEFINE

statement that suppress the display of
a column

SHOWALL

Specify the number of blank characters

between columns

SPACING=*

Display one value from each column of

the report, on consecutive lines if
necessary, before displaying another
value from the rst column

WRAP

Customize column headers

Underline all column headers and the
spaces between them

HEADLINE*

Write a blank line beneath all column

headers

HEADSKIP*

Suppress column headers

NOHEADER

Write name= in front of each value in

the report, where name= is the column
header for the value

NAMED

Specify the split character

SPLIT=

871

872

PROC REPORT Statement

Chapter 42

To do this

Use this option

Control ODS output

Specify one or more style elements (for
the Output Delivery System) to use for
different parts of the report

STYLE=

Specify text for the HTML or PDF

table of contents entry for the output

CONTENTS=

Store and retrieve report denitions, PROC REPORT statements, and your report prole
Write to the SAS log the PROC
REPORT code that creates the current
report

LIST

Suppress the building of the report

NOEXEC

Store in the specied catalog the report

denition that is dened by the PROC
REPORT step that you submit

OUTREPT=

Identify the report prole to use

PROFILE=

Specify the report denition to use

REPORT=

Control the windowing environment

Display command lines rather than
menu bars in all REPORT windows

COMMAND

Identify the library and catalog

containing user-dened help for the
report

HELP=

Open the REPORT window and start

the PROMPT facility

PROMPT

* Traditional SAS monospace output only.

Options
BOX

uses formatting characters to add line-drawing characters to the report. These

characters
3 surround each page of the report
3 separate column headers from the body of the report
3 separate rows and columns from each other
3 separate values in a summary line from other values in the same columns
3 separate a customized summary from the rest of the report.
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: You cannot use BOX if you use WRAP in the PROC REPORT
statement or in the ROPTIONS window or if you use FLOW in any item denition.
See also: the discussion of FORMCHAR= on page 874
Featured in: Example 12 on page 980

The REPORT Procedure

PROC REPORT Statement

873

CENTER|NOCENTER

species whether to center or left-justify the report and summary text (customized
break lines).
PROC REPORT honors the rst of these centering specications that it nds:

3 the CENTER or NOCENTER option in the PROC REPORT statement or the

CENTER toggle in the ROPTIONS window

3 the CENTER or NOCENTER option stored in the report denition that is

loaded with REPORT= in the PROC REPORT statement

3 the SAS system option CENTER or NOCENTER.

Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: When CENTER is in effect, PROC REPORT ignores spacing that

precedes the leftmost variable in the report.

COLWIDTH=column-width

species the default number of characters for columns containing computed variables
or numeric data set variables.
Default: 9
Range: 1 to the linesize
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: When setting the width for a column, PROC REPORT rst looks at

WIDTH= in the denition for that column. If WIDTH= is not present, then PROC
REPORT uses a column width large enough to accommodate the format for the
item. (For information about formats see the discussion of FORMAT= on page 901.)
If no format is associated with the item, then the column width depends on
variable type:
If the variable is a

Then the column width is the

character variable in the input data set

length of the variable

numeric variable in the input data set

value of the COLWIDTH= option

computed variable (numeric or character)

value of the COLWIDTH= option

Featured in:

Example 2 on page 951

COMMAND

displays command lines rather than menu bars in all REPORT windows.
After you have started PROC REPORT in the windowing environment, you can
display the menu bars in the current window by issuing the COMMAND command.
You can display the menu bars in all PROC REPORT windows by issuing the
PMENU command. The PMENU command affects all the windows in your SAS
session. Both of these commands are toggles.
You can store a setting of COMMAND in your report prole. PROC REPORT
honors the rst of these settings that it nds:

3 the COMMAND option in the PROC REPORT statement

3 the setting in your report prole.
Restriction: This option has no effect in the nonwindowing environment.

874

PROC REPORT Statement

Chapter 42

COMPLETECOLS|NOCOMPLETECOLS

creates all possible combinations for the values of the across variables even if one or
more of the combinations do not occur within the input data set. Consequently, the
column headings are the same for all logical pages of the report within a single BY
group.
Default: COMPLETECOLS
Interaction: The PRELOADFMT option in the DEFINE statement ensures that
PROC REPORT uses all user-dened format ranges for the combinations of across
variables, even when a frequency is zero.
COMPLETEROWS|NOCOMPLETEROWS

displays all possible combinations of the values of the group variables, even if one or
more of the combinations do not occur in the input data set. Consequently, the row
headings are the same for all logical pages of the report within a single BY group.
Default: NOCOMPLETEROWS
Interaction: The PRELOADFMT option in the DEFINE statement ensures that
PROC REPORT uses all user-dened format ranges for the combinations of group
variables, even when a frequency is zero.
CONTENTS=link-text

species the text for the entries in the HTML contents le or PDF table of contents
for the output that is produced by PROC REPORT. For information on HTML and
PDF output, see Output Delivery System on page 32.
Note: A hexadecimal value (such as DFx) that is specied within link-text will
not resolve because it is specied within quotation marks. To resolve a hexadecimal
value, use the %sysfunc(byte(num)) function, where num is the hexadecimal value.
Be sure to enclose link-text in double quotation marks (" ") so that the macro function
will resolve. 4
Restriction: For HTML output, the CONTENTS= option has no effect on the
HTML body le. It affects only the HTML contents le.
DATA=SAS-data-set

species the input data set.

Main discussion: Input Data Sets on page 19
EXCLNPWGT

excludes observations with nonpositive weight values (zero or negative) from the
analysis. By default, PROC REPORT treats observations with negative weights like
those with zero weights and counts them in the total number of observations.
Alias: EXCLNPWGTS
Requirement: You must use a WEIGHT statement.
See also: WEIGHT Statement on page 912
FORMCHAR <(position(s))>=formatting-character(s)

denes the characters to use as line-drawing characters in the report.

position(s)
identies the position of one or more characters in the SAS formatting-character
string. A space or a comma separates the positions.
Default: Omitting (position(s)) is the same as specifying all 20 possible SAS
formatting characters, in order.
Range: PROC REPORT uses 12 of the 20 formatting characters that SAS
provides. Table 42.4 on page 875 shows the formatting characters that PROC
REPORT uses. Figure 42.8 on page 876 illustrates the use of some commonly
used formatting character in the output from PROC REPORT.

The REPORT Procedure

PROC REPORT Statement

875

formatting-character(s)
lists the characters to use for the specied positions. PROC REPORT assigns
characters in formatting-character(s) to position(s), in the order that they are
listed. For instance, the following option assigns the asterisk (*) to the third
formatting character, the pound sign (#) to the seventh character, and does not
alter the remaining characters:
formchar(3,7)=*#

Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: The SAS system option FORMCHAR= species the default formatting

Tip:

formchar(3,7)=2D7Cx

Table 42.4

Formatting Characters Used by PROC REPORT

Position

Default

Used to draw

the right and left borders and

the vertical separators
between columns

the top and bottom borders

and the horizontal separators
between rows; also
underlining and overlining in
break lines as well as the
underlining that the
HEADLINE option draws

the top character in the left

border

the top character in a line of

characters that separates
columns

the top character in the right

border

the leftmost character in a

row of horizontal separators

the intersection of a column of

vertical characters and a row
of horizontal characters

the rightmost character in a

row of horizontal separators

876

PROC REPORT Statement

Chapter 42

the bottom character in the

left border

the bottom character in a line

of characters that separate
columns

the bottom character in the

right border

double overlining and double

underlining in break lines

Figure 42.8

Formatting Characters in PROC REPORT Output

Sales for Northern Sectors

Sector
Manager
Sales
-----------------------------Northeast

Northwest

Alomar
Andrews

Brown
Pelfrey
Reveiz

786.00
1,045.00
---------1,831.00
----------

598.00
746.00
1,110.00
---------2,454.00
---------==========
4,285.00
==========

HEADLINE

underlines all column headers and the spaces between them at the top of each page
of the report.
The HEADLINE option underlines with the second formatting character. (See the
discussion of FORMCHAR= on page 874 .)
Default: hyphen (-)
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Tip: In traditional (monospace) SAS output, you can underline column headers
without underlining the spaces between them, by using two hyphens (--) as
the last line of each column header instead of using HEADLINE.
Featured in: Example 2 on page 951 and Example 8 on page 968
HEADSKIP

writes a blank line beneath all column headers (or beneath the underlining that the
HEADLINE option writes) at the top of each page of the report.
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.

The REPORT Procedure

PROC REPORT Statement

877

Example 2 on page 951

Featured in:

HELP=libref.catalog

identies the library and catalog containing user-dened help for the report. This
help can be in CBT or HELP catalog entries. You can write a CBT or HELP entry for
each item in the report with the BUILD procedure in SAS/AF software. Store all
such entries for a report in the same catalog.
Specify the entry name for help for a particular report item in the DEFINITION
window for that report item or in a DEFINE statement.
Restriction: This option has no effect in the nonwindowing environment or on ODS
destinations other than traditional SAS monospace output.
LIST

writes to the SAS log the PROC REPORT code that creates the current report. This
listing may differ in these ways from the statements that you submit:
3 It shows some defaults that you may not have specied.
3 It omits some statements that are not specic to the REPORT procedure,
whether you submit them with the PROC REPORT step or had previously
submitted them. These statements include
BY
FOOTNOTE
FREQ
TITLE
WEIGHT
WHERE

3 It omits these PROC REPORT statement options:

LIST
OUT=
OUTREPT=
PROFILE=
REPORT=
WINDOWS|NOWINDOWS
3 It omits SAS system options.
3 It resolves automatic macro variables.
Restriction: This option has no effect in the windowing environment. In the

windowing environment, you can write the report denition for the report that is
currently in the REPORT window to the SOURCE window by selecting
Tools

Report Statements

LS=line-size

species the length of a line of the report.

PROC REPORT honors the rst of these line size specications that it nds:

3 the LS= option in the PROC REPORT statement or Linesize= in the

ROPTIONS window

3 the LS= setting stored in the report denition loaded with REPORT= in the
PROC REPORT statement
3 the SAS system option LINESIZE=.

878

PROC REPORT Statement

Chapter 42

64-256 (integer)
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Featured in: Example 6 on page 964 and Example 8 on page 968
Range:

MISSING

considers missing values as valid values for group, order, or across variables. Special
missing values used to represent numeric values (the letters A through Z and the
underscore (_) character) are each considered as a different value. A group for each
missing value appears in the report. If you omit the MISSING option, then PROC
REPORT does not include observations with a missing value for any group, order, or
across variables in the report.
See also: For information about special missing values, see the section on missing
values in SAS Language Reference: Concepts.
Featured in: Example 11 on page 977
NAMED

writes name= in front of each value in the report, where name is the column header
for the value.
Interaction: When you use the NAMED option, PROC REPORT automatically uses
the NOHEADER option.
Tip: Use NAMED in conjunction with the WRAP option to produce a report that
wraps all columns for a single row of the report onto consecutive lines rather than
placing columns of a wide report on separate pages.
Featured in: Example 7 on page 966
NOALIAS

lets you use a report that was created before compute blocks required aliases (before
Release 6.11). If you use NOALIAS, then you cannot use aliases in compute blocks.
NOCENTER

See CENTER|NOCENTER on page 873.

NOCOMPLETECOLS

See COMPLETECOLS|NOCOMPLETECOLS on page 874.

NOCOMPLETEROWS

See COMPLETEROWS|NOCOMPLETEROWS on page 874.

NOEXEC

suppresses the building of the report. Use NOEXEC with OUTREPT= to store a
report denition in a catalog entry. Use NOEXEC with LIST and REPORT= to
display a listing of the specied report denition.
NOHEADER

suppresses column headers, including those that span multiple columns.

When you suppress the display of column headers in the windowing environment,
you cannot select any report items.
NOTHREADS

See THREADS | NOTHREADS on page 883.

NOWINDOWS

NOWD
See WINDOWS|NOWINDOWS on page 884.
Alias:

OUT=SAS-data-set

names the output data set. If this data set does not exist, then PROC REPORT
creates it. The data set contains one observation for each detail row of the report and

The REPORT Procedure

PROC REPORT Statement

879

one observation for each unique summary line. If you use both customized and
default summaries at the same place in the report, then the output data set contains
only one observation because the two summaries differ only in how they present the
data. Information about customization (underlining, color, text, and so forth) is not
data and is not saved in the output data set.
The output data set contains one variable for each column of the report. PROC
REPORT tries to use the name of the report item as the name of the corresponding
variable in the output data set. However, this is not possible if a data set variable is
under or over an across variable or if a data set variable appears multiple times in
the COLUMN statement without aliases. In these cases, the name of the variable is
based on the column number (_C1_, _C2_, and so forth).
Output data set variables that are derived from input data set variables retain the
formats of their counterparts in the input data set. PROC REPORT derives labels for
these variables from the corresponding column headers in the report unless the only
item dening the column is an across variable. In that case, the variables have no
label. If multiple items are stacked in a column, then the labels of the corresponding
output data set variables come from the analysis variable in the column.
The output data set also contains a character variable named _BREAK_. If an
observation in the output data set derives from a detail row in the report, then the
value of _BREAK_ is missing. If the observation derives from a summary line, then
the value of _BREAK_ is the name of the break variable that is associated with the
summary line, or _RBREAK_. If the observation derives from a COMPUTE BEFORE
_PAGE_ or COMPUTE AFTER _PAGE_ statement, then the value of _BREAK_ is
_PAGE_. Note, however, that for COMPUTE BEFORE _PAGE_ and COMPUTE
AFTER _PAGE_, the _PAGE_ value is written to the output data set only; it is not
available as a value of the automatic variable _BREAK_ during execution of the
procedure.
Interaction: You cannot use OUT= in a PROC REPORT step that uses a BY
statement.
Featured in: Example 12 on page 980 and Example 13 on page 983
OUTREPT=libref.catalog.entry

stores in the specied catalog entry the REPORT denition that is dened by the
PROC REPORT step that you submit. PROC REPORT assigns the entry a type of
REPT.
The stored report denition may differ in these ways from the statements that you
submit:
3 It omits some statements that are not specic to the REPORT procedure,
whether you submit them with the PROC REPORT step or whether they are
already in effect when you submit the step. These statements include
BY
FOOTNOTE
FREQ
TITLE
WEIGHT
WHERE
3 It omits these PROC REPORT statement options:
LIST
NOALIAS
OUT=

880

PROC REPORT Statement

Chapter 42

OUTREPT=
PROFILE=
REPORT=
WINDOWS|NOWINDOWS

3 It omits SAS system options.

3 It resolves automatic macro variables.
Note: The current version of SAS will correctly read REPORT entries that were
created with earlier versions. However, earlier versions of SAS will not correctly read
REPORT entries that are created with the current version. 4
Featured in:

Example 7 on page 966

PANELS=number-of-panels

species the number of panels on each page of the report. If the width of a report is
less than half of the line size, then you can display the data in multiple sets of
columns so that rows that would otherwise appear on multiple pages appear on the
same page. Each set of columns is a panel. A familiar example of this kind of report
is a telephone book, which contains multiple panels of names and telephone numbers
on a single page.
When PROC REPORT writes a multipanel report, it lls one panel before
beginning the next.
The number of panels that ts on a page depends on the

3 width of the panel

3 space between panels
3 line size.
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output. However, the COLUMNS= option in the ODS PRINTER
or ODS PDF statement produces similar results. For details, see the chapter on
ODS statements in SAS Output Delivery System: Users Guide.
Default: 1

If number-of-panels is larger than the number of panels that can t on the

page, then PROC REPORT creates as many panels as it can. Let PROC REPORT
put your data in the maximum number of panels that can t on the page by
specifying a large number of panels (for example, 99).

Tip:

See also: For information about the space between panels and the line size, see the

discussions of PSPACE= on page 881 and the discussion of LS= on page 877.
Featured in:

Example 8 on page 968

PCTLDEF=

See QNTLDEF= on page 882.

PROFILE=libref.catalog

identies the report prole to use. A prole

3 species the location of menus that dene alternative menu bars and pull-down
menus for the REPORT and COMPUTE windows.

3 sets defaults for WINDOWS, PROMPT, and COMMAND.

PROC REPORT uses the entry REPORT.PROFILE in the catalog that you specify
as your prole. If no such entry exists, or if you do not specify a prole, then PROC
REPORT uses the entry REPORT.PROFILE in SASUSER.PROFILE. If you have no
prole, then PROC REPORT uses default menus and the default settings of the
options.

The REPORT Procedure

PROC REPORT Statement

881

You create a prole from the PROFILE window while using PROC REPORT in a
windowing environment. To create a prole
1 Invoke PROC REPORT with the WINDOWS option.
2 Select

Tools

Report Prole

3 Fill in the elds to suit your needs.

4 Select OK to exit the PROFILE window. When you exit the window, PROC

REPORT stores the prole in SASUSER.PROFILE.REPORT.PROFILE. Use the

CATALOG procedure or the Explorer window to copy the prole to another
location.
Note: If, after opening the PROFILE window, you decide not to create a prole,
then select CANCEL to close the window.

PROMPT

opens the REPORT window and starts the PROMPT facility. This facility guides you
through creating a new report or adding more data set variables or statistics to an
existing report.
If you start PROC REPORT with prompting, then the rst window gives you a
chance to limit the number of observations that are used during prompting. When
you exit the prompter, PROC REPORT removes the limit.
Restriction: When you use the PROMPT option, you open the REPORT window.

When the REPORT window is open, you cannot send procedure output to any ODS
destination.
You can store a setting of PROMPT in your report prole. PROC REPORT
honors the rst of these settings that it nds:

Tip:

3 the PROMPT option in the PROC REPORT statement

3 the setting in your report prole.
If you omit PROMPT from the PROC REPORT statement, then the procedure uses
the setting in your report prole, if you have one. If you do not have a report prole,
then PROC REPORT does not use the prompt facility. For information on report
proles, see PROFILE on page 926.
PS=page-size

species the number of lines in a page of the report.

PROC REPORT honors the rst of these page size specications that it nds:

3 the PS= option in the PROC REPORT statement

3 the PS= setting in the report denition specied with REPORT= in the PROC
REPORT statement

3 the SAS system option PAGESIZE=.

Range: 15-32,767 (integer)
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Featured in:

Example 6 on page 964 and Example 8 on page 968

PSPACE=space-between-panels

species the number of blank characters between panels. PROC REPORT separates
all panels in the report by the same number of blank characters. For each panel, the
sum of its width and the number of blank characters separating it from the panel to
its left cannot exceed the line size.

882

PROC REPORT Statement

Chapter 42

Default: 4
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Featured in:

Example 8 on page 968

QMARKERS=number
2

species the default number of markers to use for the P estimation method. The
number of markers controls the size of xed memory space.
Default: The default value depends on which quantiles you request. For the median

(P50), number is 7. For the quartiles (P25 and P75), number is 25. For the
quantiles P1, P5, P10, P90, P95, or P99, number is 105. If you request several
quantiles, then PROC REPORT uses the largest default value of number.
any odd integer greater than 3

Range:

Increase the number of markers above the default settings to improve the
accuracy of the estimates; you can reduce the number of markers to conserve
computing resources.

Tip:

QMETHOD=OS|P2

species the method that PROC REPORT uses to process the input data when it
computes quantiles. If the number of observations is less than or equal to the value
of the QMARKERS= option, and the value of the QNTLDEF= option is 5, then both
methods produce the same results.
OS
uses order statistics. This is the technique that PROC UNIVARIATE uses.
Note: This technique can be very memory intensive.

P2
2
uses the P method to approximate the quantile.
Default: OS
Restriction: When QMETHOD=P2, PROC REPORT does not compute weighted

quantiles.
When QMETHOD=P2, reliable estimates of some quantiles (P1, P5, P95, P99)
might not be possible for some data sets such as those with heavily tailed or
skewed distributions.

Tip:

QNTLDEF=1|2|3|4|5

species the mathematical denition that the procedure uses to calculate quantiles
when the value of the QMETHOD= option is OS. When QMETHOD=P2, you must
use QNTLDEF=5.
Default: 5
Alias:

PCTLDEF=

Main discussion: Quantile and Related Statistics on page 1345

REPORT=libref.catalog.entry

species the report denition to use. PROC REPORT stores all report denitions as
entries of type REPT in a SAS catalog.
Interaction: If you use REPORT=, then you cannot use the COLUMN statement.
See also:

OUTREPT= on page 879

Featured in:

Example 7 on page 966

SHOWALL

overrides options in the DEFINE statement that suppress the display of a column.
See also: NOPRINT and NOZERO in DEFINE Statement on page 897

The REPORT Procedure

PROC REPORT Statement

883

SPACING=space-between-columns

species the number of blank characters between columns. For each column, the sum
of its width and the blank characters between it and the column to its left cannot
exceed the line size.
Default: 2
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Interaction: PROC REPORT separates all columns in the report by the number of
blank characters specied by SPACING= in the PROC REPORT statement unless
you use SPACING= in the DEFINE statement to change the spacing to the left of
a specic item.
Interaction: When CENTER is in effect, PROC REPORT ignores spacing that
precedes the leftmost variable in the report.
Featured in: Example 2 on page 951
SPLIT=character

species the split character. PROC REPORT breaks a column header when it
reaches that character and continues the header on the next line. The split character
itself is not part of the column header although each occurrence of the split character
counts toward the 256-character maximum for a label.
Default: slash (/)
Interaction: The FLOW option in the DEFINE statement honors the split character.
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Featured in: Example 5 on page 960
STYLE<(location(s))>=<style-element-name><[style-attribute-specication(s)]>

species the style element to use for the specied locations in the report. See Using
Style Elements in PROC REPORT on page 863 for details.
Restriction: This option affects only the HTML, RTF, and Printer output.
Featured in: Example 15 on page 989 and Example 16 on page 994
THREADS | NOTHREADS

enables or disables parallel processing of the input data set. This option overrides
the SAS system option THREADS | NOTHREADS. See SAS Language Reference:
Concepts for more information about parallel processing.
Default: value of SAS system option THREADS | NOTHREADS.
Interaction: PROC REPORT uses the value of the SAS system option THREADS
except when a BY statement is specied or the value of the SAS system option
CPUCOUNT is less than 2. You can use THREADS in the PROC REPORT
statement to force PROC REPORT to use parallel processing in these situations.
VARDEF=divisor

species the divisor to use in the calculation of the variance and standard deviation.
Table 42.5 on page 883 shows the possible values for divisor and associated divisors.

Table 42.5

Possible Values for VARDEF=

Value

Divisor

degrees of freedom

number of observations

Formula for Divisor

n1
n

884

PROC REPORT Statement

Chapter 42

Value

Divisor

Formula for Divisor

WDF

sum of weights minus one

WEIGHT|WGT

sum of weights

(6i wi) 1

6i wi

P (x 0 x) CSS=divisor

CSS

The procedure computes the variance as

, where
is the corrected
2
. When you weight the analysis variables,
sums of squares and equals
i
2
, where w is the weighted mean.
equals
i
w
i
Default: DF
Requirement: To compute the standard error of the mean and Students t-test, use
the default value of VARDEF=.
Tip: When you use the WEIGHT statement and VARDEF=DF, the variance is an
estimate of 2 , where the variance of the ith observation is var (xi ) = 2 =wi and
wi is the weight for the ith observation. This yields an estimate of the variance of
an observation with unit weight.
Tip: When you use the WEIGHT statement and VARDEF=WGT, the computed
variance is asymptotically (for large n) an estimate of 2 =w, where w is the
average weight. This yields an asymptotic estimate of the variance of an
observation with average weight.
See also: WEIGHT on page 63

CSS

P w (x 0 x )

WINDOWS|NOWINDOWS

selects a windowing or nonwindowing environment.

When you use WINDOWS, SAS opens the REPORT window, which enables you to
modify a report repeatedly and to see the modications immediately. When you use
NOWINDOWS, PROC REPORT runs without the REPORT window and sends its
output to the open output destination(s).
Alias: WD|NOWD
Restriction: When you use the WINDOWS option, you cannot send procedure
output to the HTML, RTF, or Printer destination.
Tip: You can store a setting of WINDOWS in your report prole, if you have one. If
you do not specify WINDOWS or NOWINDOWS in the PROC REPORT statement,
then the procedure uses the setting in your report prole. If you do not have a
report prole, then PROC REPORT looks at the setting of the SAS system option
DMS. If DMS is ON, then PROC REPORT uses the windowing environment; if
DMS is OFF, then it uses the nonwindowing environment.
See also: For a discussion of the report prole see the discussion of PROFILE= on
page 880.
Featured in: Example 1 on page 948
WRAP

displays one value from each column of the report, on consecutive lines if necessary,
before displaying another value from the rst column. By default, PROC REPORT
displays values for only as many columns as it can t on one page. It lls a page
with values for these columns before starting to display values for the remaining
columns on the next page.
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Interaction: When WRAP is in effect, PROC REPORT ignores PAGE in any item
denitions.
Tip: Typically, you use WRAP in conjunction with the NAMED option in order to
avoid wrapping column headers.

The REPORT Procedure

Featured in:

BREAK Statement

885

Example 7 on page 966

BREAK Statement
Produces a default summary at a break (a change in the value of a group or order variable). The
information in a summary applies to a set of observations. The observations share a unique
combination of values for the break variable and all other group or order variables to the left of
the break variable in the report.
Featured in: Example 4 on page 957 and Example 5 on page 960.

BREAK location break-variable< / option(s)>;

To do this

Use this option

Specify the color of the break lines in the REPORT window

COLOR=

Double overline each value

DOL*

Double underline each value

DUL*

Overline each value

OL*

Start a new page after the last break line

PAGE

Write a blank line for the last break line

SKIP

Specify a style element for default summary lines, customized

summary lines or both

STYLE=

Write a summary line in each group of break lines

SUMMARIZE

Suppress the printing of the value of the break variable in the

summary line and of any underlining or overlining in the break lines
in the column containing the break variable

SUPPRESS

Underline each value

UL*

* Traditional SAS monospace output only.

Required Arguments
location

controls the placement of the break lines and is either

AFTER
places the break lines immediately after the last row of each set of rows that have
the same value for the break variable.
BEFORE
places the break lines immediately before the rst row of each set of rows that
have the same value for the break variable.

886

BREAK Statement

Chapter 42

break-variable

is a group or order variable. The REPORT procedure writes break lines each time
the value of this variable changes.

Options
COLOR=color

species the color of the break lines in the REPORT window. You can use the
following colors:
BLACK

MAGENTA

BLUE

ORANGE

BROWN

PINK

CYAN

RED

GRAY

WHITE

GREEN

YELLOW

Default: The color of Foreground in the SASCOLOR window. (For more

information, see the online help for the SASCOLOR window.)

Restriction: This option affects output in the windowing environment only.
Note: Not all operating environments and devices support all colors, and on some
operating systems and devices, one color may map to another color. For example, if
the DEFINITION window displays the word BROWN in yellow characters, then
selecting BROWN results in a yellow item. 4
DOL

Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Featured in: Example 2 on page 951, Example 4 on page 957, Example 5 on page
960, and Example 8 on page 968
STYLE<location(s)>=<style-element-name><[style-attribute-specication(s)]>

species the style element to use for default summary lines that are created with the
BREAK statement. See Using Style Elements in PROC REPORT on page 863 for
details.
Restriction: This option affects only the HTML, RTF, and Printer output.
SUMMARIZE

writes a summary line in each group of break lines. A summary line for a set of
observations contains values for
3 the break variable (which you can suppress with the SUPPRESS option)
3 other group or order variables to the left of the break variable
3 statistics
3 analysis variables
3 computed variables.
The following table shows how PROC REPORT calculates the value for each kind
of report item in a summary line that is created by the BREAK statement:
If the report item is

Then its value is

the break variable

the current value of the variable (or a missing value if

you use SUPPRESS)

a group or order variable to the left of

the break variable

the current value of the variable

a group or order variable to the right

of the break variable, or a display
variable anywhere in the report

missing*

888

BREAK Statement

Chapter 42

If the report item is

Then its value is

a statistic

the value of the statistic over all observations in the set

an analysis variable

the value of the statistic specied as the usage option in

the items denition. PROC REPORT calculates the
value of the statistic over all observations in the set.
The default usage is SUM.

a computed variable

the results of the calculations based on the code in the

corresponding compute block (see COMPUTE
Statement on page 895).

If you reference a variable with a missing value in a customized summary line, then PROC
REPORT displays that variable as a blank (for character variables) or a period (for numeric
variables).

Note: PROC REPORT cannot create groups in a report that contains order or
display variables. 4
Featured in:

Example 2 on page 951, Example 4 on page 957, and Example 9 on

page 971
SUPPRESS

suppresses printing of

3 the value of the break variable in the summary line

3 any underlining and overlining in the break lines in the column that contains
the break variable.
Interaction: If you use SUPPRESS, then the value of the break variable is

unavailable for use in customized break lines unless you assign a value to it in the
compute block that is associated with the break (see COMPUTE Statement on
page 895).
Featured in:

Example 4 on page 957

(for underlining) uses the second formatting character to underline each value

3 that appears in the summary line

3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: hyphen (-)
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: If you specify both the UL and DUL options, then PROC REPORT

honors only UL.

See also: the discussion of FORMCHAR= on page 874.

Order of Break Lines

When a default summary contains more than one break line, the order in which the
break lines appear is
1 overlining or double overlining (OL or DOL)
2 summary line (SUMMARIZE)
3 underlining or double underlining (UL or DUL)

The REPORT Procedure

BY Statement

889

4 skipped line (SKIP)

5 page break (PAGE).

Note: If you dene a customized summary for the break, then customized break
lines appear after underlining or double underlining. For more information about
customized break lines, see COMPUTE Statement on page 895 and LINE Statement
on page 907. 4

BY Statement
Creates a separate report on a separate page for each BY group.
Restriction: If you use the BY statement, then you must use the NOWINDOWS option in
the PROC REPORT statement.
Restriction:

You cannot use the OUT= option when you use a BY statement.

Interaction: If you use the RBREAK statement in a report that uses BY processing, then
PROC REPORT creates a default summary for each BY group. In this case, you cannot
summarize information for the whole report.

Using the BY statement does not make the FIRST. and LAST. variables available
in compute blocks.

Tip:

Main discussion:

BY on page 58

BY <DESCENDING> variable-1
<< DESCENDING> variable-n> <NOTSORTED>;

Required Arguments
variable

species the variable that the procedure uses to form BY groups. You can specify
more than one variable. If you do not use the NOTSORTED option in the BY
statement, then the observations in the data set either must be sorted by all the
variables that you specify or must be indexed appropriately. Variables in a BY
statement are called BY variables.

Options
DESCENDING

species that the data set is sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data are grouped in another way, for example, chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same

890

CALL DEFINE Statement

Chapter 42

values for all BY variables. If observations with the same values for the BY variables
are not contiguous, then the procedure treats each contiguous set as a separate BY
group.

CALL DEFINE Statement

Sets the value of an attribute for a particular column in the current row.
Restriction:

Valid only in a compute block that is attached to a report item.

Featured in:

Example 4 on page 957

CALL DEFINE (column-id | _ROW_ , attribute-name, value);

The CALL DEFINE statement is often used to write report denitions that other
people will use in a windowing environment. Only the FORMAT, URL, URLBP, and
URLP attributes have an effect in the nonwindowing environment. In fact, URL,
URLBP, and URLP are effective only in the nonwindowing environment. The STYLE=
and URL attributes are effective only when you are using the Output Delivery System
to create HTML, RTF, or Printer output. (See Table 42.6 on page 891 for descriptions of
the available attributes.)

Required Arguments
column-id

species a column name or a column number (that is, the position of the column from
the left edge of the report). A column ID can be one of the following:

3 a character literal (in quotation marks) that is the column name

3 a character expression that resolves to the column name
3 a numeric literal that is the column number
3 a numeric expression that resolves to the column number
3 a name of the form _Cn_, where n is the column number
3 the automatic variable _COL_, which identies the column that contains the
report item that the compute block is attached to
attribute-name

is the attribute to dene. For attribute names, refer to Table 42.6 on page 891.
_ROW_

is an automatic variable that indicates the entire current row.

value

sets the value for the attribute. For values for each attribute, refer to Table 42.6 on
page 891.

The REPORT Procedure

Table 42.6

CALL DEFINE Statement

Attribute Descriptions

Attribute

Description

Values

Affects

BLINK

Controls blinking of current

value

1 turns blinking on; 0 turns

it off

windowing environment

COLOR

Controls the color of the current

value in the REPORT window

blue, red, pink, green,

cyan, yellow, white,
orange, black, magenta,
gray, brown

windowing environment

COMMAND

Species that a series of

commands follows

a quoted string of SAS

commands to submit to the
command line

windowing environment

FORMAT

Species a format for the column

a SAS format or a
user-dened format

windowing and
nonwindowing environments

HIGHLIGHT

Controls highlighting of the

current value

1 turns highlighting on; 0

turns it off

windowing environment

RVSVIDEO

Controls display of the current

value

1 turns reverse video on; 0

turns it off

windowing environment

STYLE=

Species the style element for

the Output Delivery System

See Using the STYLE=

Attribute on page 892

HTML, RTF, and Printer

output

URL

Makes the contents of each cell

of the column a link to the
specied Uniform Resource
Locator (URL)*

a quoted URL (either single

or double quotation marks
can be used)

HTML, RTF, and Printer

output

891

892

CALL DEFINE Statement

Chapter 42

Attribute

Description

Values

Affects

URLBP

Makes the contents of each cell

of the column a link. The link
points to a Uniform Resource
Locator that is a concatenation of

a quoted URL (either single

or double quotation marks
can be used)

HTML output

a quoted URL (either single

or double quotation marks
can be used)

HTML output

1 the string that is specied

by the BASE= option in
the ODS HTML statement

2 the string that is specied

by the PATH= option in
the ODS HTML statement

3 the value of the URLBP

*,#

attribute
URLP

Makes the contents of each cell

of the column a link. The link
points to a Uniform Resource
Locator that is a concatenation of

1 the string that is specied

by the PATH= option in
the ODS HTML statement

2 the value of the URLP

*,#

attribute

* The total length of the URL that you specify (including any characters that come from the BASE= and PATH=
options) cannot exceed the line size. Use the LS= option in the PROC REPORT statement to alter the line size
for the PROC REPORT step.
# For information on the BASE= and PATH= options, see the documentation for the ODS HTML statement in
SAS Output Delivery System: Users Guide.

Note: The attributes BLINK, HIGHLIGHT, and RVSVIDEO do not work on all
devices. 4

Using the STYLE= Attribute

The STYLE= attribute species the style element to use in the cells that are affected
by the CALL DEFINE statement.
The STYLE= attribute functions like the STYLE= option in other statements in
PROC REPORT. However, instead of acting as an option in a statement, it becomes the
value for the STYLE= attribute. For instance, the following CALL DEFINE statement
sets the background color to yellow and the font size to 7 for the specied column:
call define(_col_, "style",
"style=[background=yellow font_size=7]");

See Using Style Elements in PROC REPORT on page 863 for details.
Restriction: This option affects only the HTML, RTF, Printer destinations.
Interaction: If you set a style element for the CALLDEF location in the PROC

REPORT statement and you want to use that exact style element in a CALL
DEFINE statement, then use an empty string as the value for the STYLE
attribute, as shown here:
call define (_col_, "STYLE", "" );

The REPORT Procedure

COLUMN Statement

893

Featured in: Example 16 on page 994

COLUMN Statement
Describes the arrangement of all columns and of headers that span more than one column.
Restriction: You cannot use the COLUMN statement if you use REPORT= in the PROC
REPORT statement.
Featured in: Example 1 on page 948, Example 3 on page 954, Example 5 on page 960,
Example 6 on page 964, Example 10 on page 975, and Example 11 on page 977

COLUMN column-specication(s);

Required Arguments
column-specication(s)

is one or more of the following:

3
3
3
3

report-item(s)
report-item-1, report-item-2 <. . . , report-item-n>
(header-1 < . . . header-n > report-item(s) )
report-item=name

where report-item is the name of a data set variable, a computed variable, or a

statistic. See Statistics That Are Available in PROC REPORT on page 857 for a list
of available statistics.
report-item(s)
identies items that each form a column in the report.
Featured in: Example 1 on page 948 and Example 11 on page 977

report-item-1, report-item-2 <. . . , report-item-n>

identies report items that collectively determine the contents of the column or
columns. These items are said to be stacked in the report because each item
generates a header, and the headers are stacked one above the other. The header
for the leftmost item is on top. If one of the items is an analysis variable, a
computed variable, a group variable, or a statistic, then its values ll the cells in
that part of the report. Otherwise, PROC REPORT lls the cells with frequency
counts.
If you stack a statistic with an analysis variable, then the statistic that you
name in the column statement overrides the statistic in the denition of the
analysis variable. For example, the following PROC REPORT step produces a
report that contains the minimum value of Sales for each sector:
proc report data=grocery;
column sector sales,min;
define sector/group;
define sales/analysis sum;
run;

894

COLUMN Statement

Chapter 42

If you stack a display variable under an across variable, then all the values of
that display variable appear in the report.
Interaction: A series of stacked report items can include only one analysis

variable or statistic. If you include more than one analysis variable or statistic,
then PROC REPORT returns an error because it cannot determine which
values to put in the cells of the report.
Tip: You can use parentheses to group report items whose headers should appear

at the same level rather than stacked one above the other.
Featured in: Example 5 on page 960, Example 6 on page 964, and Example 10 on

page 975
(header-1 < header-n > report-item(s))
creates one or more headers that span multiple columns.
header
is a string of characters that spans one or more columns in the report. PROC
REPORT prints each header on a separate line. You can use split characters in
a header to split one header over multiple lines. See the discussion of SPLIT=
on page 883.
In traditional (monospace) SAS output, if the rst and last characters of a
header are one of the following characters, then PROC REPORT uses that
character to expand the header to ll the space over the column or columns:
: = \_ .* +

Similarly, if the rst character of a header is < and the last character is >, or
vice-versa, then PROC REPORT expands the header to ll the space over the
column by repeating the rst character before the text of the header and the
last character after it.
Note: A hexadecimal value (such as DFx) that is specied within header
will not resolve because it is specied within quotation marks. To resolve a
hexadecimal value, use the %sysfunc(byte(num)) function, where num is the
hexadecimal value. Be sure to enclose header in double quotation marks (" ") so
that the macro function will resolve. 4
report-item(s)
species the columns to span.
Featured in: Example 10 on page 975

report-item=name
species an alias for a report item. You can use the same report item more than
once in a COLUMN statement. However, you can use only one DEFINE statement
for any given name. (The DEFINE statement designates characteristics such as
formats and customized column headers. If you omit a DEFINE statement for an
item, then the REPORT procedure uses defaults.) Assigning an alias in the
COLUMN statement does not by itself alter the report. However, it does enable you
to use separate DEFINE statements for each occurrence of a variable or statistic.
Featured in: Example 3 on page 954

Note: You cannot always use an alias. When you refer in a compute block to a report
item that has an alias, you must usually use the alias. However, if the report item
shares a column with an across variable, then you must reference the column by column
number (see Four Ways to Reference Report Items in a Compute Block on page 859). 4

The REPORT Procedure

COMPUTE Statement

895

COMPUTE Statement
Starts a compute block. A compute block contains one or more programming statements that
PROC REPORT executes as it builds the report.
Interaction: An ENDCOMP statement must mark the end of the group of statements in
the compute block.
Featured in: Example 2 on page 951, Example 3 on page 954, Example 4 on page 957,
Example 5 on page 960, Example 9 on page 971, and Example 10 on page 975

COMPUTE location <target>

</ STYLE=<style-element-name>
<[style-attribute-specication(s)]>>;
LINE specication(s);
. . . select SAS language elements . . .
ENDCOMP;
COMPUTE report-item </ type-specication>;
CALL DEFINE (column-id, attribute-name, value);
. . . select SAS language elements . . .
ENDCOMP;
A compute block can be associated with a report item or with a location (at the top or
bottom of a report; at the top or bottom of a page; before or after a set of observations).
You create a compute block with the COMPUTE window or with the COMPUTE
statement. One form of the COMPUTE statement associates the compute block with a
report item. Another form associates the compute block with a location.
For a list of the SAS language elements that you can use in compute blocks, see The
Contents of Compute Blocks on page 859.

Required Arguments
You must specify either a location or a report item in the COMPUTE statement.
location

determines where the compute block executes in relation to target.

AFTER
executes the compute block at a break in one of the following places:

3 immediately after the last row of a set of rows that have the same value for
the variable that you specify as target or, if there is a default summary on
that variable, immediately after the creation of the preliminary summary line
(see How PROC REPORT Builds a Report on page 936).

3 except in Printer and RTF output, near the bottom of each page, immediately
before any footnotes, if you specify _PAGE_ as target.

3 at the end of the report if you omit a target.

BEFORE
executes the compute block at a break in one of the following places:

3 immediately before the rst row of a set of rows that have the same value for
the variable that you specify as target or, if there is a default summary on

896

COMPUTE Statement

Chapter 42

that variable, immediately after the creation of the preliminary summary line
(see How PROC REPORT Builds a Report on page 936).

3 except in Printer and RTF output, near the top of each page, between any
titles and the column headings, if you specify _PAGE_ as target.

3 immediately before the rst detail row if you omit a target.

Note: If a report contains more columns than will t on a printed page, then
PROC REPORT generates an additional page or pages to contain the remaining
columns. In this case, when you specify _PAGE_ as target, the COMPUTE block does
NOT re-execute for each of these additional pages; the COMPUTE block re-executes
only after all columns have been printed. 4
Featured in:

Example 3 on page 954 and Example 9 on page 971

report-item

species a data set variable, a computed variable, or a statistic to associate the

compute block with. If you are working in the nonwindowing environment, then you
must include the report item in the COLUMN statement. If the item is a computed
variable, then you must include a DEFINE statement for it.
Featured in:

Example 4 on page 957 and Example 5 on page 960

Note: The position of a computed variable is important. PROC REPORT assigns

values to the columns in a row of a report from left to right. Consequently, you cannot
base the calculation of a computed variable on any variable that appears to its right in
the report. 4

Options
STYLE<(location(s))>=<style-element-name><[style-attribute-specication(s)]>

species the style to use for the text that is created by any LINE statements in this
compute block. See Using Style Elements in PROC REPORT on page 863 for
details.
Restriction: This option affects only the HTML, RTF, and Printer destinations.
Featured in:

Example 16 on page 994

target

controls when the compute block executes. If you specify a location (BEFORE or
AFTER) for the COMPUTE statement, then you can also specify target, which can be
one of the following:
break-variable
is a group or order variable.
When you specify a break variable, PROC REPORT executes the statements in
the compute block each time the value of the break variable changes.
_PAGE_ </ justication>
except in Printer and RTF output, causes the compute block to execute once for
each page, either immediately after printing any titles or immediately before
printing any footnotes. justication controls the placement of text and values. It
can be one of the following:
CENTER

centers each line that the compute block writes.

LEFT

left-justies each line that the compute block writes.

RIGHT

right-justies each line that the compute block writes.

Default: CENTER

The REPORT Procedure

Featured in:

DEFINE Statement

897

Example 9 on page 971

type-specication

species the type and, optionally, the length of report-item. If the report item that is
associated with a compute block is a computed variable, then PROC REPORT
assumes that it is a numeric variable unless you use a type specication to specify
that it is a character variable. A type specication has the form
CHARACTER < LENGTH=length>
where
CHARACTER
species that the computed variable is a character variable. If you do not specify a
length, then the variables length is 8.
Alias: CHAR
Featured in: Example 10 on page 975
LENGTH=length
species the length of a computed character variable.
Default: 8
Range: 1 to 200
Interaction: If you specify a length, then you must use CHARACTER to indicate
that the computed variable is a character variable.
Featured in: Example 10 on page 975

DEFINE Statement
Describes how to use and display a report item.
If you do not use a DEFINE statement, then PROC REPORT uses default
characteristics.
Featured in: Example 2 on page 951, Example 3 on page 954, Example 4 on page 957,
Example 5 on page 960, Example 6 on page 964, Example 9 on page 971, and Example
10 on page 975
Tip:

DEFINE report-item / <option(s)>;

To do this

Use this option

Specify how to use a report item (see Usage of Variables in a Report on page 853)
Dene the item, which must be a data set variable, as an
across variable

ACROSS

Dene the item, which must be a data set variable, as an

analysis variable

ANALYSIS

Dene the item as a computed variable

COMPUTED

Dene the item, which must be a data set variable, as a

display variable

DISPLAY

898

DEFINE Statement

Chapter 42

To do this

Use this option

Dene the item, which must be a data set variable, as a

group variable

GROUP

Dene the item, which must be a data set variable, as an

order variable

ORDER

Customize the appearance of a report item

Exclude all combinations of the item that are not found in
the preloaded range of user-dened formats

EXCLUSIVE

Assign a SAS or user-dened format to the item

FORMAT=

Reference a HELP or CBT entry that contains Help

information for the report item

ITEMHELP=

Consider missing values as valid values for the item

MISSING

Order the values of a group, order, or across variable

according to the specied order

ORDER=

Specify that all formats are preloaded for the item.

PRELOADFMT

For traditional SAS monospace output, dene the number

of blank characters to leave between the column being
dened and the column immediately to its left

SPACING=

Associate a statistic with an analysis variable

statistic

Specify a style element (for the Output Delivery System)

for the report item

STYLE=

Specify a numeric variable whose values weight the value

of the analysis variable

WEIGHT=

Dene the width of the column in which PROC REPORT

displays the report item

WIDTH=

Specify options for a report item

Reverse the order in which PROC REPORT displays rows
or values of a group, order, or across variable

DESCENDING

Wrap the value of a character variable in its column

FLOW

Specify that the item that you are dening is an ID

variable

Suppress the display of the report item

NOPRINT

Suppress the display of the report item if its values are

all zero or missing

NOZERO

Insert a page break just before printing the rst column

containing values of the report item

PAGE

Control the placement of values and column headers

Center the formatted values of the report item within the
column width and center the column header over the
values

CENTER

Left-justify the formatted values of the report item within

the column width and left-justify the column headers over
the values

LEFT

The REPORT Procedure

To do this

DEFINE Statement

899

Use this option

Right-justify the formatted values of the report item

within the column width and right-justify the column
headers over the values

RIGHT

Specify the color in the REPORT window of the column

header and of the values of the item that you dene

COLOR=

Dene the column header for the report item

column-header

Required Arguments
report-item

species the name or alias (established in the COLUMN statement) of the data set
variable, computed variable, or statistic to dene.
Note: Do not specify a usage option in the denition of a statistic. The name of the
statistic tells PROC REPORT how to use it. 4

Options
ACROSS

denes report-item, which must be a data set variable, as an across variable. (See
Across Variables on page 854.)
Featured in: Example 5 on page 960
ANALYSIS

denes report-item, which must be a data set variable, as an analysis variable. (See
Analysis Variables on page 855.)
By default, PROC REPORT calculates the Sum statistic for an analysis variable.
Specify an alternate statistic with the statistic option in the DEFINE statement.
Note: Naming a statistic in the DEFINE statement implies the ANALYSIS
option, so you never need to specify ANALYSIS. However, specifying ANALYSIS may
make your code easier for novice users to understand. 4
Featured in: Example 2 on page 951, Example 3 on page 954, and Example 4 on
page 957
CENTER

centers the formatted values of the report item within the column width and centers
the column header over the values. This option has no effect on the CENTER option
in the PROC REPORT statement, which centers the report on the page.
COLOR=color

species the color in the REPORT window of the column header and of the values of
the item that you are dening. You can use the following colors:
BLACK

MAGENTA

BLUE

ORANGE

BROWN

PINK

CYAN

RED

900

DEFINE Statement

Chapter 42

GRAY

WHITE

GREEN

YELLOW

Default: The color of Foreground in the SASCOLOR window. (For more

information, see the online Help for the SASCOLOR window.)

Restriction: This option affects output in the windowing environment only.

Note: Not all operating environments and devices support all colors, and in some
operating environments and devices, one color may map to another color. For
example, if the DEFINITION window displays the word BROWN in yellow
characters, then selecting BROWN results in a yellow item. 4
column-header

denes the column header for the report item. Enclose each header in single or
double quotation marks. When you specify multiple column headers, PROC REPORT
uses a separate line for each one. The split character also splits a column header
over multiple lines.
In traditional (monospace) SAS output, if the rst and last characters of a heading
are one of the following characters, then PROC REPORT uses that character to
expand the heading to ll the space over the column:
: = \_ .* +

Similarly, if the rst character of a header is < and the last character is >, or
vice-versa, then PROC REPORT expands the header to ll the space over the column
by repeating the rst character before the text of the header and the last character
after it.
Note: A hexadecimal value (such as DFx) that is specied within
column-header will not resolve because it is specied within quotation marks. To
resolve a hexadecimal value, use the %sysfunc(byte(num)) function, where num is
the hexadecimal value. Be sure to enclose column-header in double quotation marks
(" ") so that the macro function will resolve. 4
Default:

Item

Header

variable without a label

variable name

variable with a label

variable label

statistic

statistic name

If you want to use names when labels exist, then submit the following SAS
statement before invoking PROC REPORT:

Tip:

options nolabel;

HEADLINE underlines all column headers and the spaces between them. In
traditional (monospace) SAS output, you can underline column headers without
underlining the spaces between them, by using the special characters -- as the
last line of each column header instead of using HEADLINE (see Example 4 on
page 957).

Tip:

The REPORT Procedure

DEFINE Statement

901

Example 3 on page 954, Example 4 on page 957, and Example 5 on

page 960
COMPUTED

denes the specied item as a computed variable. Computed variables are variables
that you dene for the report. They are not in the input data set, and PROC
REPORT does not add them to the input data set.
In the windowing environment, you add a computed variable to a report from the
COMPUTED VAR window.
In the nonwindowing environment, you add a computed variable by

3 including the computed variable in the COLUMN statement

3 dening the variables usage as COMPUTED in the DEFINE statement
3 computing the value of the variable in a compute block associated with the
variable.
Featured in:

Example 5 on page 960 and Example 10 on page 975

DESCENDING

reverses the order in which PROC REPORT displays rows or values of a group, order,
or across variable.
By default, PROC REPORT orders group, order, and across variables by their
formatted values. Use the ORDER= option in the DEFINE statement to specify an
alternate sort order.

Tip:

DISPLAY

denes report-item, which must be a data set variable, as a display variable. (See
Display Variables on page 853.)
EXCLUSIVE

excludes from the report and the output data set all combinations of the group
variables and the across variables that are not found in the preloaded range of
user-dened formats.
Requirement: You must specify the PRELOADFMT option in the DEFINE
statement in order to preload the variable formats.
FLOW

wraps the value of a character variable in its column. The FLOW option honors the
split character. If the text contains no split character, then PROC REPORT tries to
split text at a blank.
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Featured in:

Example 10 on page 975

FORMAT=format

assigns a SAS or user-dened format to the item. This format applies to report-item
as PROC REPORT displays it; the format does not alter the format associated with a
variable in the data set. For data set variables, PROC REPORT honors the rst of
these formats that it nds:

3 the format that is assigned with FORMAT= in the DEFINE statement

3 the format that is assigned in a FORMAT statement when you invoke PROC
REPORT
3 the format that is associated with the variable in the data set.
If none of these is present, then PROC REPORT uses BESTw. for numeric
variables and $w. for character variables. The value of w is the default column

902

DEFINE Statement

Chapter 42

width. For character variables in the input data set, the default column width is the
variables length. For numeric variables in the input data set and for computed
variables (both numeric and character), the default column width is the value
specied by COLWIDTH= in the PROC REPORT statement or in the ROPTIONS
window.
In the windowing environment, if you are unsure what format to use, then type a
question mark (?) in the format eld in the DEFINITION window to access the
FORMATS window.
Featured in: Example 2 on page 951 and Example 6 on page 964
GROUP

denes report-item, which must be a data set variable, as a group variable. (See
Group Variables on page 854.)
Featured in: Example 4 on page 957, Example 6 on page 964, and Example 14 on
page 986
ID

species that the item that you are dening is an ID variable. An ID variable and all
columns to its left appear at the left of every page of a report. ID ensures that you
can identify each row of the report when the report contains more columns than will
t on one page.
Featured in: Example 6 on page 964
ITEMHELP=entry-name

references a HELP or CBT entry that contains help information for the report item.
Use PROC BUILD in SAS/AF software to create a HELP or CBT entry for a report
item. All HELP and CBT entries for a report must be in the same catalog, and you
must specify that catalog with the HELP= option in the PROC REPORT statement
or from the User Help elds in the ROPTIONS window.
Of course, you can access these entries only from a windowing environment. To
access a Help entry from the report, select the item and issue the HELP command.
PROC REPORT rst searches for and displays an entry named entry-name.CBT. If
no such entry exists, then PROC REPORT searches for entry-name.HELP. If neither
a CBT nor a HELP entry for the selected item exists, then the opening frame of the
Help for PROC REPORT is displayed.
LEFT

left-justies the formatted values of the report item within the column width and
left-justies the column headers over the values. If the format width is the same as
the width of the column, then the LEFT option has no effect on the placement of
values.
MISSING

considers missing values as valid values for the report item. Special missing values
that represent numeric values (the letters A through Z and the underscore (_)
character) are each considered as a separate value.
Default: If you omit the MISSING option, then PROC REPORT excludes from the
report and the output data sets all observations that have a missing value for any
group, order, or across variable.
NOPRINT

suppresses the display of the report item. Use this option

3 if you do not want to show the item in the report but you need to use its values
to calculate other values that you use in the report
3 to establish the order of rows in the report
3 if you do not want to use the item as a column but want to have access to its
values in summaries (see Example 9 on page 971).

The REPORT Procedure

DEFINE Statement

903

Interaction: Even though the columns that you dene with NOPRINT do not

appear in the report, you must count them when you are referencing columns by
number (see Four Ways to Reference Report Items in a Compute Block on page
859).
Interaction: SHOWALL in the PROC REPORT statement or the ROPTIONS
window overrides all occurrences of NOPRINT.
Featured in: Example 3 on page 954, Example 9 on page 971, and Example 12 on
page 980
NOZERO

suppresses the display of the report item if its values are all zero or missing.
Interaction: Even though the columns that you dene with NOZERO do not appear
in the report, you must count them when you are referencing columns by number
(see Four Ways to Reference Report Items in a Compute Block on page 859).
Interaction: SHOWALL in the PROC REPORT statement or in the ROPTIONS
window overrides all occurrences of NOZERO.
ORDER

denes report-item, which must be a data set variable, as an order variable. (See
Order Variables on page 853.)
Featured in: Example 2 on page 951
ORDER=DATA|FORMATTED|FREQ|INTERNAL

orders the values of a group, order, or across variable according to the specied order,
where
DATA
orders values according to their order in the input data set.
FORMATTED
orders values by their formatted (external) values. If no format has been assigned
to a class variable, then the default format, BEST12., is used.
FREQ
orders values by ascending frequency count.
INTERNAL
orders values by their unformatted values, which yields the same order that PROC
SORT would yield. This order is operating environment-dependent. This sort
sequence is particularly useful for displaying dates chronologically.
Default: FORMATTED
Interaction: DESCENDING in the items denition reverses the sort sequence for
an item. By default, the order is ascending.
Featured in: Example 2 on page 951
Note: The default value for the ORDER= option in PROC REPORT is not the
same as the default value in other SAS procedures. In other SAS procedures, the
default is ORDER=INTERNAL. The default for the option in PROC REPORT may
change in a future release to be consistent with other procedures. Therefore, in
production jobs where it is important to order report items by their formatted values,
specify ORDER=FORMATTED even though it is currently the default. Doing so
ensures that PROC REPORT will continue to produce the reports you expect even if
the default changes. 4
PAGE

inserts a page break just before printing the rst column containing values of the
report item.

904

DEFINE Statement

Chapter 42

Interaction: PAGE is ignored if you use WRAP in the PROC REPORT statement or

in the ROPTIONS window.

PRELOADFMT

species that the format is preloaded for the variable.

Restriction: PRELOADFMT applies only to group and across variables.
Requirement: PRELOADFMT has no effect unless you specify either EXCLUSIVE
or ORDER=DATA and you assign a format to the variable.
Interaction: To limit the report to the combination of formatted variable values
that are present in the input data set, use the EXCLUSIVE option in the DEFINE
statement.
Interaction To include all ranges and values of the user-dened formats in the
output, use the COMPLETEROWS option in the PROC REPORT statement.
Note: If you do not specify NOCOMPLETECOLS when you dene the across
variables, then the report includes a column for every formatted variable. If you
specify COMPLETEROWS when you dene the group variables, then the report
includes a row for every formatted value. Some combinations of rows and columns
might not make sense when the report includes a column for every formatted value
of the across variable and a row for every formatted value of the group variable. 4
RIGHT

right-justies the formatted values of the specied item within the column width and
right-justies the column headers over the values. If the format width is the same as
the width of the column, then RIGHT has no effect on the placement of values.
SPACING=horizontal-positions

denes the number of blank characters to leave between the column being dened
and the column immediately to its left. For each column, the sum of its width and
the blank characters between it and the column to its left cannot exceed the line size.
Default: 2
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Interaction: When PROC REPORTs CENTER option is in effect, PROC REPORT
ignores spacing that precedes the leftmost variable in the report.
Interaction: SPACING= in an items denition overrides the value of SPACING= in
the PROC REPORT statement or in the ROPTIONS window.
statistic

associates a statistic with an analysis variable. You must associate a statistic with
every analysis variable in its denition. PROC REPORT uses the statistic that you
specify to calculate values for the analysis variable for the observations that are
represented by each cell of the report. You cannot use statistic in the denition of any
other kind of variable.
See Statistics That Are Available in PROC REPORT on page 857 for a list of
available statistics.
Default: SUM
Featured in: Example 2 on page 951, Example 3 on page 954, and Example 4 on
page 957
Note: PROC REPORT uses the name of the analysis variable as the default
header for the column. You can customize the column header with the column-header
option in the DEFINE statement. 4
STYLE<(location(s))>=<style-element-name><[style-attribute-specication(s)]>

species the style element to use for column headers and for text inside cells for this
report item. See Using Style Elements in PROC REPORT on page 863 for details.

The REPORT Procedure

DEFINE Statement

905

Restriction: This option affects only the HTML, RTF, and Printer destinations.
Featured in:

Example 16 on page 994

WEIGHT=weight-variable

species a numeric variable whose values weight the values of the analysis variable
that is specied in the DEFINE statement. The variable value does not have to be an
integer. The following table describes how PROC REPORT treats various values of
the WEIGHT variable.
Weight
Value

PROC REPORT Response

counts the observation in the total number of observations

less than 0

converts the value to zero and counts the observation in the total number of
observations

missing

excludes the observation

To exclude observations that contain negative and zero weights from the analysis,
use the EXCLNPWGT option in the PROC REPORT statement. Note that most
SAS/STAT procedures, such as PROC GLM, exclude negative and zero weights by
default.
Restriction: to compute weighted quantiles, use QMETHOD=OS in the PROC
REPORT statement.
Tip: When you use the WEIGHT= option, consider which value of the VARDEF=
option in the PROC REPORT statement is appropriate.
Tip: Use the WEIGHT= option in separate variable denitions in order to specify
different weights for the variables.
Note: Prior to Version 7 of SAS, the REPORT procedure did not exclude the
observations with missing weights from the count of observations. 4
WIDTH=column-width

denes the width of the column in which PROC REPORT displays report-item.
Default: A column width that is just large enough to handle the format. If there is
no format, then PROC REPORT uses the value of the COLWIDTH= option in the
PROC REPORT statement.
Range: 1 to the value of the SAS system option LINESIZE=
Restriction: This option has no effect on ODS destinations other than traditional
SAS monospace output.
Interaction: WIDTH= in an item denition overrides the value of COLWIDTH= in
the PROC REPORT statement or the ROPTIONS window.
Tip: When you stack items in the same column in a report, the width of the item
that is at the bottom of the stack determines the width of the column.
Featured in: Example 10 on page 975

906

ENDCOMP Statement

Chapter 42

ENDCOMP Statement
Marks the end of one or more programming statements that PROC REPORT executes as it builds
the report.
Restriction:

A COMPUTE statement must precede the ENDCOMP statement.

ENDCOMP;

Example 2 on page 951

FREQ Statement
Treats observations as if they appear multiple times in the input data set.
Tip: The effects of the FREQ and WEIGHT statements are similar except when
calculating degrees of freedom.
See also: For an example that uses the FREQ statement, see Example on page 62

FREQ variable;

Required Arguments
variable

species a numeric variable whose value represents the frequency of the observation.
If you use the FREQ statement, then the procedure assumes that each observation
represents n observations, where n is the value of variable. If n is not an integer,
then SAS truncates it. If n is less than 1 or is missing, then the procedure does not
use that observation to calculate statistics.

Frequency Information Is Not Saved

When you store a report denition, PROC REPORT does not store the FREQ
statement.

The REPORT Procedure

LINE Statement

907

LINE Statement
Provides a subset of the features of the PUT statement for writing customized summaries.
Restriction: This statement is valid only in a compute block that is associated with a
location in the report.
Restriction: You cannot use the LINE statement in conditional statements (IF-THEN,
IF-THEN/ELSE, and SELECT) because it is not executed until PROC REPORT has
executed all other statements in the compute block.
Featured in:

Example 2 on page 951, Example 3 on page 954, and Example 9 on page 971

LINE specication(s);

Required Arguments
specication(s)

can have one of the following forms. You can mix different forms of specications in
one LINE statement.
item item-format
species the item to display and the format to use to display it, where
item
is the name of a data set variable, a computed variable, or a statistic in the
report. For information about referencing report items see Four Ways to
Reference Report Items in a Compute Block on page 859.
item-format
is a SAS format or user-dened format. You must specify a format for each item.
Featured in: Example 2 on page 951

character-string
species a string of text to display. When the string is a blank and nothing else is
in specication(s), PROC REPORT prints a blank line.
Note: A hexadecimal value (such as DFx) that is specied within
character-string will not resolve because it is specied within quotation marks. To
resolve a hexadecimal value, use the %sysfunc(byte(num)) function, where num
is the hexadecimal value. Be sure to enclose character-string in double quotation
marks (" ") so that the macro function will resolve. 4
Featured in: Example 2 on page 951

number-of-repetitions*character-string
species a character string and the number of times to repeat it.
Featured in: Example 3 on page 954

pointer-control
species the column in which PROC REPORT displays the next specication. You
can use either of the following forms for pointer controls:
@column-number
species the number of the column in which to begin displaying the next item in
the specication list.

908

RBREAK Statement

Chapter 42

+column-increment
species the number of columns to skip before beginning to display the next
item in the specication list.
Both column-number and column-increment can be either a variable or a literal
value.
Restriction: The pointer controls are designed for monospace output. They have
no effect on the HTML, RTF, or Printer output.
Featured in: Example 3 on page 954 and Example 5 on page 960

Differences between the LINE and PUT Statements

The LINE statement does not support the following features of the PUT statement:
3 automatic labeling signaled by an equals sign (=), also known as named output
3 the _ALL_, _INFILE_, and _PAGE_ arguments and the OVERPRINT option
3 grouping items and formats to apply one format to a list of items
3 pointer control using expressions
3 line pointer controls (# and /)
3 trailing at signs (@ and @@)
3 format modiers
3 array elements.

RBREAK Statement
Produces a default summary at the beginning or end of a report or at the beginning or end of each
BY group.
Featured in:

Example 1 on page 948 and Example 10 on page 975

RBREAK location </ option(s)>;

To do this

Use this option

Specify the color of the break lines in the REPORT window

COLOR=

Double overline each value

DOL*

Double underline each value

DUL*

Overline each value

OL*

Start a new page after the last break line of a break located at the
beginning of the report

PAGE

Write a blank line for the last break line of a break located at the
beginning of the report

SKIP*

Specify a style element (for the Output Delivery System) for

default summary lines, customized summary lines, or both

STYLE=

Include a summary line as one of the break lines

SUMMARIZE

Underline each value

UL*

* Traditional SAS monospace output only.

The REPORT Procedure

RBREAK Statement

909

Required Arguments
location

controls the placement of the break lines and is either of the following:
AFTER
places the break lines at the end of the report.
BEFORE
places the break lines at the beginning of the report.

Options
COLOR=color

species the color of the break lines in the REPORT window. You can use the
following colors:
BLACK

MAGENTA

BLUE

ORANGE

BROWN

PINK

CYAN

RED

GRAY

WHITE

GREEN

YELLOW

Default: The color of Foreground in the SASCOLOR window. (For more

information, see the online Help for the SASCOLOR window.)

Restriction: This option affects output in the windowing environment only.

(for double overlining) uses the thirteenth formatting character to overline each value

3 that appears in the summary line

3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: equals sign (=)
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: If you specify both the OL and DOL options, then PROC REPORT

honors only OL.

See also: the discussion of FORMCHAR= on page 874.

910

RBREAK Statement

Chapter 42

Featured in:

Example 1 on page 948

DUL

(for double underlining) uses the thirteenth formatting character to underline each
value

3 that appears in the summary line

3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: equals sign (=)
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: If you specify both the UL and DUL options, then PROC REPORT

honors only UL.

See also: the discussion of FORMCHAR= on page 874.
OL

(for overlining) uses the second formatting character to overline each value

3 that appears in the summary line

3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: hyphen (-)
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: If you specify both the OL and DOL options, then PROC REPORT

honors only OL.

See also: the discussion of FORMCHAR= on page 874.
Featured in:

Example 10 on page 975

PAGE

starts a new page after the last break line of a break located at the beginning of the
report.
SKIP

writes a blank line after the last break line of a break located at the beginning of the
report.
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

STYLE<(location(s))>=<style-element-name><[style-attribute-specication(s)]>

species the style element to use for default summary lines that are created with the
RBREAK statement. See Using Style Elements in PROC REPORT on page 863 for
details.
Restriction: This option affects only the HTML, RTF, and Printer destinations.
SUMMARIZE

includes a summary line as one of the break lines. A summary line at the beginning
or end of a report contains values for

3 statistics
3 analysis variables
3 computed variables.
The following table shows how PROC REPORT calculates the value for each kind
of report item in a summary line created by the RBREAK statement:

The REPORT Procedure

RBREAK Statement

If the report item is

Then its value is

a statistic

the value of the statistic over all observations in

the set

an analysis variable

the value of the statistic specied as the usage

option in the DEFINE statement. PROC REPORT
calculates the value of the statistic over all
observations in the set. The default usage is SUM.

a computed variable

911

the results of the calculations based on the code in

the corresponding compute block (see COMPUTE
Statement on page 895).

Featured in:

Example 1 on page 948 and Example 10 on page 975

(for underlining) uses the second formatting character to underline each value
3 that appears in the summary line
3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: hyphen (-)
Restriction: This option has no effect on ODS destinations other than traditional

SAS monospace output.

Interaction: If you specify both the UL and DUL options, then PROC REPORT
honors only UL.
See also: the discussion of FORMCHAR= on page 874.

Order of Break Lines

When a default summary contains more than one break line, the order in which the
break lines appear is
1 overlining or double overlining (OL or DOL, traditional SAS monospace output

only)
2 summary line (SUMMARIZE)
3 underlining or double underlining (UL or DUL, traditional SAS monospace output

only)
4 skipped line (SKIP, traditional SAS monospace output only)
5 page break (PAGE).

912

WEIGHT Statement

Chapter 42

WEIGHT Statement
Species weights for analysis variables in the statistical calculations.
See also: For information about calculating weighted statistics see Calculating
Weighted Statistics on page 64. For an example that uses the WEIGHT statement, see
Weighted Statistics Example on page 65.

WEIGHT variable;

Required Arguments
variable

species a numeric variable whose values weight the values of the analysis variables.
The value of the variable does not have to be an integer. If the value of variable is
Weight value

PROC REPORT

counts the observation in the total number of observations

less than 0

converts the value to zero and counts the observation in the

total number of observations

missing

excludes the observation

To exclude observations that contain negative and zero weights from the analysis,
use EXCLNPWGT. Note that most SAS/STAT procedures, such as PROC GLM,
exclude negative and zero weights by default.
When you use the WEIGHT statement, consider which value of the VARDEF=
option is appropriate. See VARDEF= on page 883 and the calculation of weighted
statistics in Keywords and Formulas on page 1340 for more information.

Tip:

Note: Prior to Version 7 of SAS, the procedure did not exclude the observations
with missing weights from the count of observations. 4

Weight Information Is Not Saved

When you store a report denition, PROC REPORT does not store the WEIGHT
statement.

REPORT Procedure Windows

The windowing environment in PROC REPORT provides essentially the same
functionality as the statements, with one major exception: you cannot use the Output
Delivery System from the windowing environment.

The REPORT Procedure

BREAK

913

BREAK
Controls PROC REPORTs actions at a change in the value of a group or order variable or at the
top or bottom of a report.

Path
Edit

Summarize information

After you select Summarize Information, PROC REPORT offers you four choices for
the location of the break:

3
3
3
3

Before Item
After Item
At the top
At the bottom.

After you select a location, the BREAK window opens.

Note: To create a break before or after detail lines (when the value of a group or
order variable changes), you must select a variable before you open the BREAK
window. 4

Description

Note: For information about changing the formatting characters that are used by the
line drawing options in this window, see the discussion of FORMCHAR= on page 874. 4

Options
Overline summary

uses the second formatting character to overline each value

914

BREAK

Chapter 42

3 that appears in the summary line

3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: hyphen (-)
Interaction: If you specify options to overline and to double overline, then PROC

REPORT overlines.
Double overline summary

uses the thirteenth formatting character to overline each value

3 that appears in the summary line
3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: equals sign (=)
Interaction: If you specify options to overline and to double overline, then PROC

REPORT overlines.
Underline summary

uses the second formatting character to underline each value

3 that appears in the summary line
3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: hyphen (-)
Interaction: If you specify options to underline and to double underline, then

PROC REPORT underlines.

Double underline summary

uses the thirteenth formatting character to underline each value

3 that appears in the summary line
3 that would appear in the summary line if you specied the SUMMARIZE option.
Default: equals sign (=)
Interaction: If you specify options to underline and to double underline, then

PROC REPORT underlines.

Skip line after break

writes a blank line for the last break line.

This option has no effect if you use it in a break at the end of a report.
Page after break

starts a new page after the last break line. This option has no effect in a break at the
end of a report.
Interaction: If you use this option in a break on a variable and you create a break at
the end of the report, then the summary for the whole report is on a separate page.
Summarize analysis columns

writes a summary line in each group of break lines. A summary line contains values
for
3 statistics
3 analysis variables
3 computed variables.
A summary line between sets of observations also contains
3 the break variable (which you can suppress with Suppress break value)
3 other group or order variables to the left of the break variable.
The following table shows how PROC REPORT calculates the value for each kind
of report item in a summary line created by the BREAK window:

The REPORT Procedure

BREAK

If the report item is

Then its value is

the break variable

the current value of the variable (or a missing value if

you select suppress break value)

a group or order variable to the left of

the break variable

the current value of the variable

a group or order variable to the right of

the break variable, or a display
variable anywhere in the report

missing*

a statistic

the value of the statistic over all observations in the

set

an analysis variable

the value of the statistic specied as the usage option

in the items denition. PROC REPORT calculates the
value of the statistic over all observations in the set.
The default usage is SUM.

a computed variable

915

the results of the calculations based on the code in the

corresponding compute block (see COMPUTE
Statement on page 895).

If you reference a variable with a missing value in a customized summary line, then PROC
REPORT displays that variable as a blank (for character variables) or a period (for numeric
variables).

Suppress break value

suppresses printing of

3 the value of the break variable in the summary line

3 any underlining and overlining in the break lines in the column containing the
break variable.
If you select Suppress break value, then the value of the break variable is
unavailable for use in customized break lines unless you assign it a value in the
compute block that is associated with the break.

Color
From the list of colors, select the one to use in the REPORT window for the column
header and the values of the item that you are dening.
Default: The color of Foreground in the SASCOLOR window. (For more

information, see the online Help for the SASCOLOR window.)

Not all operating environments and devices support all colors, and in some
operating environments and devices, one color may map to another color. For
example, if the DEFINITION window displays the word BROWN in yellow
characters, then selecting BROWN results in a yellow item.

Note:

916

COMPUTE

Chapter 42

Buttons
Edit Program

opens the COMPUTE window and enables you to associate a compute block with a
location in the report.
OK

applies the information in the BREAK window to the report and closes the window.
Cancel

closes the BREAK window without applying information to the report.

COMPUTE
Attaches a compute block to a report item or to a location in the report. Use the SAS Text Editor
commands to manipulate text in this window.

Path
From Edit Program in the COMPUTED VAR, DEFINITION, or BREAK window.

Description
For information about the SAS language features that you can use in the COMPUTE
window, see The Contents of Compute Blocks on page 859.

COMPUTED VAR
Adds a variable that is not in the input data set to the report.

Path
Select a column. Then select
Edit

Add Item

Computed Column

After you select Computed Column, PROC REPORT prompts you for the location of
the computed column relative to the column that you have selected. After you select a
location, the COMPUTED VAR window opens.

Description
Enter the name of the variable at the prompt. If it is a character variable, then
select the Character data check box and, if you want, enter a value in the Length
eld. The length can be any integer between 1 and 200. If you leave the eld blank,
then PROC REPORT assigns a length of 8 to the variable.

The REPORT Procedure

DATA SELECTION

917

After you enter the name of the variable, select Edit Program to open the COMPUTE
window. Use programming statements in the COMPUTE window to dene the
computed variable. After closing the COMPUTE and COMPUTED VAR windows, open
the DEFINITION window to describe how to display the computed variable.
Note: The position of a computed variable is important. PROC REPORT assigns
values to the columns in a row of a report from left to right. Consequently, you cannot
base the calculation of a computed variable on any variable that appears to its right in
the report. 4

DATA COLUMNS
Lists all variables in the input data set so that you can add one or more data set variables to the
report.

Path
Select a report item. Then select
Edit

Add Item

Data Column

After you select Data column, PROC REPORT prompts you for the location of the
computed column relative to the column that you have selected. After you select a
location, the DATA COLUMNS window opens.

Description
Select one or more variables to add to the report. When you select the rst variable,
it moves to the top of the list in the window. If you select multiple variables, then
subsequent selections move to the bottom of the list of selected variables. An asterisk
(*) identies each selected variable. The order of selected variables from top to bottom
determines their order in the report from left to right.

DATA SELECTION
Loads a data set into the current report denition.

Path
File

Open Data Set

Description
The rst list box in the DATA SELECTION window lists all the librefs dened for
your SAS session. The second one lists all the SAS data sets in the selected library.

918

DEFINITION

Chapter 42

Note: You must use data that is compatible with the current report denition. The
data set that you load must contain variables whose names are the same as the
variable names in the current report denition. 4

Buttons
OK

loads the selected data set into the current report denition.
Cancel

closes the DATA SELECTION window without loading new data.

DEFINITION
Displays the characteristics associated with an item in the report and lets you change them.

Path
Select a report item. Then select
Edit

Dene

Note: Alternatively, double-click on the selected item. (Not all operating

environments support this method of opening the DEFINITION window.) 4

Description

Usage
For an explanation of each type of usage see Laying Out a Report on page 852.
DISPLAY

denes the selected item as a display variable. DISPLAY is the default for character
variables.
ORDER

denes the selected item as an order variable.

The REPORT Procedure

DEFINITION

919

GROUP

denes the selected item as a group variable.

ACROSS

denes the selected item as an across variable.

ANALYSIS

denes the selected item as an analysis variable. You must specify a statistic (see the
discussion of the Statistic= attribute on page 920) for an analysis variable.
ANALYSIS is the default for numeric variables.
COMPUTED

denes the selected item as a computed variable. Computed variables are variables
that you dene for the report. They are not in the input data set, and PROC
REPORT does not add them to the input data set. However, computed variables are
included in an output data set if you create one.
In the windowing environment, you add a computed variable to a report from the
COMPUTED VAR window.

Attributes
Format=

assigns a SAS or user-dened format to the item. This format applies to the selected
item as PROC REPORT displays it; the format does not alter the format that is
associated with a variable in the data set. For data set variables, PROC REPORT
honors the rst of these formats that it nds:
3 the format that is assigned with FORMAT= in the DEFINITION window
3 the format that is assigned in a FORMAT statement when you start PROC
REPORT
3 the format that is associated with the variable in the data set.
If none of these is present, then PROC REPORT uses BESTw. for numeric
variables and $w. for character variables. The value of w is the default column
width. For character variables in the input data set, the default column width is the
variables length. For numeric variables in the input data set and for computed
variables (both numeric and character), the default column width is the value of the
COLWIDTH= attribute in the ROPTIONS window.
If you are unsure what format to use, then type a question mark (?) in the format
eld in the DEFINITION window to access the FORMATS window.
Spacing=

denes the number of blank characters to leave between the column being dened
and the column immediately to its left. For each column, the sum of its width and
the blank characters between it and the column to its left cannot exceed the line size.
Default: 2
Interaction: When PROC REPORTs CENTER option is in effect, PROC REPORT
ignores spacing that precedes the leftmost variable in the report.
Interaction: SPACING= in an item denition overrides the value of SPACING= in
the PROC REPORT statement or the ROPTIONS window.
Width=

denes the width of the column in which PROC REPORT displays the selected item.
Range: 1 to the value of the SAS system option LINESIZE=
Default: A column width that is just large enough to handle the format. If there is
no format, then PROC REPORT uses the value of COLWIDTH=.

920

DEFINITION

Chapter 42

Note: When you stack items in the same column in a report, the width of the
item that is at the bottom of the stack determines the width of the column. 4
Statistic=

Note: PROC REPORT uses the name of the analysis variable as the default
header for the column. You can customize the column header with the Header eld of
the DEFINITION window. 4
You can use the following values for statistic:
Descriptive statistic keywords
CSS

PCTSUM

RANGE

MAX

STDDEV|STD

MEAN

STDERR

MIN

SUM

SUMWGT

NMISS

USS

PCTN

VAR

Quantile statistic keywords

MEDIAN|Q2|P50

Q3|P75

P90

P95

P10

P99

Q1|P25

QRANGE

Hypothesis testing keyword

PROBT

Explanations of the keywords, the formulas that are used to calculate them, and
the data requirements are discussed in Appendix 1, SAS Elementary Statistics
Procedures, on page 1339.
To compute standard error and the Students t-test you must use the
default value of VARDEF= which is DF.

Requirement:

See also: For denitions of these statistics, see Keywords and Formulas on page

1340.
Order=

orders the values of a GROUP, ORDER, or ACROSS variable according to the

specied order, where

The REPORT Procedure

DEFINITION

921

DATA
orders values according to their order in the input data set.
FORMATTED
orders values by their formatted (external) values. By default, the order is
ascending.
FREQ
orders values by ascending frequency count.
INTERNAL
orders values by their unformatted values, which yields the same order that PROC
SORT would yield. This order is operating environment-dependent. This sort
sequence is particularly useful for displaying dates chronologically.
Default: FORMATTED
Interaction: DESCENDING in the items denition reverses the sort sequence for
an item.
Note: The default value for the ORDER= option in PROC REPORT is not the
same as the default value in other SAS procedures. In other SAS procedures, the
default is ORDER=INTERNAL. The default for the option in PROC REPORT may
change in a future release to be consistent with other procedures. Therefore, in
production jobs where it is important to order report items by their formatted values,
specify ORDER=FORMATTED even though it is currently the default. Doing so
ensures that PROC REPORT will continue to produce the reports you expect even if
the default changes. 4
Justify=

You can justify the placement of the column header and of the values of the item that
you are dening within a column in one of three ways:
LEFT
left-justies the formatted values of the item that you are dening within the
column width and left-justies the column header over the values. If the format
width is the same as the width of the column, then LEFT has no effect on the
placement of values.
RIGHT
right-justies the formatted values of the item that you are dening within the
column width and right-justies the column header over the values. If the format
width is the same as the width of the column, then RIGHT has no effect on the
placement of values.
CENTER
centers the formatted values of the item that you are dening within the column
width and centers the column header over the values. This option has no effect on
the setting of the SAS system option CENTER.
When justifying values, PROC REPORT justies the eld width dened by the
format of the item within the column. Thus, numbers are always aligned.
Data type=

shows you if the report item is numeric or character. You cannot change this eld.
Item Help=

references a HELP or CBT entry that contains help information for the selected item.
Use PROC BUILD in SAS/AF software to create a HELP or CBT entry for a report
item. All HELP and CBT entries for a report must be in the same catalog, and you
must specify that catalog with the HELP= option in the PROC REPORT statement
or from the User Help elds in the ROPTIONS window.

922

DEFINITION

Chapter 42

To access a help entry from the report, select the item and issue the HELP
command. PROC REPORT rst searches for and displays an entry named
entry-name.CBT. If no such entry exists, then PROC REPORT searches for
entry-name.HELP. If neither a CBT nor a HELP entry for the selected item exists,
then the opening frame of the help for PROC REPORT is displayed.
Alias=

By entering a name in the Alias eld, you create an alias for the report item that
you are dening. Aliases let you distinguish between different uses of the same
report item. When you refer in a compute block to a report item that has an alias,
you must use the alias (see Example 3 on page 954).

Options
NOPRINT

suppresses the display of the item that you are dening. Use this option
3 if you do not want to show the item in the report but you need to use the values
in it to calculate other values that you use in the report
3 to establish the order of rows in the report
3 if you do not want to use the item as a column but want to have access to its
values in summaries (see Example 9 on page 971).
Interaction: Even though the columns that you dene with NOPRINT do not

suppresses the display of the item that you are dening if its values are all zero or
missing.
Interaction: Even though the columns that you dene with NOZERO do not appear
in the report, you must count them when you are referencing columns by number
(see Four Ways to Reference Report Items in a Compute Block on page 859).
Interaction: SHOWALL in the PROC REPORT statement or the ROPTIONS
window overrides all occurrences of NOZERO.
DESCENDING

reverses the order in which PROC REPORT displays rows or values of a group, order,
or across variable.
PAGE

inserts a page break just before printing the rst column containing values of the
selected item.
Interaction: PAGE is ignored if you use WRAP in the PROC REPORT statement or
in the ROPTIONS window.
FLOW

wraps the value of a character variable in its column. The FLOW option honors the
split character. If the text contains no split character, then PROC REPORT tries to
split text at a blank.
ID column

species that the item that you are dening is an ID variable. An ID variable and all
columns to its left appear at the left of every page of a report. ID ensures that you

The REPORT Procedure

DISPLAY PAGE

923

can identify each row of the report when the report contains more columns than will
t on one page.

Color
From the list of colors, select the one to use in the REPORT window for the column
header and the values of the item that you are dening.

Default: The color of Foreground in the SASCOLOR window. (For more

information, see the online Help for the SASCOLOR window.)

Note:

Buttons
Apply

applies the information in the open window to the report and keeps the window open.
Edit Program

opens the COMPUTE window and enables you to associate a compute block with the
variable that you are dening.
OK

applies the information in the DEFINITION window to the report and closes the
window.
Cancel

closes the DEFINITION window without applying changes made with APPLY .

DISPLAY PAGE
Displays a particular page of the report.

Path
View

Display Page

Description
You can get to the last page of the report by entering a large number for the page
number. When you are on the last page of the report, PROC REPORT sends a note to
the message line of the REPORT window.

924

EXPLORE

Chapter 42

EXPLORE
Lets you experiment with your data.
Restriction: You cannot open the EXPLORE window unless your report contains at least
one group or order variable.

Path
Edit

Explore Data

Description
In the EXPLORE window you can

3 subset the data with list boxes

3 suppress the display of a column with the Remove Column check box
3 change the order of the columns with Rotate columns .
Note: The results of your manipulations in the EXPLORE window appear in the
REPORT window but are not saved in report denitions. 4

Window Features
list boxes

The EXPLORE window contains three list boxes. These boxes contain the value All
levels as well as actual values for the rst three group or order variables in your
report. The values reect any WHERE clause processing that is in effect. For
example, if you use a WHERE clause to subset the data so that it includes only the
northeast and northwest sectors, then the only values that appear in the list box for
Sector are All levels, Northeast, and Northwest. Selecting All levels in this
case displays rows of the report for only the northeast and northwest sectors. To see
data for all the sectors, you must clear the WHERE clause before you open the
EXPLORE window.
Selecting values in the list boxes restricts the display in the REPORT window to
the values that you select. If you select incompatible values, then PROC REPORT
returns an error.
Remove Column

Above each list box in the EXPLORE window is a check box labeled Remove Column.
Selecting this check box and applying the change removes the column from the
REPORT window. You can easily restore the column by clearing the check box and
applying that change.

Buttons
OK

applies the information in the EXPLORE window to the report and closes the window.

The REPORT Procedure

LOAD REPORT

925

Apply

applies the information in the EXPLORE window to the report and keeps the window
open.
Rotate columns

changes the order of the variables displayed in the list boxes. Each variable that can
move one column to the left does; the leftmost variable moves to the third column.
Cancel

closes the EXPLORE window without applying changes made with APPLY .

FORMATS
Displays a list of formats and provides a sample of each one.

Path
From the DEFINE window, type a question mark (?) in the Format eld and select
any of the Buttons except Cancel, or press RETURN.

Description
When you select a format in the FORMATS window, a sample of that format appears
in the Sample: eld. Select the format that you want to use for the variable that you
are dening.

Buttons
OK

writes the format that you have selected into the Format eld in the DEFINITION
window and closes the FORMATS window. To see the format in the report, select
Apply in the DEFINITION window.
Cancel

closes the FORMATS window without writing a format into the Format eld.

LOAD REPORT
Loads a stored report denition.

Path
File

Open Report

Description
The rst list box in the LOAD REPORT window lists all the librefs that are dened
for your SAS session. The second list box lists all the catalogs that are in the selected

926

MESSAGES

Chapter 42

library. The third list box lists descriptions of all the stored report denitions (entry
types of REPT) that are in the selected catalog. If there is no description for an entry,
then the list box contains the entrys name.

Buttons
OK

loads the current data into the selected report denition.

Cancel

closes the LOAD REPORT window without loading a new report denition.
Note: Issuing the END command in the REPORT window returns you to the
previous report denition (with the current data). 4

MESSAGES
Automatically opens to display notes, warnings, and errors returned by PROC REPORT.
You must close the MESSAGES window by selecting OK before you can continue to
use PROC REPORT.

PROFILE
Customizes some features of the PROC REPORT environment by creating a report prole.

Path
Tools

Report Prole

Description
The PROFILE window creates a report prole that

3 species the SAS library, catalog, and entry that dene alternative menus to use
in the REPORT and COMPUTE windows. Use PROC PMENU to create catalog
entries of type PMENU that dene these menus. PMENU entries for both
windows must be in the same catalog.

3 sets defaults for WINDOWS, PROMPT, and COMMAND. PROC REPORT uses the
default option whenever you start the procedure unless you specically override
the option in the PROC REPORT statement.
Specify the catalog that contains the prole to use with the PROFILE= option in the
PROC REPORT statement (see the discussion of PROFILE= on page 880).

The REPORT Procedure

PROMPTER

927

Buttons
OK

stores your prole in a le that is called SASUSER.PROFILE.REPORT.PROFILE.

Note: Use PROC CATALOG or the EXPLORER window to copy the prole to
another location. 4
Cancel

closes the window without storing the prole.

PROMPTER
Prompts you for information as you add items to a report.

Path
Specify the PROMPT option when you start PROC REPORT or select PROMPT from
the ROPTIONS window. The PROMPTER window opens the next time that you add an
item to the report.

Description
The prompter guides you through parts of the windows that are most commonly used
to build a report. As the content of the PROMPTER window changes, the title of the
window changes to the name of the window that you would use to perform a task if you
were not using the prompter. The title change is to help you begin to associate the
windows with their functions and to learn what window to use if you later decide to
change something.
If you start PROC REPORT with prompting, then the rst window gives you a
chance to limit the number of observations that are used during prompting. When you
exit the prompter, PROC REPORT removes the limit.

Buttons
OK

applies the information in the open window to the report and continues the
prompting process.
Note: When you select OK from the last prompt window, PROC REPORT
removes any limit on the number of observations that it is working with. 4
Apply

applies the information in the open window to the report and keeps the window open.
Backup

returns you to the previous PROMPTER window.

Exit Prompter

closes the PROMPTER window without applying any more changes to the report. If
you have limited the number of observations to use during prompting, then PROC
REPORT removes the limit.

928

REPORT

Chapter 42

REPORT
Is the surface on which the report appears.

Path
Use WINDOWS or PROMPT in the PROC REPORT statement.

Description
You cannot write directly in any part of the REPORT window except column headers.
To change other aspects of the report, you select a report item (for example, a column
heading) as the target of the next command and issue the command. To select an item,
use a mouse or cursor keys to position the cursor over it. Then click the mouse button
or press ENTER. To execute a command, make a selection from the menu bar at the top
of the REPORT window. PROC REPORT displays the effect of a command immediately
unless the DEFER option is on.
Note: Issuing the END command in the REPORT window returns you to the
previous report denition with the current data. If there is no previous report
denition, then END closes the REPORT window. 4

ROPTIONS
Displays choices that control the layout and display of the entire report and identies the SAS data
library and catalog containing CBT or HELP entries for items in the report.

Path
Tools

Options

Report

The REPORT Procedure

ROPTIONS

929

Description

Modes
DEFER

stores the information for changes and makes the changes all at once when you turn
DEFER mode off or select
View

Refresh

DEFER is particularly useful when you know that you need to make several
changes to the report but do not want to see the intermediate reports.
By default, PROC REPORT redisplays the report in the REPORT window each
time you redene the report by adding or deleting an item, by changing information
in the DEFINITION window, or by changing information in the BREAK window.
PROMPT

opens the PROMPTER window the next time that you add an item to the report.

Options
CENTER

centers the report and summary text (customized break lines). If CENTER is not
selected, then the report is left-justied.
PROC REPORT honors the rst of these centering specications that it nds:

3 the CENTER or NOCENTER option in the PROC REPORT statement or the

CENTER toggle in the ROPTIONS window

3 the CENTER or NOCENTER option stored in the report denition loaded with
REPORT= in the PROC REPORT statement

3 the SAS system option CENTER or NOCENTER.

When PROC REPORTs CENTER option is in effect, PROC REPORT ignores
spacing that precedes the leftmost variable in the report.
HEADLINE

underlines all column headers and the spaces between them at the top of each page
of the report.

930

ROPTIONS

Chapter 42

HEADLINE underlines with the second formatting character. (See the discussion
of FORMCHAR= on page 874.)
Default: hyphen (-)

In traditional (monospace) SAS output, you can underline column headers

without underlining the spaces between them, by using -- as the last line of
each column header instead of using HEADLINE.

Tip:

HEADSKIP

writes a blank line beneath all column headers (or beneath the underlining that the
HEADLINE option writes) at the top of each page of the report.
NAMED

writes name= in front of each value in the report, where name is the column header
for the value.
Use NAMED in conjunction with WRAP to produce a report that wraps all
columns for a single row of the report onto consecutive lines rather than placing
columns of a wide report on separate pages.

Tip:

Interaction: When you use NAMED, PROC REPORT automatically uses

NOHEADER.
NOHEADER

suppresses column headers, including those that span multiple columns.

Once you suppress the display of column headers in the windowing environment,
you cannot select any report items.
SHOWALL

overrides the parts of a denition that suppress the display of a column (NOPRINT
and NOZERO). You dene a report item with a DEFINE statement or in the
DEFINITION window.
WRAP

denitions.
Typically, you use WRAP in conjunction with NAMED to avoid wrapping
column headers.

Tip:
BOX

uses formatting characters to add line-drawing characters to the report. These

characters

3 surround each page of the report

3 separate column headers from the body of the report
3 separate rows and columns from each other.
Interaction: You cannot use BOX if you use WRAP in the PROC REPORT

statement or ROPTIONS window or if you use FLOW in any items denition.

See also: For information about formatting characters, see the discussion of

FORMCHAR= on page 874.

MISSING

considers missing values as valid values for group, order, or across variables. Special
missing values that are used to represent numeric values (the letters A through Z

The REPORT Procedure

ROPTIONS

931

and the underscore (_) character) are each considered as a different value. A group
for each missing value appears in the report. If you omit the MISSING option, then
PROC REPORT does not include observations with a missing value for one or more
group, order, or across variables in the report.

Attributes
Linesize

species the line size for a report. PROC REPORT honors the rst of these line-size
specications that it nds:

3 LS= in the PROC REPORT statement or Linesize= in the ROPTIONS window

3 the LS= setting stored in the report denition loaded with REPORT= in the
PROC REPORT statement

3 the SAS system option LINESIZE=.

Range: 64-256 (integer)

If the line size is greater than the width of the REPORT window, then use SAS
windowing environment commands RIGHT and LEFT to display portions of the
report that are not currently in the display.

Tip:

Pagesize

species the page size for a report. PROC REPORT honors the rst of these page
size specications that it nds:

3 PS= in the PROC REPORT statement or Pagesize= in the ROPTIONS window

3 the PS= setting stored in the report denition loaded with REPORT= in the
PROC REPORT statement

3 the SAS system option PAGESIZE=.

Range: 15-32,767 (integer)
Colwidth

species the default number of characters for columns containing computed variables
or numeric data set variables.
Range: 1 to the linesize
Default: 9
Interaction: When setting the width for a column, PROC REPORT rst looks at

WIDTH= in the denition for that column. If WIDTH= is not present, then PROC
REPORT uses a column width large enough to accommodate the format for the
item. (For information about formats, see the discussion of Format= on page 919.)
If no format is associated with the item, then the column width depends on
variable type:

If the variable is a

Then the column width is the

character variable in the input data set

length of the variable

numeric variable in the input data set

value of the COLWIDTH= option

computed variable (numeric or character)

value of the COLWIDTH= option

932

ROPTIONS

Chapter 42

SPACING=space-between-columns

blank characters specied by SPACING= in the PROC REPORT statement or the

ROPTIONS window unless you use SPACING= in the denition of a particular
item to change the spacing to the left of that item.
Interaction: When CENTER is in effect, PROC REPORT ignores spacing that

precedes the leftmost variable in the report.

SPLIT=character

If you are typing over a header (rather than entering one from the
PROMPTER or DEFINITION window), then you do not see the effect of the split
character until you refresh the screen by adding or deleting an item, by changing
the contents of a DEFINITION or a BREAK window, or by selecting

Note:

View

Refresh

PANELS=number-of-panels

3 width of the panel

3 space between panels
3 line size.
Default: 1

If number-of-panels is larger than the number of panels that can t on the

page, then PROC REPORT creates as many panels as it can. Let PROC REPORT
put your data in the maximum number of panels that can t on the page by
specifying a large number of panels (for example, 99).

Tip:

See also: For information about specifying the space between panels see the

discussion of PSPACE= on page 932. For information about setting the linesize,
see the discussion of Linesize on page 931).
PSPACE=space-between-panels

species the number of blank characters between panels. PROC REPORT separates
all panels in the report by the same number of blank characters. For each panel, the

The REPORT Procedure

SAVE DEFINITION

933

sum of its width and the number of blank characters separating it from the panel to
its left cannot exceed the line size.
Default: 4
User Help

identies the library and catalog containing user-dened help for the report. This
help can be in CBT or HELP catalog entries. You can write a CBT or HELP entry for
each item in the report with the BUILD procedure in SAS/AF software. You must
store all such entries for a report in the same catalog.
Specify the entry name for help for a particular report item in the DEFINITION
window for that report item or in a DEFINE statement.

SAVE DATA SET

Lets you specify an output data set in which to store the data from the current report.

Path
File

Save Data Set

Description
To specify an output data set, enter the name of the SAS data library and the name
of the data set (called member in the window) that you want to create in the Save Data
Set window.

Buttons
OK

Creates the output data set and closes the Save Data Set window.
Cancel

Closes the Save Data Set window without creating an output data set.

SAVE DEFINITION
Saves a report denition for subsequent use with the same data set or with a similar data set.

Path
File

Save Report

Description
The SAVE DEFINITION window prompts you for the complete name of the catalog
entry in which to store the denition of the current report and for an optional

934

SOURCE

Chapter 42

description of the report. This description shows up in the LOAD REPORT window and
helps you to select the appropriate report.
SAS stores the report denition as a catalog entry of type REPT. You can use a report
denition to create an identically structured report for any SAS data set that contains
variables with the same names as those used in the report denition.

Buttons
OK

Creates the report denition and closes the SAVE DEFINITION window.
Cancel

Closes the SAVE DEFINITION window without creating a report denition.

SOURCE
Lists the PROC REPORT statements that build the current report.

Path
Tools

Report Statements

STATISTICS
Displays statistics that are available in PROC REPORT.

Path
Edit

Add item

Statistic

After you select Statistic, PROC REPORT prompts you for the location of the
statistic relative to the column that you have selected. After you select a location, the
STATISTICS window opens.

Description
Select the statistics that you want to include in your report and close the window.
When you select the rst statistic, it moves to the top of the list in the window. If you
select multiple statistics, then subsequent selections move to the bottom of the list of
selected statistics. An asterisk (*) indicates each selected statistic. The order of selected
statistics from top to bottom determines their order in the report from left to right.
Note: If you double-click on a statistic, then PROC REPORT immediately adds it to
the report. The STATISTICS window remains open. 4

The REPORT Procedure

WHERE ALSO

935

To compute standard error and the Students t test you must use the default value of
VARDEF= which is DF.
To add all selected statistics to the report, select
File

Accept Selection

Selecting
File

closes the STATISTICS window without adding the selected statistics to the report.

WHERE
Selects observations from the data set that meet the conditions that you specify.

Path
Subset

Where

Description
Enter a where-expression in the Enter where clause eld. A where-expression is an
arithmetic or logical expression that generally consists of a sequence of operands and
operators. For information about constructing a where-expression, see the
documentation of the WHERE statement in the section on statements in SAS Language
Reference: Dictionary.
Note: You can clear all where-expressions by leaving the Enter where clause eld
empty and by selecting OK . 4

Buttons
OK

Applies the where-expression to the report and closes the WHERE window.
Cancel

Closes the WHERE window without altering the report.

WHERE ALSO
Selects observations from the data set that meet the conditions that you specify and any other
conditions that are already in effect.

936

How PROC REPORT Builds a Report

Chapter 42

Path

Subset

Where Also

Description
Enter a where-expression in the Enter where also clause eld. A
where-expression is an arithmetic or logical expression that generally consists of a
sequence of operands and operators. For information about constructing a
where-expression, see the documentation of the WHERE statement in the chapter on
statements in SAS Language Reference: Dictionary.

Buttons
OK

Adds the where-expression to any other where-expressions that are already in effect
and applies them all to the report. It also closes the WHERE ALSO window.
Cancel

Closes the WHERE ALSO window without altering the report.

How PROC REPORT Builds a Report

Sequence of Events
This section explains the general process of building a report. For examples that
illustrate this process, see Report-Building Examples on page 937. The sequence of
events is the same whether you use programming statements or the windowing
environment.
To understand the process of building a report, you must understand the difference
between report variables and temporary variables. Report variables are variables that
are specied in the COLUMN statement. A report variable can come from the input
data set or can be computed (that is, the DEFINE statement for that variable species
the COMPUTED option). A report variable might or might not appear in a compute
block. Variables that appear only in one or more compute blocks are temporary
variables. Temporary variables do not appear in the report and are not written to the
output data set (if one is requested).
PROC REPORT constructs a report as follows:
1 It consolidates the data by group, order, and across variables. It calculates all
statistics for the report, those for detail rows as well as those for summary lines in
breaks. Statistics include those computed for analysis variables. PROC REPORT
calculates statistics for summary lines whether or not they appear in the report. It
stores all this information in a temporary le.
2 It initializes all temporary variables to missing.
3 It begins constructing the rows of the report.
a At the beginning of each row, it initializes all report variables to missing.

The REPORT Procedure

Report-Building Examples

937

b It lls in values for report variables from left to right.

3 Values for computed variables come from executing the statements in

the corresponding compute blocks.

3 Values for all other variables come from the temporary le that was
created at the beginning of the report-building process.
c Whenever it comes to a break, PROC REPORT rst constructs the break

lines that are created with the BREAK or RBREAK statement or with
options in the BREAK window. If there is a compute block attached to the
break, then PROC REPORT then executes the statements in the compute
block. See Construction of Summary Lines on page 937 for details.
Note: Because of the way PROC REPORT builds a report, you can
3 use group statistics in compute blocks for a break before the group variable.
3 use statistics for the whole report in a compute block at the beginning of the
report.
This document references these statistics with the appropriate compound name.
For information about referencing report items in a compute block, see Four Ways
to Reference Report Items in a Compute Block on page 859. 4

Construction of Summary Lines

PROC REPORT constructs a summary line for a break if either of the following
conditions is true:
3 You summarize numeric variables in the break.
3 You use a compute block at the break. (You can attach a compute block to a break
without using a BREAK or RBREAK statement or without selecting any options in
the BREAK window.)
For more information about using compute blocks, see Using Compute Blocks
on page 858 and COMPUTE Statement on page 895.
The summary line that PROC REPORT constructs at this point is preliminary. If no
compute block is attached to the break, then the preliminary summary line becomes the
nal summary line. However, if a compute block is attached to the break, then the
statements in the compute block can alter the values in the preliminary summary line.
PROC REPORT prints the summary line only if you summarize numeric variables in
the break.

Report-Building Examples
Building a Report That Uses Groups and a Report Summary
The report in Output 42.2 contains ve columns:
3 Sector and Department are group variables.
3 Sales is an analysis variable that is used to calculate the Sum statistic.
3 Prot is a computed variable whose value is based on the value of Department.
3 The N statistic indicates how many observations each row represents.
At the end of the report a break summarizes the statistics and computed variables in
the report and assigns to Sector the value of TOTALS:.
The following statements produce Output 42.2. The user-dened formats that are
used are created by a PROC FORMAT step on page 949.

938

Report-Building Examples

Chapter 42

libname proclib SAS-data-library;

options nodate pageno=1 linesize=64

pagesize=60 fmtsearch=(proclib);
proc report data=grocery headline headskip;
column sector department sales Profit N;
define sector / group format=$sctrfmt.;
define department
/ group format=$deptfmt.;
define sales / analysis sum
format=dollar9.2;
define profit / computed format=dollar9.2;
compute profit;
if department=np1 or department=np2
then profit=0.4*sales.sum;
else profit=0.25*sales.sum;
endcomp;
rbreak after / dol dul summarize;
compute after;
sector=TOTALS:;
endcomp;
where sector contains n;
title Report for Northeast and Northwest Sectors;
run;

Output 42.2

Report with Groups and a Report Summary

Report for Northeast and Northwest Sectors

Sector
Department
Sales
Profit
N
-----------------------------------------------------Northeast

Northwest

=========
TOTALS:
=========

Canned
Meat/Dairy
Paper
Produce
Canned
Meat/Dairy
Paper
Produce

$840.00
$490.00
$290.00
$211.00
$1,070.00
$1,055.00
$150.00
$179.00
=========
$4,285.00
=========

$336.00
$122.50
$116.00
$52.75
$428.00
$263.75
$60.00
$44.75
=========
$1,071.25
=========

2
2
2
2
3
3
3
3
=========
20
=========

A description of how PROC REPORT builds this report follows:

1 PROC REPORT starts building the report by consolidating the data (Sector and

Department are group variables) and by calculating the statistics (Sales.sum and
N) for each detail row and for the break at the end of the report. It stores these
values in a temporary le.
2 Now, PROC REPORT is ready to start building the rst row of the report. This

report does not contain a break at the beginning of the report or a break before

The REPORT Procedure

Report-Building Examples

939

any groups, so the rst row of the report is a detail row. The procedure initializes
all report variables to missing, as Figure 42.9 on page 939 illustrates. Missing
values for a character variable are represented by a blank, and missing values for
a numeric variable are represented by a period.

Figure 42.9

First Detail Row with Values Initialized

Sector

Department

Sales

Profit

3 Figure 42.10 on page 939 illustrates the construction of the rst three columns of

the row. PROC REPORT lls in values for the row from left to right. Values come
from the temporary le that is created at the beginning of the report-building
process.

Figure 42.10 First Detail Row with Values Filled in from Left to Right

Sector

Department

Profit

Northeast

Sales

Sector

Department

Sales

Profit

Northeast

Canned

Sector

Department

Sales

Profit

Northeast

Canned

$840.00

4 The next column in the report contains the computed variable Prot. When it gets

to this column, PROC REPORT executes the statements in the compute block that
is attached to Prot. Nonperishable items (which have a value of np1 or np2)
return a prot of 40%; perishable items (which have a value of p1 or p2) return a
prot of 25%.
if department=np1 or department=np2
then profit=0.4*sales.sum;
else profit=0.25*sales.sum;

The row now looks like Figure 42.11 on page 940.

Note: The position of a computed variable is important. PROC REPORT
assigns values to the columns in a row of a report from left to right. Consequently,

940

Report-Building Examples

Chapter 42

you cannot base the calculation of a computed variable on any variable that
appears to its right in the report. 4
Figure 42.11 A Computed Variable Added to the First Detail Row

Sector

Department

Sales

sector=TOTALS:;

This statement replaces the value of Sector, which in the summary line is
missing by default, with the word TOTALS:. After PROC REPORT executes the
statement, it modies the summary line to reect this change to the value of
Sector. The nal version of the summary line appears in Figure 42.14 on page 941.

Figure 42.14 Final Summary Line

Sector
TOTALS:

Department

Sales

Profit

$4,285.00

$1,071.25

10 Finally, PROC REPORT writes all the break lines, with underlining, overlining,

and the nal summary line, to the report.

Building a Report That Uses Temporary Variables

PROC REPORT initializes report variables to missing at the beginning of each row of
the report. The value for a temporary variable is initialized to missing before PROC
REPORT begins to construct the rows of the report, and it remains missing until you
specically assign a value to it. PROC REPORT retains the value of a temporary
variable from the execution of one compute block to another.
Because all compute blocks share the current values of all variables, you can
initialize temporary variables at a break at the beginning of the report or at a break
before a break variable. This report initializes the temporary variable Sctrtot at a
break before Sector.
Note: PROC REPORT creates a preliminary summary line for a break before it
executes the corresponding compute block. If the summary line contains computed
variables, then the computations are based on the values of the contributing variables
in the preliminary summary line. If you want to recalculate computed variables based
on values that you set in the compute block, then you must do so explicitly in the
compute block. This report illustrates this technique.
If no compute block is attached to a break, then the preliminary summary line
becomes the nal summary line. 4
The report in Output 42.3 contains ve columns:

3 Sector and Department are group variables.

3 Sales is an analysis variable that is used twice in this report: once to calculate the
Sum statistic, and once to calculate the Pctsum statistic.

3 Sctrpct is a computed variable whose values are based on the values of Sales and a
temporary variable, Sctrtot, which is the total sales for a sector.
At the beginning of the report, a customized report summary tells what the sales for
all stores are. At a break before each group of observations for a department, a default
summary summarizes the data for that sector. At the end of each group a break inserts
a blank line.
The following statements produce Output 42.3. The user-dened formats that are
used are created by a PROC FORMAT step on page 949.

942

Report-Building Examples

Chapter 42

Note: Calculations of the percentages do not multiply their results by 100 because
PROC REPORT prints them with the PERCENT. format. 4

libname proclib SAS-data-library;

options nodate pageno=1 linesize=64

pagesize=60 fmtsearch=(proclib);
proc report data=grocery noheader nowindows;
column sector department sales
Sctrpct sales=Salespct;
define sector

/ Sector group
format=$sctrfmt.;
define department / group format=$deptfmt.;
define sales
/ analysis sum
format=dollar9.2 ;
define sctrpct
/ computed
format=percent9.2 ;
define salespct
/ pctsum format=percent9.2;
compute before;
line ;
line @16 Total for all stores is
sales.sum dollar9.2;
line ;
line @29 Sum of @40 Percent
@51 Percent of;
line @6 Sector @17 Department
@29 Sales
@40 of Sector @51 All Stores;
line @6 55*=;
line ;
endcomp;
break before sector / summarize ul;
compute before sector;
sctrtot=sales.sum;
sctrpct=sales.sum/sctrtot;
endcomp;
compute sctrpct;
sctrpct=sales.sum/sctrtot;
endcomp;
break after sector/skip;
where sector contains n;
title Report for Northeast and Northwest Sectors;
run;

The REPORT Procedure

Output 42.3

Report-Building Examples

943

Report with Temporary Variables

Report for Northeast and Northwest Sectors

Total for all stores is $4,285.00

Sum of
Percent
Percent of
Sector
Department Sales
of Sector All Stores
=======================================================
Northeast
--------Northeast

Northwest
--------Northwest

Canned
Meat/Dairy
Paper
Produce

$1,831.00
--------$840.00
$490.00
$290.00
$211.00

100.00%
--------45.88%
26.76%
15.84%
11.52%

42.73%
--------19.60%
11.44%
6.77%
4.92%

Canned
Meat/Dairy
Paper
Produce

$2,454.00
--------$1,070.00
$1,055.00
$150.00
$179.00

100.00%
--------43.60%
42.99%
6.11%
7.29%

57.27%
--------24.97%
24.62%
3.50%
4.18%

A description of how PROC REPORT builds this report follows:

1 PROC REPORT starts building the report by consolidating the data (Sector and

Department are group variables) and by calculating the statistics (Sales.sum and
Sales.pctsum) for each detail row, for the break at the beginning of the report, for
the breaks before each group, and for the breaks after each group. It stores these
values in a temporary le.
2 PROC REPORT initializes the temporary variable, Sctrtot, to missing (see Figure

42.15 on page 943).

Figure 42.15 Initialized Temporary Variables

Report Variables
Sector

Department

Temporary
Variable

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

3 Because this PROC REPORT step contains a COMPUTE BEFORE statement, the

procedure constructs a preliminary summary line for the break at the beginning of
the report. This preliminary summary line contains values for the statistics
(Sales.sum and Sales.pctsum) and the computed variable (Sctrpct).
At this break, Sales.sum is the sales for all stores, and Sales.pctsum is the
percentage those sales represent for all stores (100%). PROC REPORT takes the
values for these statistics from the temporary le that it created at the beginning
of the report-building process.
The value for Sctrpct comes from executing the statements in the corresponding
compute block. Because the value of Sctrtot is missing, PROC REPORT cannot
calculate a value for Sctrpct. Therefore, in the preliminary summary line (which is

944

Report-Building Examples

Chapter 42

not printed in this case), this variable also has a missing value (see Figure 42.16
on page 944).
The statements in the COMPUTE BEFORE block do not alter any variables.
Therefore, the nal summary line is the same as the preliminary summary line.
Note: The COMPUTE BEFORE statement creates a break at the beginning of
the report. You do not need to use an RBREAK statement.

Figure 42.16 Preliminary and Final Summary Line for the Break at the Beginning
of the Report
Report Variables
Sector

Department

Temporary
Variable

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

$4,285.00

100.00%

4 Because the program does not include an RBREAK statement with the

SUMMARIZE option, PROC REPORT does not write the nal summary line to the
report. Instead, it uses LINE statements to write a customized summary that
embeds the value of Sales.sum into a sentence and to write customized column
headers. (The NOHEADER option in the PROC REPORT statement suppresses
the default column headers, which would have appeared before the customized
summary.)
5 Next, PROC REPORT constructs a preliminary summary line for the break before

the rst group of observations. (This break both uses the SUMMARIZE option in
the BREAK statement and has a compute block attached to it. Either of these
conditions generates a summary line.) The preliminary summary line contains
values for the break variable (Sector), the statistics (Sales.sum and Sales.pctsum),
and the computed variable (Sctrpct). At this break, Sales.sum is the sales for one
sector (the northeast sector). PROC REPORT takes the values for Sector,
Sales.sum, and Sales.pctsum from the temporary le that it created at the
beginning of the report-building process.
The value for Sctrpct comes from executing the statements in the corresponding
compute blocks. Because the value of Sctrtot is still missing, PROC REPORT
cannot calculate a value for Sctrpct. Therefore, in the preliminary summary line,
Sctrpct has a missing value (see Figure 42.17 on page 944).

Figure 42.17 Preliminary Summary Line for the Break before the First Group of
Observations
Report Variables
Sector
Northeast

Department

Temporary
Variable

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

$1,831.00

42.73%

6 PROC REPORT creates the nal version of the summary line by executing the

statements in the COMPUTE BEFORE SECTOR compute block. These

statements execute once each time the value of Sector changes.

3 The rst statement assigns the value of Sales.sum, which in that part of the
report represents total sales for one Sector, to the variable Sctrtot.

The REPORT Procedure

Report-Building Examples

945

3 The second statement completes the summary line by recalculating Sctrpct

from the new value of Sctrtot. Figure 42.18 on page 945 shows the nal
summary line.
Note: In this example, you must recalculate the value for Sctrpct in the nal
summary line. If you do not recalculate the value for Sctrpct, then it will be
missing because the value of Sctrtot is missing at the time that the COMPUTE
Sctrpct block executes. 4
Figure 42.18 Final Summary Line for the Break before the First Group of
Observations
Report Variables
Sector

Department

Temporary
Variable
Sctrpct

Sales.pctsum

Sctrtot

$1,831.00

Northeast

Sales.sum

100.00%

42.73%

$1,831.00

7 Because the program contains a BREAK BEFORE statement with the

SUMMARIZE option, PROC REPORT writes the nal summary line to the report.
The UL option in the BREAK statement underlines the summary line.
8 Now, PROC REPORT is ready to start building the rst detail row of the report. It
initializes all report variables to missing. Values for temporary variables do not
change. Figure 42.19 on page 945 illustrates the rst detail row at this point.

Figure 42.19 First Detail Row with Initialized Values

Report Variables
Sector

Department

Temporary
Variable

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

$1,831.00

9 Figure 42.20 on page 945 illustrates the construction of the rst three columns of

the row. PROC REPORT lls in values for the row from left to right. The values
come from the temporary le that it created at the beginning of the report-building
process.

Figure 42.20 Filling in Values from Left to Right

Report Variables
Sector

Department

Temporary
Variable
Sctrpct

Sales.pctsum

Sctrtot

Northeast

Sales.sum

$1,831.00

Report Variables

Temporary
Variable

Sector

Department

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

Northeast

Canned

$1,831.00

Report Variables

Temporary
Variable

Sector

Department

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

Northeast

Canned

$840.00

$1,831.00

946

Report-Building Examples

Chapter 42

10 The next column in the report contains the computed variable Sctrpct. When it

gets to this column, PROC REPORT executes the statement in the compute block
attached to Sctrpct. This statement calculates the percentage of the sectors total
sales that this department accounts for:
sctrpct=sales.sum/sctrtot;

The row now looks like Figure 42.21 on page 946.

Figure 42.21 First Detail Row with the First Computed Variable Added
Report Variables

Temporary
Variable

Sector

Department

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

Northeast

Canned

$840.00

45.88%

$1,831.00

11 The next column in the report contains the statistic Sales.pctsum. PROC REPORT

gets this value from the temporary le. The rst detail row is now complete (see
Figure 42.22 on page 946).

Figure 42.22 First Complete Detail Row

Report Variables

Temporary
Variable

Sector

Department

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

Northeast

Canned

$840.00

45.88%

19.60%

$1,831.00

12 PROC REPORT writes the detail row to the report. It repeats steps 8, 9, 10, 11,

and 12 for each detail row in the group.

13 After writing the last detail row in the group to the report, PROC REPORT
constructs the default group summary. Because no compute block is attached to
this break and because the BREAK AFTER statement does not include the
SUMMARIZE option, PROC REPORT does not construct a summary line. The
only action at this break is that the SKIP option in the BREAK AFTER statement
writes a blank line after the last detail row of the group.
14 Now the value of the break variable changes from Northeast to Northwest.
PROC REPORT constructs a preliminary summary line for the break before this
group of observations. As at the beginning of any row, PROC REPORT initializes
all report variables to missing but retains the value of the temporary variable.
Next, it completes the preliminary summary line with the appropriate values for
the break variable (Sector), the statistics (Sales.sum and Sales.pctsum), and the
computed variable (Sctrpct). At this break, Sales.sum is the sales for the
Northwest sector. Because the COMPUTE BEFORE Sector block has not yet
executed, the value of Sctrtot is still $1,831.00, the value for the Northeast sector.
Thus, the value that PROC REPORT calculates for Sctrpct in this preliminary
summary line is incorrect (see Figure 42.23 on page 947). The statements in the
compute block for this break calculate the correct value (see the following step).

The REPORT Procedure

Report-Building Examples

947

Figure 42.23 Preliminary Summary Line for the Break before the Second Group of
Observations
Report Variables
Sector

Department

Temporary
Variable

Sales.sum

Sales.pctsum

Sctrtot

$2,454.00

Northwest

Sctrpct
134.00%

57.27%

$1,831.00

CAUTION:

Synchronize values for computed variables in break lines to prevent incorrect results.
If the PROC REPORT step does not recalculate Sctrpct in the compute block
that is attached to the break, then the value in the nal summary line will not
be synchronized with the other values in the summary line, and the report will
be incorrect. 4
15 PROC REPORT creates the nal version of the summary line by executing the

statements in the COMPUTE BEFORE Sector compute block. These statements

execute once each time the value of Sector changes.

3 The rst statement assigns the value of Sales.sum, which in that part of the
report represents sales for the Northwest sector, to the variable Sctrtot.

3 The second statement completes the summary line by recalculating Sctrpct

from the new, appropriate value of Sctrtot. Figure 42.24 on page 947 shows
the nal summary line.

Figure 42.24 Final Summary Line for the Break before the Second Group of
Observations
Report Variables
Sector
Northwest

Department

Temporary
Variable

Sales.sum

Sctrpct

Sales.pctsum

Sctrtot

$2,454.00

100.00%

57.27%

$2,454.00

Because the program contains a BREAK BEFORE statement with the

SUMMARIZE option, PROC REPORT writes the nal summary line to the report.
The UL option in the BREAK statement underlines the summary line.
16 Now, PROC REPORT is ready to start building the rst row for this group of

observations. It repeats steps 8 through 16 until it has processed all observations

in the input data set (stopping with step 14 for the last group of observations).

948

Examples: REPORT Procedure

Chapter 42

Examples: REPORT Procedure

Example 1: Selecting Variables for a Report

Procedure features:

PROC REPORT statement options:

NOWD
COLUMN statement
default variable usage
RBREAK statement options:
DOL
SUMMARIZE
Other features:

FORMAT statement
FORMAT procedure:
LIBRARY=
SAS system options:
FMTSEARCH=
Automatic macro variables:
SYSDATE

This example uses a permanent data set and permanent formats to create a report
that contains

3 one row for every observation

3 a default summary for the whole report.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

The REPORT Procedure

Program

949

Create the GROCERY data set. GROCERY contains one days sales gures for eight stores in
the Grocery Mart chain. Each observation contains one days sales data for one department in
one store.
data grocery;
input Sector $
datalines;
se 1 np1 50
se
se 2 np1 40
se
nw 3 np1 60
nw
nw 4 np1 45
nw
nw 9 np1 45
nw
sw 5 np1 53
sw
sw 6 np1 40
sw
ne 7 np1 90
ne
ne 8 np1 200
ne
;

Manager $ Department $ Sales @@;

1
2
3
4
9
5
6
7
8

p1
p1
p1
p1
p1
p1
p1
p1
p1

100
300
600
250
205
130
350
190
300

se
se
nw
nw
nw
sw
sw
ne
ne

1
2
3
4
9
5
6
7
8

np2
np2
np2
np2
np2
np2
np2
np2
np2

120
220
420
230
420
120
225
420
420

se
se
nw
nw
nw
sw
sw
ne
ne

1
2
3
4
9
5
6
7
8

p2
p2
p2
p2
p2
p2
p2
p2
p2

80
70
30
73
76
50
80
86
125

Create the $SCTRFMT., $MGRFMT., and $DEPTFMT. formats. PROC FORMAT creates
permanent formats for Sector, Manager, and Department. The LIBRARY= option species a
permanent storage location so that the formats are available in subsequent SAS sessions. These
formats are used for examples throughout this section.
proc format library=proclib;
value $sctrfmt se = Southeast
ne = Northeast
nw = Northwest
sw = Southwest;
value $mgrfmt 1
3
5
7
9

=
=
=
=
=

value $deptfmt np1

np2
p1
p2
run;

Smith
2
Reveiz 4
Taylor 6
Alomar 8
Pelfrey;
=
=
=
=

=
=
=
=

Jones
Brown
Adams
Andrews

Paper
Canned
Meat/Dairy
Produce;

Specify the format search library. The SAS system option FMTSEARCH= adds the SAS
data library PROCLIB to the search path that is used to locate formats.
options fmtsearch=(proclib);

Specify the report options. The NOWD option runs the REPORT procedure without the
REPORT window and sends its output to the open output destination(s).
proc report data=grocery nowd;

950

Output

Chapter 42

Specify the report columns. The report contains a column for Manager, Department, and
Sales. Because there is no DEFINE statement for any of these variables, PROC REPORT uses
the character variables (Manager and Department) as display variables and the numeric
variable (Sales) as an analysis variable that is used to calculate the sum statistic.
column manager department sales;

Produce a report summary. The RBREAK statement produces a default summary at the end
of the report. DOL writes a line of equal signs (=) above the summary information.
SUMMARIZE sums the value of Sales for all observations in the report.
rbreak after / dol summarize;

Select the observations to process. The WHERE statement selects for the report only the
observations for stores in the southeast sector.
where sector=se;

Format the report columns. The FORMAT statement assigns formats to use in the report.
You can use the FORMAT statement only with data set variables.
format manager $mgrfmt. department $deptfmt.
sales dollar11.2;

Specify the titles. SYSDATE is an automatic macro variable that returns the date when the
SAS job or SAS session began. The TITLE2 statement uses double rather than single quotation
marks so that the macro variable resolves.
title Sales for the Southeast Sector;
title2 "for &sysdate";
run;

Output
Sales for the Southeast Sector
for 04JAN02
Manager
Smith
Smith
Smith
Smith
Jones
Jones
Jones
Jones

Department
Paper
Meat/Dairy
Canned
Produce
Paper
Meat/Dairy
Canned
Produce

Sales
$50.00
$100.00
$120.00
$80.00
$40.00
$300.00
$220.00
$70.00
===========
$980.00

The REPORT Procedure

Program

Example 2: Ordering the Rows in a Report

Procedure features:

PROC REPORT statement options:

COLWIDTH=
HEADLINE
HEADSKIP
SPACING=
BREAK statement options:
OL
SKIP
SUMMARIZE
COMPUTE statement arguments:
AFTER
DEFINE statement options:
ANALYSIS
FORMAT=
ORDER
ORDER=
SUM
ENDCOMP statement
LINE statement:
with quoted text
with variable values
Data set:

GROCERY on page 949

Formats: $MGRFMT. and $DEPTFMT. on page 949

This example

3 arranges the rows alphabetically by the formatted values of Manager and the
internal values of Department (so that sales for the two departments that sell
nonperishable goods precede sales for the two departments that sell perishable
goods)

3
3
3
3

controls the default column width and the spacing between columns
underlines the column headers and writes a blank line beneath the underlining
creates a default summary of Sales for each manager
creates a customized summary of Sales for the whole report.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

951

952

Program

Chapter 42

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). COLWIDTH=10 sets the default
column width to 10 characters. SPACING= puts ve blank characters between columns.
HEADLINE underlines all column headers and the spaces between them at the top of each page
of the report. HEADSKIP writes a blank line beneath the underlining that HEADLINE writes.
proc report data=grocery nowd
colwidth=10
spacing=5
headline headskip;

Specify the report columns. The report contains a column for Manager, Department, and
Sales.
column manager department sales;

Dene the sort order variables. The values of all variables with the ORDER option in the
DEFINE statement determine the order of the rows in the report. In this report, PROC
REPORT arranges the rows rst by the value of Manager (because it is the rst variable in the
COLUMN statement) and then by the values of Department.
ORDER= species the sort order for a variable. This report arranges the rows according to the
formatted values of Manager and the internal values of Department (np1, np2, p1, and p2).
FORMAT= species the formats to use in the report.
define manager / order order=formatted format=$mgrfmt.;
define department / order order=internal format=$deptfmt.;

Dene the analysis variable. Sum calculates the sum statistic for all observations that are
represented by the current row. In this report each row represents only one observation.
Therefore, the Sum statistic is the same as the value of Sales for that observation in the input
data set. Using Sales as an analysis variable in this report enables you to summarize the values
for each group and at the end of the report.
define sales / analysis sum format=dollar7.2;

Produce a report summary. This BREAK statement produces a default summary after the
last row for each manager. OL writes a row of hyphens above the summary line. SUMMARIZE
writes the value of Sales (the only analysis or computed variable) in the summary line. PROC
REPORT sums the values of Sales for each manager because Sales is an analysis variable that
is used to calculate the Sum statistic. SKIP writes a blank line after the summary line.

The REPORT Procedure

Output

953

break after manager / ol

summarize
skip;

Produce a customized summary. This COMPUTE statement begins a compute block that
produces a customized summary at the end of the report. The LINE statement places the quoted
text and the value of Sales.sum (with the DOLLAR9.2 format) in the summary. An ENDCOMP
statement must end the compute block.
compute after;
line Total sales for these stores were:
sales.sum dollar9.2;
endcomp;

Select the observations to process. The WHERE statement selects for the report only the
observations for stores in the southeast sector.
where sector=se;

Specify the title.

title Sales for the Southeast Sector;
run;

Output
Sales for the Southeast Sector
Manager
Department
Sales
---------------------------------Jones

Paper
Canned
Meat/Dairy
Produce

$40.00
$220.00
$300.00
$70.00
------$630.00

Paper
Canned
Meat/Dairy
Produce

$50.00
$120.00
$100.00
$80.00
------$350.00

------Jones
Smith

------Smith

Total sales for these stores were:

$980.00

954

Example 3: Using Aliases to Obtain Multiple Statistics for the Same Variable

Chapter 42

Example 3: Using Aliases to Obtain Multiple Statistics for the Same Variable
Procedure features:

COLUMN statement:
with aliases
COMPUTE statement arguments:
AFTER
DEFINE statement options:
ANALYSIS
MAX
MIN
NOPRINT
customizing column headers
LINE statement:
pointer controls
quoted text
repeating a character string
variable values and formats
writing a blank line
Other features:

automatic macro variables:

SYSDATE
Data set:

GROCERY on page 949

Formats:

$MGRFMT. and $DEPTFMT. on page 949

The customized summary at the end of this report displays the minimum and
maximum values of Sales over all departments for stores in the southeast sector. To
determine these values, PROC REPORT needs the MIN and MAX statistic for Sales in
every row of the report. However, to keep the report simple, the display of these
statistics is suppressed.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

The REPORT Procedure

Program

955

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). HEADLINE underlines all
column headers and the spaces between them at the top of each page of the report. HEADSKIP
writes a blank line beneath the underlining that HEADLINE writes.
proc report data=grocery nowd headline headskip;

Specify the report columns. The report contains columns for Manager and Department. It
also contains three columns for Sales. The column specications SALES=SALESMIN and
SALES=SALESMAX create aliases for Sales. These aliases enable you to use a separate
denition of Sales for each of the three columns.
column manager department sales
sales=salesmin
sales=salesmax;

Dene the sort order variables. The values of all variables with the ORDER option in the
DEFINE statement determine the order of the rows in the report. In this report, PROC REPORT
arranges the rows rst by the value of Manager (because it is the rst variable in the COLUMN
statement) and then by the values of Department. The ORDER= option species the sort order
for a variable. This report arranges the values of Manager by their formatted values and
arranges the values of Department by their internal values (np1, np2, p1, and p2). FORMAT=
species the formats to use in the report. Text in quotation marks species column headings.
define manager / order
order=formatted
format=$mgrfmt.
Manager;
define department
/ order
order=internal
format=$deptfmt.
Department;

Dene the analysis variable. The value of an analysis variable in any row of a report is the
value of the statistic that is associated with it (in this case Sum), calculated for all observations
that are represented by that row. In a detail report each row represents only one observation.
Therefore, the Sum statistic is the same as the value of Sales for that observation in the input
data set.
define sales / analysis sum format=dollar7.2 Sales;

Dene additional analysis variables for use in the summary. These DEFINE statements
use aliases from the COLUMN statement to create separate columns for the MIN and MAX
statistics for the analysis variable Sales. NOPRINT suppresses the printing of these statistics.
Although PROC REPORT does not print these values in columns, it has access to them so that
it can print them in the summary.
define salesmin / analysis min noprint;
define salesmax / analysis max noprint;

956

Program

Chapter 42

Print a horizontal line at the end of the report. This COMPUTE statement begins a
compute block that executes at the end of the report. The rst LINE statement writes a blank
line. The second LINE statement writes 53 hyphens (-), beginning in column 7. Note that the
pointer control (@) has no effect on ODS destinations other than traditional SAS monospace
output.
compute after;
line ;
line @7 53*-;

Produce a customized summary. The rst line of this LINE statement writes the text in
quotation marks, beginning in column 7. The second line writes the value of Salesmin with the
DOLLAR7.2 format, beginning in the next column. The cursor then moves one column to the
right (+1), where PROC REPORT writes the text in quotation marks. Again, the cursor moves
one column to the right, and PROC REPORT writes the value of Salesmax with the DOLLAR7.2
format. (Note that the program must reference the variables by their aliases.) The third line
writes the text in quotation marks, beginning in the next column. Note that the pointer control
(@) is designed for the Listing destination (traditional SAS output). It has no effect on ODS
destinations other than traditional SAS monospace output. The ENDCOMP statement ends the
compute block.
line @7 | Departmental sales ranged from
salesmin dollar7.2 +1 to +1 salesmax dollar7.2
. |;
line @7 53*-;
endcomp;

Select the observations to process. The WHERE statement selects for the report only the
observations for stores in the southeast sector.
where sector=se;

The REPORT Procedure

Example 4: Consolidating Multiple Observations into One Row of a Report

957

Output
Sales for the Southeast Sector
for 04JAN02

Manager Department
Sales
---------------------------Jones

Smith

Paper
Canned
Meat/Dairy
Produce
Paper
Canned
Meat/Dairy
Produce

$40.00
$220.00
$300.00
$70.00
$50.00
$120.00
$100.00
$80.00

----------------------------------------------------| Departmental sales ranged from $40.00 to $300.00. |

-----------------------------------------------------

Example 4: Consolidating Multiple Observations into One Row of a Report

Specify the title.

title Sales Figures for Northern Sectors;
run;

960

Output

Chapter 42

Output
Sales Figures for Northern Sectors

Sector
Manager
Sales
-----------------------------Northeast

Alomar
Andrews

786.00
1,045.00
---------$1,831.00

Northwest

Brown
Pelfrey
Reveiz

598.00
746.00
1,110.00
---------$2,454.00

Combined sales for the northern sectors were $4,285.00.

Example 5: Creating a Column for Each Value of a Variable

Procedure features:

PROC REPORT statement options:

SPLIT=
BREAK statement options:
SKIP
COLUMN statement:
stacking variables
COMPUTE statement arguments:
with a computed variable as report-item
AFTER
DEFINE statement options:
ACROSS
ANALYSIS
COMPUTED
SUM
LINE statement:
pointer controls
Data set: GROCERY on page 949
Formats: $SCTRFMT., $MGRFMT., and $DEPTFMT. on page 949

The report in this example

3 consolidates multiple observations into one row
3 contains a column for each value of Department that is selected for the report (the
departments that sell perishable items)
3 contains a variable that is not in the input data set
3 uses customized column headers, some of which contain blank lines

The REPORT Procedure

Program

961

3 double-spaces between detail rows

3 uses pointer controls to control the placement of text and variable values in a
customized summary.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). HEADLINE underlines the
column headings. HEADSKIP writes a blank line beneath the underlining that HEADLINE
writes. SPLIT= denes the split character as an asterisk (*) because the default split character
(/) is part of the name of a department.
proc report data=grocery nowd
headline
headskip
split=*;

Specify the report columns. Department and Sales are separated by a comma in the
COLUMN statement, so they collectively determine the contents of the column that they dene.
Each item generates a header, but the header for Sales is set to blank in its denition. Because
Sales is an analysis variable, its values ll the cells that are created by these two variables.
column sector manager department,sales perish;

Dene the group variables. In this report, Sector and Manager are group variables. Each
detail row of the report consolidates the information for all observations with the same values of
the group variables. FORMAT= species the formats to use in the report. Text in quotation
marks in the DEFINE statements species column headings. These statements illustrate two
ways to write a blank line in a column header. Sector writes a blank line because each
quoted string is a line of the column heading. The two adjacent quotation marks write a blank
line for the second line of the heading. Manager* writes a blank line because the split
character (*) starts a new line of the heading. That line contains only a blank.
define sector / group format=$sctrfmt. Sector ;
define manager / group format=$mgrfmt. Manager* ;

962

Program

Chapter 42

Dene the across variable. PROC REPORT creates a column and a column heading for each
formatted value of the across variable Department. PROC REPORT orders the columns by these
values. PROC REPORT also generates a column heading that spans all these columns. Quoted
text in the DEFINE statement for Department customizes this heading. In traditional
(monospace) SAS output, PROC REPORT expands the heading with underscores to ll all
columns that are created by the across variable.
define department / across format=$deptfmt. _Department_;

Dene the analysis variable. Sales is an analysis variable that is used to calculate the sum
statistic. In each case, the value of Sales is the sum of Sales for all observations in one
department in one group. (In this case, the value represents a single observation.)
define sales / analysis sum format=dollar11.2 ;

Dene the computed variable. The COMPUTED option indicates that PROC REPORT must
compute values for Perish. You compute the variables values in a compute block that is
associated with Perish.
define perish / computed format=dollar11.2
Perishable*Total;

Produce a report summary. This BREAK statement creates a default summary after the last
row for each value of Manager. The only option that is in use is SKIP, which writes a blank line.
You can use this technique to double-space in many reports that contains a group or order
variable.
break after manager / skip;

Calculate values for the computed variable. This compute block computes the value of
Perish from the values for the Meat/Dairy department and the Produce department. Because
the variables Sales and Department collectively dene these columns, there is no way to
identify the values to PROC REPORT by name. Therefore, the assignment statement uses
column numbers to unambiguously specify the values to use. Each time PROC REPORT needs a
value for Perish, it sums the values in the third and fourth columns of that row of the report.
compute perish;
perish=sum(_c3_, _c4_);
endcomp;

Produce a customized summary. This compute block creates a customized summary at the
end of the report. The rst LINE statement writes 57 hyphens (-) starting in column 4.
Subsequent LINE statements write the quoted text in the specied columns and the values of
the variables _C3_, _C4_, and _C5_ with the DOLLAR11.2 format. Note that the pointer control
(@) is designed for the Listing destination. It has no effect on ODS destinations other than
traditional SAS monospace output.

The REPORT Procedure

Output

Select the observations to process. The WHERE statement selects for the report only the
observations for departments p1 and p2 in stores in the northeast or northwest sector.
where sector contains n
and (department=p1 or department=p2);

Specify the title.

title Sales Figures for Perishables in Northern Sectors;
run;

Output
Sales Figures for Perishables in Northern Sectors
_______Department_______
Meat/Dairy
Produce

Sector

Manager

Perishable
Total
--------------------------------------------------------Northeast

Alomar

$190.00

$86.00

$276.00

Andrews

$300.00

$125.00

$425.00

Brown

$250.00

$73.00

$323.00

Pelfrey

$205.00

$76.00

$281.00

Reveiz

$600.00

$30.00

$630.00

Northwest

--------------------------------------------------------|
Combined sales for meat and dairy :
$1,545.00
|
|
Combined sales for produce :
$390.00
|
|
|
|
Combined sales for all perishables:
$1,935.00
|
---------------------------------------------------------

963

964

Example 6: Displaying Multiple Statistics for One Variable

Chapter 42

Example 6: Displaying Multiple Statistics for One Variable

Procedure features:

PROC REPORT statement options:

LS=
PS=
COLUMN statement:
specifying statistics for stacked variables
DEFINE statement options:
FORMAT=
GROUP
ID
Data set:

GROCERY on page 949

Formats:

$MGRFMT. on page 949

The report in this example displays six statistics for the sales for each managers
store. The output is too wide to t all the columns on one page, so three of the statistics
appear on the second page of the report. In order to make it easy to associate the
statistics on the second page with their group, the report repeats the values of Manager
and Sector on every page of the report.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=80 pagesize=60
fmtsearch=(proclib);

The REPORT Procedure

Output

965

Specify the report columns. This COLUMN statement creates a column for Sector, Manager,
and each of the six statistics that are associated with Sales.
column sector manager (Sum Min Max Range Mean Std),sales;

Dene the group variables and the analysis variable. ID species that Manager is an ID
variable. An ID variable and all columns to its left appear at the left of every page of a report.
In this report, Sector and Manager are group variables. Each detail row of the report
consolidates the information for all observations with the same values of the group variables.
FORMAT= species the formats to use in the report.
define manager / group format=$mgrfmt. id;
define sector / group format=$sctrfmt.;
define sales / format=dollar11.2 ;

Specify the title.

title Sales Statistics for All Sectors;
run;

Output
Sales Statistics for All Sectors

Sum
Min
Max
Sector
Manager
Sales
Sales
Sales
--------------------------------------------------------Northeast
Northwest

Southeast
Southwest

Alomar
Andrews
Brown
Pelfrey
Reveiz
Jones
Smith
Adams
Taylor

$786.00
$1,045.00
$598.00
$746.00
$1,110.00
$630.00
$350.00
$695.00
$353.00

$86.00
$125.00
$45.00
$45.00
$30.00
$40.00
$50.00
$40.00
$50.00

$420.00
$420.00
$250.00
$420.00
$600.00
$300.00
$120.00
$350.00
$130.00

Sales Statistics for All Sectors

Range
Mean
Std
Sector
Manager
Sales
Sales
Sales
--------------------------------------------------------Northeast
Northwest

Southeast
Southwest

Alomar
Andrews
Brown
Pelfrey
Reveiz
Jones
Smith
Adams
Taylor

$334.00
$295.00
$205.00
$375.00
$570.00
$260.00
$70.00
$310.00
$80.00

$196.50
$261.25
$149.50
$186.50
$277.50
$157.50
$87.50
$173.75
$88.25

$156.57
$127.83
$105.44
$170.39
$278.61
$123.39
$29.86
$141.86
$42.65

966

Example 7: Storing and Reusing a Report Denition

Chapter 42

Example 7: Storing and Reusing a Report Denition

Procedure features:

PROC REPORT statement options:

NAMED
OUTREPT=
REPORT=
WRAP
Other features:

TITLE statement
WHERE statement
Data set: GROCERY on page 949
Formats:

$SCTRFMT., $MGRFMT. and $DEPTFMT. on page 949

The rst PROC REPORT step in this example creates a report that displays one
value from each column of the report, using two rows to do so, before displaying another
value from the rst column. (By default, PROC REPORT displays values for only as
many columns as it can t on one page. It lls a page with values for these columns
before starting to display values for the remaining columns on the next page.)
Each item in the report is identied in the body of the report rather than in a column
header.
The report denition created by the rst PROC REPORT step is stored in a catalog
entry. The second PROC REPORT step uses it to create a similar report for a different
sector of the city.

Program to Store a Report Denition

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=80 pagesize=60
fmtsearch=(proclib);

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). NAMED writes name= in front
of each value in the report, where name= is the column heading for the value. When you use
NAMED, PROC REPORT suppresses the display of column headings at the top of each page.
proc report data=grocery nowd
named

The REPORT Procedure

Output

967

wrap
ls=64 ps=36
outrept=proclib.reports.namewrap;

Specify the report columns. The report contains a column for Sector, Manager, Department,
and Sales.
column sector manager department sales;

Dene the display and analysis variables. Because no usage is specied in the DEFINE
statements, PROC REPORT uses the defaults. The character variables (Sector, Manager, and
Department) are display variables. Sales is an analysis variable that is used to calculate the
sum statistic. FORMAT= species the formats to use in the report.
define
define
define
define

sector / format=$sctrfmt.;
manager / format=$mgrfmt.;
department / format=$deptfmt.;
sales / format=dollar11.2;

Select the observations to process. A report denition might differ from the SAS program
that creates the report. In particular, PROC REPORT stores neither WHERE statements nor
TITLE statements.
where manager=1;

Specify the title. SYSDATE is an automatic macro variable that returns the date when the
SAS job or SAS session began. The TITLE statement uses double rather than single quotation
marks so that the macro variable resolves.
title "Sales Figures for Smith on &sysdate";
run;

Output

This is the output from the rst PROC REPORT step, which creates the
report denition.

Sales Figures for Smith on 04JAN02

Sector=Southeast
Sales=
$50.00
Sector=Southeast
Sales=
$100.00
Sector=Southeast
Sales=
$120.00
Sector=Southeast
Sales=
$80.00

Manager=Smith

Department=Paper

Manager=Smith

Department=Meat/Dairy

Manager=Smith

Department=Canned

Manager=Smith

Department=Produce

968

Program to Use a Report Denition

Chapter 42

Program to Use a Report Denition

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. FMTSEARCH= species the
library to include when searching for user-created formats.
options nodate pageno=1 fmtsearch=(proclib);

Specify the report options, load the report denition, and select the observations to
process. REPORT= uses the report denition that is stored in
PROCLIB.REPORTS.NAMEWRAP to produce the report. The second report differs from the
rst one because it uses different WHERE and TITLE statements.
proc report data=grocery report=proclib.reports.namewrap
nowd;
where sector=sw;
title "Sales Figures for the Southwest Sector on &sysdate";
run;

Output
Sales Figures for the Southwest Sector on 04JAN02
Sector=Southwest
Sector=Southwest
Sector=Southwest
Sector=Southwest
Sector=Southwest
Sector=Southwest
Sector=Southwest
Sector=Southwest

Manager=Taylor
Manager=Taylor
Manager=Taylor
Manager=Taylor
Manager=Adams
Manager=Adams
Manager=Adams
Manager=Adams

Department=Paper
Department=Meat/Dairy
Department=Canned
Department=Produce
Department=Paper
Department=Meat/Dairy
Department=Canned
Department=Produce

Sales Figures for the Southwest Sector on 04JAN02

Sales=
Sales=
Sales=
Sales=
Sales=
Sales=
Sales=
Sales=

$53.00
$130.00
$120.00
$50.00
$40.00
$350.00
$225.00
$80.00

Example 8: Condensing a Report into Multiple Panels

Procedure features:

PROC REPORT statement options:

The REPORT Procedure

Program

969

FORMCHAR=
HEADLINE
LS=
PANELS=
PS=
PSPACE=
BREAK statement options:
SKIP
Other features:

SAS system option FORMCHAR=

GROCERY on page 949
Formats: $MGRFMT. and $DEPTFMT. on page 949
Data set:

The report in this example

3 uses panels to condense a two-page report to one page. Panels compactly present
information for long, narrow reports by placing multiple rows of information side
by side.
3 uses a default summary to place a blank line after the last row for each manager.
3 changes the default underlining character for the duration of this PROC REPORT
step.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=80 pagesize=60
fmtsearch=(proclib);

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). HEADLINE underlines all
column headings and the spaces between them at the top of each panel of the report.
FORMCHAR= sets the value of the second formatting character (the one that HEADLINE
uses) to the tilde (~). Therefore, the tilde underlines the column headings in the output.
HEADSKIP writes a blank line beneath the underlining that HEADLINE writes. LS= sets the
line size for the report to 64, and PS= sets the page size to 18. PANELS= creates a multipanel
report. Specifying PANELS=99 ensures that PROC REPORT ts as many panels as possible on
one page. PSPACE=6 places six spaces between panels.
proc report data=grocery nowd headline
formchar(2)=~

970

Program

Chapter 42

panels=99 pspace=6
ls=64 ps=18;

Specify the report columns. The report contains a column for Manager, Department, and
Sales.
column manager department sales;

Dene the sort order and analysis columns. The values of all variables with the ORDER
option in the DEFINE statement determine the order of the rows in the report. In this report,
PROC REPORT arranges the rows rst by the value of Manager (because it is the rst variable
in the COLUMN statement) and then, within each value of Manager, by the values of
Department. The ORDER= option species the sort order for a variable. This report arranges
the values of Manager by their formatted values and arranges the values of Department by their
internal values (np1, np2, p1, and p2). FORMAT= species the formats to use in the report.
define manager / order
order=formatted
format=$mgrfmt.;
define department / order
order=internal
format=$deptfmt.;
define sales / format=dollar7.2;

Produce a report summary. This BREAK statement produces a default summary after the
last row for each manager. Because SKIP is the only option in the BREAK statement, each
break consists of only a blank line.
break after manager / skip;

Select the observations to process. The WHERE statement selects for the report only the
observations for stores in the northwest or southwest sector.
where sector=nw or sector=sw;

Specify the title.

title Sales for the Western Sectors;
run;

The REPORT Procedure

Example 9: Writing a Customized Summary on Each Page

971

Output
Sales for the Western Sectors
Manager Department
Sales
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Adams
Paper
$40.00
Canned
$225.00
Meat/Dairy $350.00
Produce
$80.00
Brown

Pelfrey

Paper
Canned
Meat/Dairy
Produce

$45.00
$230.00
$250.00
$73.00

Paper
Canned
Meat/Dairy
Produce

Manager Department
Sales
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Reveiz

Paper
Canned
Meat/Dairy
Produce

$60.00
$420.00
$600.00
$30.00

Taylor

Paper
Canned
Meat/Dairy
Produce

$53.00
$120.00
$130.00
$50.00

$45.00
$420.00
$205.00
$76.00

Example 9: Writing a Customized Summary on Each Page

Procedure features:

BREAK statement options:

OL
PAGE
SUMMARIZE
COMPUTE statement arguments:
with a computed variable as report-item
BEFORE break-variable
AFTER break-variable with conditional logic
BEFORE _PAGE_
DEFINE statement options:
NOPRINT
LINE statement:
pointer controls
quoted text
repeating a character string
variable values and formats
Data set: GROCERY on page 949
Formats: $SCTRFMT., $MGRFMT., and $DEPTFMT. on page 949

The report in this example displays a record of one days sales for each store. The
rows are arranged so that all the information about one store is together, and the
information for each store begins on a new page. Some variables appear in columns.
Others appear only in the page header that identies the sector and the stores manager.
The header that appears at the top of each page is created with the _PAGE_
argument in the COMPUTE statement.
Prot is a computed variable based on the value of Sales and Department.

972

Program

Chapter 42

The text that appears at the bottom of the page depends on the total of Sales for the
store. Only the rst two pages of the report appear here.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=30
fmtsearch=(proclib);

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). NOHEADER in the PROC
REPORT statement suppresses the default column headings.
proc report data=grocery nowd
headline headskip;

Specify the title.

title Sales for Individual Stores;

Specify the report columns. The report contains a column for Sector, Manager, Department,
Sales, and Prot, but the NOPRINT option suppresses the printing of the columns for Sector and
Manager. The page heading (created later in the program) includes their values. To get these
variable values into the page heading, Sector and Manager must be in the COLUMN statement.
column sector manager department sales Profit;

Dene the group, computed, and analysis variables. In this report, Sector, Manager, and
Department are group variables. Each detail row of the report consolidates the information for
all observations with the same values of the group variables. Prot is a computed variable
whose values are calculated in the next section of the program. FORMAT= species the formats
to use in the report.
define
define
define
define
define

sector / group noprint;

manager / group noprint;
profit / computed format=dollar11.2;
sales / analysis sum format=dollar11.2;
department / group format=$deptfmt.;

The REPORT Procedure

Program

973

Calculate the computed variable. Prot is computed as a percentage of Sales. For

nonperishable items, the prot is 40% of the sale price. For perishable items the prot is 25%.
Notice that in the compute block you must reference the variable Sales with a compound name
(Sales.sum) that identies both the variable and the statistic that you calculate with it.
compute profit;
if department=np1 or department=np2
then profit=0.4*sales.sum;
else profit=0.25*sales.sum;
endcomp;

Create a customized page header. This compute block executes at the top of each page, after
PROC REPORT writes the title. It writes the page heading for the current managers store. The
LEFT option left-justies the text in the LINE statements. Each LINE statement writes the
text in quotation marks just as it appears in the statement. The rst two LINE statements
write a variable value with the format specied immediately after the variables name.
compute before _page_ / left;
line sector $sctrfmt. Sector;
line Store managed by manager $mgrfmt.;
line ;
line ;
line ;
endcomp;

Produce a report summary. This BREAK statement creates a default summary after the last
row for each manager. OL writes a row of hyphens above the summary line. SUMMARIZE
writes the value of Sales (the only analysis or computed variable) in the summary line. The
PAGE option starts a new page after each default summary so that the page heading that is
created in the preceding compute block always pertains to the correct manager.
break after manager / ol summarize page;

Produce a customized summary. This compute block places conditional text in a customized
summary that appears after the last detail row for each manager.
compute after manager;

Specify the length of the customized summary text. The LENGTH statement assigns a
length of 35 to the temporary variable TEXT. In this particular case, the LENGTH statement is
unnecessary because the longest version appears in the rst IF/THEN statement. However,
using the LENGTH statement ensures that even if the order of the conditional statements
changes, TEXT will be long enough to hold the longest version.
length text $ 35;

974

Output

Chapter 42

Specify the conditional logic for the customized summary text. You cannot use the LINE
statement in conditional statements (IF-THEN, IF-THEN/ELSE, and SELECT) because it does
not take effect until PROC REPORT has executed all other statements in the compute block.
These IF-THEN/ELSE statements assign a value to TEXT based on the value of Sales.sum in
the summary row. A LINE statement writes that variable, whatever its value happens to be.
if sales.sum lt 500 then
text=Sales are below the target region.;
else if sales.sum ge 500 and sales.sum lt 1000 then
text=Sales are in the target region.;
else if sales.sum ge 1000 then
text=Sales exceeded goal!;
line ;
line text $35.;
endcomp;
run;

Output

Sales for Individual Stores

Northeast Sector
Store managed by Alomar

Department
Sales
Profit
-----------------------------------Canned
Meat/Dairy
Paper
Produce

$420.00
$190.00
$90.00
$86.00
----------$786.00

$168.00
$47.50
$36.00
$21.50
----------$196.50

Sales are in the target region.

Sales for Individual Stores

Northeast Sector
Store managed by Andrews

Department
Sales
Profit
-----------------------------------Canned
Meat/Dairy
Paper
Produce

$420.00
$300.00
$200.00
$125.00
----------$1,045.00

$168.00
$75.00
$80.00
$31.25
----------$261.25

Sales exceeded goal!

The REPORT Procedure

Program

975

Example 10: Calculating Percentages

Procedure features:

COLUMN statement arguments:

PCTSUM
SUM
spanning headers
COMPUTE statement options:
CHAR
LENGTH=
DEFINE statement options:
COMPUTED
FLOW
WIDTH=
RBREAK statement options:
OL
SUMMARIZE
Other features:

TITLE statement
Data set: GROCERY on page 949
Formats: $MGRFMT. and $DEPTFMT. on page 949

The summary report in this example shows the total sales for each store and the
percentage that these sales represent of sales for all stores. Each of these columns has
its own header. A single header also spans all the columns. This header looks like a
title, but it differs from a title because it would be stored in a report denition. You
must submit a null TITLE statement whenever you use the report denition, or the
report will contain both a title and the spanning header.
The report includes a computed character variable, COMMENT, that ags stores
with an unusually high percentage of sales. The text of COMMENT wraps across
multiple rows. It makes sense to compute COMMENT only for individual stores.
Therefore, the compute block that does the calculation includes conditional code that
prevents PROC REPORT from calculating COMMENT on the summary line.

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

976

Program

Chapter 42

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). HEADLINE underlines all
column headings and the spaces between them at the top of each page of the report. The null
TITLE statement suppresses the title of the report.
proc report data=grocery nowd headline;
title;

Specify the report columns. The COLUMN statement uses the text in quotation marks as a
spanning heading. The heading spans all the columns in the report because they are all
included in the pair of parentheses that contains the heading. The COLUMN statement
associates two statistics with Sales: Sum and Pctsum. The Sum statistic sums the values of
Sales for all observations that are included in a row of the report. The Pctsum statistic shows
what percentage of Sales that sum is for all observations in the report.
column (Individual Store Sales as a Percent of All Sales
sector manager sales,(sum pctsum) comment);

Dene the group and analysis columns. In this report, Sector and Manager are group
variables. Each detail row represents a set of observations that have a unique combination of
formatted values for all group variables. Sales is, by default, an analysis variable that is used to
calculate the Sum statistic. However, because statistics are associated with Sales in the column
statement, those statistics override the default. FORMAT= species the formats to use in the
report. Text between quotation marks species the column heading.
define manager / group
format=$mgrfmt.;
define sector / group
format=$sctrfmt.;
define sales / format=dollar11.2
;
define sum / format=dollar9.2
Total Sales;

Dene the percentage and computed columns. The DEFINE statement for Pctsum
species a column heading, a format, and a column width of 8. The PERCENT. format presents
the value of Pctsum as a percentage rather than a decimal. The DEFINE statement for
COMMENT denes it as a computed variable and assigns it a column width of 20 and a blank
column heading. The FLOW option wraps the text for COMMENT onto multiple lines if it
exceeds the column width.
define pctsum / Percent of Sales format=percent6. width=8;
define comment / computed width=20 flow;

Calculate the computed variable. Options in the COMPUTE statement dene COMMENT
as a character variable with a length of 40.
compute comment / char length=40;

The REPORT Procedure

Example 11: How PROC REPORT Handles Missing Values

977

Specify the conditional logic for the computed variable. For every store where sales
exceeded 15% of the sales for all stores, this compute block creates a comment that says Sales
substantially above expectations. Of course, on the summary row for the report, the
value of Pctsum is 100. However, it is inappropriate to ag this row as having exceptional sales.
The automatic variable _BREAK_ distinguishes detail rows from summary rows. In a detail row,
the value of _BREAK_ is blank. The THEN statement executes only on detail rows where the
value of Pctsum exceeds 0.15.
if sales.pctsum gt .15 and _break_ =
then comment=Sales substantially above expectations.;
else comment= ;
endcomp;

Produce the report summary. This RBREAK statement creates a default summary at the
end of the report. OL writes a row of hyphens above the summary line. SUMMARIZE writes the
values of Sales.sum and Sales.pctsum in the summary line.
rbreak after / ol summarize;
run;

Output
1
Individual Store Sales as a Percent of All Sales
Total
Percent
Sector
Manager
Sales of Sales
------------------------------------------------------------Northeast Alomar
$786.00
12%
Andrews $1,045.00
17%
Sales substantially
above expectations.
Northwest Brown
$598.00
9%
Pelfrey
$746.00
12%
Reveiz
$1,110.00
18%
Sales substantially
above expectations.
Southeast Jones
$630.00
10%
Smith
$350.00
6%
Southwest Adams
$695.00
11%
Taylor
$353.00
6%
--------- -------$6,313.00
100%

Example 11: How PROC REPORT Handles Missing Values

Procedure features:

PROC REPORT statement options:

MISSING

978

Program with Data Set with No Missing Values

Chapter 42

COLUMN statement
with the N statistic
Other features:

TITLE statement
Formats: $MGRFMT. on page 949

This example illustrates the difference between the way PROC REPORT handles
missing values for group (or order or across) variables with and without the MISSING
option. The differences in the reports are apparent if you compare the values of N for
each row and compare the totals in the default summary at the end of the report.

Program with Data Set with No Missing Values

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

Create the GROCMISS data set. GROCMISS is identical to GROCERY except that it
contains some observations with missing values for Sector, Manager, or both.
data grocmiss;
input Sector $
datalines;
se 1 np1 50
.
se 2 np1 40
se
nw 3 np1 60
nw
nw 4 np1 45
nw
nw 9 np1 45
nw
sw 5 np1 53
sw
. . np1 40
sw
ne 7 np1 90
ne
ne 8 np1 200
ne
;

Manager $ Department $ Sales @@;

1
2
3
4
9
5
6
.
8

p1
p1
p1
p1
p1
p1
p1
p1
p1

100
300
600
250
205
130
350
190
300

se
se
.
nw
nw
sw
sw
ne
ne

.
2
3
4
9
5
6
7
8

np2
np2
np2
np2
np2
np2
np2
np2
np2

120
220
420
230
420
120
225
420
420

se
se
nw
nw
nw
sw
sw
ne
ne

1
2
3
4
9
5
6
7
8

p2
p2
p2
p2
p2
p2
p2
p2
p2

80
70
30
73
76
50
80
86
125

The REPORT Procedure

Output with No Missing Values

979

Specify the report columns. The report contains columns for Sector, Manager, the N statistic,
and Sales.
column sector manager N sales;

Dene the group and analysis variables. In this report, Sector and Manager are group
variables. Sales is, by default, an analysis variable that is used to calculate the Sum statistic.
Each detail row represents a set of observations that have a unique combination of formatted
values for all group variables. The value of Sales in each detail row is the sum of Sales for all
observations in the group. In this PROC REPORT step, the procedure does not include
observations with a missing value for the group variable. FORMAT= species formats to use in
the report.
define sector / group format=$sctrfmt.;
define manager / group format=$mgrfmt.;
define sales / format=dollar9.2;

Produce a report summary. This RBREAK statement creates a default summary at the end
of the report. DOL writes a row of equal signs above the summary line. SUMMARIZE writes the
values of N and Sales.sum in the summary line.
rbreak after / dol summarize;

Specify the title.

title Summary Report for All Sectors and Managers;
run;

Output with No Missing Values

Summary Report for All Sectors and Managers
Sector
Manager
N
Sales
---------------------------------------Northeast Alomar
3
$596.00
Andrews
4 $1,045.00
Northwest Brown
4
$598.00
Pelfrey
4
$746.00
Reveiz
3
$690.00
Southeast Jones
4
$630.00
Smith
2
$130.00
Southwest Adams
3
$655.00
Taylor
4
$353.00
========= =========
31 $5,443.00

980

Program with Data Set with Missing Values

Chapter 42

Program with Data Set with Missing Values

Include the missing values. The MISSING option in the second PROC REPORT step includes
the observations with missing values for the group variable.
proc report data=grocmiss nowd headline missing;
column sector manager N sales;
define sector / group format=$sctrfmt.;
define manager / group format=$mgrfmt.;
define sales / format=dollar9.2;
rbreak after / dol summarize;
run;

Output with Missing Values

Summary Report for All Sectors and Managers

Sector
Manager
N
Sales
---------------------------------------1
$40.00
Reveiz
1
$420.00
Smith
1
$100.00
Northeast
1
$190.00
Alomar
3
$596.00
Andrews
4 $1,045.00
Northwest Brown
4
$598.00
Pelfrey
4
$746.00
Reveiz
3
$690.00
Southeast
1
$120.00
Jones
4
$630.00
Smith
2
$130.00
Southwest Adams
3
$655.00
Taylor
4
$353.00
========= =========
36 $6,313.00

Example 12: Creating and Processing an Output Data Set

Procedure features:

PROC REPORT statement options:

BOX
OUT=
DEFINE statement options:
ANALYSIS
GROUP
NOPRINT
SUM

The REPORT Procedure

Program to Create Output Data Set

981

Other features:

Data set options:

WHERE=
Data set: GROCERY on page 949
Formats: $MGRFMT. on page 949

This example uses WHERE processing as it builds an output data set. This
technique enables you to do WHERE processing after you have consolidated multiple
observations into a single row.
The rst PROC REPORT step creates a report (which it does not display) in which
each row represents all the observations from the input data set for a single manager.
The second PROC REPORT step builds a report from the output data set. This report
uses line-drawing characters to separate the rows and columns.

Program to Create Output Data Set

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

Specify the report options and columns. The NOWD option runs PROC REPORT without
the REPORT window and sends its output to the open output destination(s). OUT= creates the
output data set TEMP. The output data set contains a variable for each column in the report
(Manager and Sales) as well as for the variable _BREAK_, which is not used in this example.
Each observation in the data set represents a row of the report. Because Manager is a group
variable and Sales is an analysis variable that is used to calculate the Sum statistic, each row
in the report (and therefore each observation in the output data set) represents multiple
observations from the input data set. In particular, each value of Sales in the output data set is
the total of all values of Sales for that manager. The WHERE= data set option in the OUT=
option lters those rows as PROC REPORT creates the output data set. Only those observations
with sales that exceed $1,000 become observations in the output data set.
proc report data=grocery nowd
out=temp( where=(sales gt 1000) );
column manager sales;

982

Output Showing the Output Data Set

Chapter 42

Dene the group and analysis variables. Because the denitions of all report items in this
report include the NOPRINT option, PROC REPORT does not print a report. However, the
PROC REPORT step does execute and create an output data set.
define manager / group noprint;
define sales / analysis sum noprint;
run;

Output Showing the Output Data Set

This is the output data set that PROC REPORT creates. It is used as
the input set in the second PROC REPORT step.

The Data Set TEMP

Manager
3
8

Sales
1110
1045

_____________BREAK______________

Program That Uses the Output Data Set

Specify the report options and columns, dene the group and analysis columns, and
specify the titles. DATA= species the output data set from the rst PROC REPORT step as
the input data set for this report. The BOX option draws an outline around the output,
separates the column headings from the body of the report, and separates rows and columns of
data. The TITLE statements specify a title for the report.
proc report data=temp box nowd;
column manager sales;
define manager / group format=$mgrfmt.;
define sales / analysis sum format=dollar11.2;
title Managers with Daily Sales;
title2 of over;
title3 One Thousand Dollars;
run;

The REPORT Procedure

Program That Creates the Output Data Set

983

Report Based on the Output Data Set

Managers with Daily Sales
of over
One Thousand Dollars

---------------------|Manager
Sales|
|--------------------|
|Andrews|
$1,045.00|
|-------+------------|
|Reveiz |
$1,110.00|
----------------------

Example 13: Storing Computed Variables as Part of a Data Set

Procedure features:

PROC REPORT statement options:

OUT=
COMPUTE statement:
with a computed variable as report-item
DEFINE statement options:
COMPUTED
Other features: CHART procedure
Data set: GROCERY on page 949
Formats: $SCTRFMT. on page 949

The report in this example

3 creates a computed variable
3 stores it in an output data set
3 uses that data set to create a chart based on the computed variable.

Program That Creates the Output Data Set

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

984

Program That Creates the Output Data Set

Chapter 42

Delete any existing titles.

title;

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window and sends its output to the open output destination(s). OUT= creates the output data
set PROFIT.
proc report data=grocery nowd out=profit;

Specify the report columns. The report contains columns for Manager, Department, Sales,
and Prot, which is not in the input data set. Because the purpose of this report is to generate
an output data set to use in another procedure, the report layout simply uses the default usage
for all the data set variables to list all the observations. DEFINE statements for the data set
variables are unnecessary.
column sector manager department sales Profit;

Dene the computed column. The COMPUTED option tells PROC REPORT that Prot is
dened in a compute block somewhere in the PROC REPORT step.
define profit / computed;

Calculate the computed column. Prot is computed as a percentage of Sales. For

nonperishable items, the prot is 40% of the sale price. For perishable items the prot is 25%.
Notice that in the compute block, you must reference the variable Sales with a compound name
(Sales.sum) that identies both the variable and the statistic that you calculate with it.
/* Compute values for Profit. */
compute profit;
if department=np1 or department=np2 then profit=0.4*sales.sum;
else profit=0.25*sales.sum;
endcomp;
run;

The REPORT Procedure

Program That Uses the Output Data Set

985

The Output Data Set

This is the output data set that is created by PROC REPORT. It is used
as input for PROC CHART.

The Data Set PROFIT

Sector
se
se
se
se
se
se
se
se
nw
nw
nw
nw
nw
nw
nw
nw
nw
nw
nw
nw
sw
sw
sw
sw
sw
sw
sw
sw
ne
ne
ne
ne
ne
ne
ne
ne

Manager
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
9
9
9
9
5
5
5
5
6
6
6
6
7
7
7
7
8
8
8
8

Department
np1
p1
np2
p2
np1
p1
np2
p2
np1
p1
np2
p2
np1
p1
np2
p2
np1
p1
np2
p2
np1
p1
np2
p2
np1
p1
np2
p2
np1
p1
np2
p2
np1
p1
np2
p2

Sales
50
100
120
80
40
300
220
70
60
600
420
30
45
250
230
73
45
205
420
76
53
130
120
50
40
350
225
80
90
190
420
86
200
300
420
125

1
Profit
20
25
48
20
16
75
88
17.5
24
150
168
7.5
18
62.5
92
18.25
18
51.25
168
19
21.2
32.5
48
12.5
16
87.5
90
20
36
47.5
168
21.5
80
75
168
31.25

_BREAK__

Program That Uses the Output Data Set

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=80 pagesize=60
fmtsearch=(proclib);

986

Output from Processing the Output Data Set

Chapter 42

Chart the data in the output data set. PROC CHART uses the output data set from the
previous PROC REPORT step to chart the sum of Prot for each sector.
proc chart data=profit;
block sector / sumvar=profit;
format sector $sctrfmt.;
format profit dollar7.2;
title Sum of Profit by Sector;
run;

Output from Processing the Output Data Set

Sum of Profit by Sector

___
/_ /|
___
|**| |
/_ /|
|**| |
|**| |
|**| |
|**| |
|**| |
|**| |
|**| |
___
___
-|**| |--------|**| |---------/_ /|---------/_ /|------/ |**| |
/ |**| |
/ |**| |
/ |**| |
/
/ |**| |
/ |**| |
/ |**| |
/ |**| |
/
/
|**| |
/
|**| |
/
|**| |
/
|**| |
/
/
|**|/
/
|**|/
/
|**|/
/
|**|/
/
/
/
/
/
/
/
$627.25 /
$796.50 /
$309.50 /
$327.70 /
/-------------/-------------/-------------/-------------/
Northeast

Northwest

Southeast

Southwest

Sector

Example 14: Using a Format to Create Groups

Procedure features:

DEFINE statement options:

GROUP
Other features: FORMAT procedure
Data set:

GROCERY on page 949

Formats:

$MGRFMT. on page 949

This example shows how to use formats to control the number of groups that PROC
REPORT creates. The program creates a format for Department that classies the four
departments as one of two types: perishable or nonperishable. Consequently, when
Department is an across variable, PROC REPORT creates only two columns instead of
four. The column header is the formatted value of the variable.

The REPORT Procedure

Program

987

Program

Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
species the library to include when searching for user-created formats.
options nodate pageno=1 linesize=64 pagesize=60
fmtsearch=(proclib);

Create the $PERISH. format. PROC FORMAT creates a format for Department. This
variable has four different values in the data set, but the format has only two values.
proc format;
value $perish p1,p2=Perishable
np1,np2=Nonperishable;
run;

Specify the report options. The NOWD option runs the REPORT procedure without the
REPORT window and sends its output to the open output destination(s). HEADLINE underlines
all column headings and the spaces between them at the top of each page of the report.
HEADSKIP writes a blank line beneath the underlining that HEADLINE writes.
proc report data=grocery nowd
headline
headskip;

Specify the report columns. Department and Sales are separated by a comma in the
COLUMN statement, so they collectively determine the contents of the column that they dene.
Because Sales is an analysis variable, its values ll the cells that are created by these two
variables. The report also contains a column for Manager and a column for Sales by itself
(which is the sales for all departments).
column manager department,sales sales;

Dene the group and across variables. Manager is a group variable. Each detail row of the
report consolidates the information for all observations with the same value of Manager.
Department is an across variable. PROC REPORT creates a column and a column heading for
each formatted value of Department. ORDER=FORMATTED arranges the values of Manager
and Department alphabetically according to their formatted values. FORMAT= species the
formats to use. The empty quotation marks in the denition of Department specify a blank
column heading, so no heading spans all the departments. However, PROC REPORT uses the
formatted values of Department to create a column heading for each individual department.

988

Output

Chapter 42

define manager / group order=formatted

format=$mgrfmt.;
define department / across order=formatted
format=$perish. ;

Dene the analysis variable. Sales is an analysis variable that is used to calculate the Sum
statistic. Sales appears twice in the COLUMN statement, and the same denition applies to both
occurrences. FORMAT= species the format to use in the report. WIDTH= species the width of
the column. Notice that the column headings for the columns that both Department and Sales
create are a combination of the heading for Department and the (default) heading for Sales.
define sales / analysis sum
format=dollar9.2 width=13;

Produce a customized summary. This COMPUTE statement begins a compute block that
produces a customized summary at the end of the report. The LINE statement places the quoted
text and the value of Sales.sum (with the DOLLAR9.2 format) in the summary. An ENDCOMP
statement must end the compute block.
compute after;
line ;
line Total sales for these stores were:
sales.sum dollar9.2;
endcomp;

Specify the title.

title Sales Summary for All Stores;
run;

Output
Sales Summary for All Stores

Nonperishable
Perishable
Manager
Sales
Sales
Sales
---------------------------------------------------Adams
Alomar
Andrews
Brown
Jones
Pelfrey
Reveiz
Smith
Taylor

$265.00
$510.00
$620.00
$275.00
$260.00
$465.00
$480.00
$170.00
$173.00

$430.00
$276.00
$425.00
$323.00
$370.00
$281.00
$630.00
$180.00
$180.00

$695.00
$786.00
$1,045.00
$598.00
$630.00
$746.00
$1,110.00
$350.00
$353.00

Total sales for these stores were: $6,313.00

The REPORT Procedure

Program

989

Example 15: Specifying Style Elements for ODS Output in the PROC REPORT
Statement
Procedure features:

STYLE= option in the PROC REPORT statement

Other features:

ODS HTML statement

ODS PDF statement
ODS RTF statement
Data set:

GROCERY on page 949

Formats: $MGRFMT. and $DEPTFMT. on page 949

This example creates HTML, PDF, and RTF les and sets the style elements for each
location in the report in the PROC REPORT statement.

Program
Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Specify the ODS output lenames. By opening multiple ODS destinations, you can produce
multiple output les in a single execution. The ODS HTML statement produces output that is
written in HTML. The ODS PDF statement produces output in Portable Document Format
(PDF). The ODS RTF statement produces output in Rich Text Format (RTF). The output from
PROC REPORT goes to each of these les.
ods html body=external-HTML-file;
ods pdf file=external-PDF-file;
ods rtf file=external-RTF-file;

Specify the report options. The NOWD option runs PROC REPORT without the REPORT
window. In this case, SAS writes the output to the traditional procedure output, the HTML body
le, and the RTF and PDF les.
proc report data=grocery nowd headline headskip

Specify the style attributes for the report. This STYLE= option sets the style element for
the structural part of the report. Because no style element is specied, PROC REPORT uses all
the style attributes of the default style element for this location except for CELLSPACING=,
BORDERWIDTH=, and BORDERCOLOR=.
style(report)=[cellspacing=5 borderwidth=10 bordercolor=blue]

990

Program

Chapter 42

Specify the style attributes for the column headings. This STYLE= option sets the style
element for all column headings. Because no style element is specied, PROC REPORT uses all
the style attributes of the default style element for this location except for those that are
specied here.
style(header)=[foreground=yellow
font_style=italic font_size=6]

Specify the style attributes for the report columns. This STYLE= option sets the style
element for all the cells in all the columns. Because no style element is specied, PROC
REPORT uses all the style attributes of the default style element for this location except for
those that are specied here.
style(column)=[foreground=moderate brown
font_face=helvetica font_size=4]

Specify the style attributes for the compute block lines. This STYLE= option sets the
style element for all the LINE statements in all compute blocks. Because no style element is
specied, PROC REPORT uses all the style attributes of the default style element for this
location except for those that are specied here.
style(lines)=[foreground=white background=black
font_style=italic font_weight=bold font_size=5]

Specify the style attributes for report summaries. This STYLE= option sets the style
element for all the default summary lines. Because no style element is specied, PROC
REPORT uses all the style attributes of the default style element for this location except for
those that are specied here.
style(summary)=[foreground=cx3e3d73 background=cxaeadd9
font_face=helvetica font_size=3 just=r];

Specify the report columns. The report contains columns for Manager, Department, and
Sales.
column manager department sales;

Dene the sort order variables. In this report Manager and Department are order variables.
PROC REPORT arranges the rows rst by the value of Manager (because it is the rst variable
in the COLUMN statement), then by the value of Department. For Manager, ORDER= species
that values of Manager are arranged according to their formatted values; similarly, for
Department, ORDER= species that values of Department are arranged according to their
internal values. FORMAT= species the format to use for each variable. Text in quotation
marks species the column headings.
define manager / order
order=formatted
format=$mgrfmt.
Manager;
define department / order
order=internal
format=$deptfmt.
Department;

The REPORT Procedure

Program

991

Produce a report summary. The BREAK statement produces a default summary after the last
row for each manager. SUMMARIZE writes the values of Sales (the only analysis or computed
variable in the report) in the summary line. PROC REPORT sums the values of Sales for each
manager because Sales is an analysis variable that is used to calculate the Sum statistic.
break after manager / summarize;

Produce a customized summary. The COMPUTE statement begins a compute block that
produces a customized summary after each value of Manager. The LINE statement places the
quoted text and the values of Manager and Sales.sum (with the formats $MGRFMT. and
DOLLAR7.2) in the summary. An ENDCOMP statement must end the compute block.
compute after manager;
line Subtotal for manager $mgrfmt. is
sales.sum dollar7.2 .;
endcomp;

Produce a customized end-of-report summary. This COMPUTE statement begins a

compute block that executes at the end of the report. The LINE statement writes the quoted
text and the value of Sales.sum (with the DOLLAR7.2 format). An ENDCOMP statement must
end the compute block.
compute after;
line Total for all departments is:
sales.sum dollar7.2 .;
endcomp;

Select the observations to process. The WHERE statement selects for the report only the
observations for stores in the southeast sector.
where sector=se;

Specify the title.

title Sales for the Southeast Sector;
run;

Close the ODS destinations.

ods html close;
ods pdf close;
ods rtf close;

992

HTML Output

Chapter 42

HTML Output

The REPORT Procedure

PDF Output

993

994

RTF Output

Chapter 42

RTF Output

Example 16: Specifying Style Elements for ODS Output in Multiple

Statements
Procedure features:

STYLE= option in
PROC REPORT statement
CALL DEFINE statement
COMPUTE statement
DEFINE statement

The REPORT Procedure

Program

995

Other features:

ODS HTML statement

ODS PDF statement
ODS RTF statement
Data set:

GROCERY on page 949

Formats: $MGRFMT. on page 949 and $DEPTFMT. on page 949

This example creates HTML, PDF, and RTF les and sets the style elements for each
location in the report in the PROC REPORT statement. It then overrides some of these
settings by specifying style elements in other statements.

Program
Declare the PROCLIB library. The PROCLIB library is used to store user-created formats.
libname proclib SAS-data-library;

Specify the style attributes for the report. This STYLE= option sets the style element for
the structural part of the report. Because no style element is specied, PROC REPORT uses all
the style attributes of the default style element for this location except for those that are
specied here.
style(report)=[cellspacing=5 borderwidth=10 bordercolor=blue]

996

Program

Chapter 42

Specify the style attributes for the report summaries. This STYLE= option sets the style
element for all the default summary lines. Because no style element is specied, PROC
REPORT uses all the style attributes of the default style element for this location except for
those that are specied here.
style(summary)=[foreground=cx3e3d73 background=cxaeadd9
font_face=helvetica font_size=3 just=r];

Specify the report columns. The report contains columns for Manager, Department, and
Sales.
column manager department sales;

Dene the rst sort order variable. In this report Manager is an order variable. PROC
REPORT arranges the rows rst by the value of Manager (because it is the rst variable in the
COLUMN statement). ORDER= species that values of Manager are arranged according to
their formatted values. FORMAT= species the format to use for this variable. Text in quotation
marks species the column headings.
define manager / order
order=formatted
format=$mgrfmt.
Manager

Specify the style attributes for the rst sort order variable column heading. The
STYLE= option sets the foreground and background colors of the column heading for Manager.
The other style attributes for the column heading will match those that were established for the
HEADER location in the PROC REPORT statement.
style(header)=[foreground=white
background=black];

The REPORT Procedure

Program

997

Dene the second sort order variable. In this report Department is an order variable.
PROC REPORT arranges the rows rst by the value of Manager (because it is the rst variable
in the COLUMN statement), then by the value of Department. ORDER= species that values of
Department are arranged according to their internal values. FORMAT= species the format to
use for this variable. Text in quotation marks species the column heading.
define department / order
order=internal
format=$deptfmt.
Department

Specify the style attributes for the second sort order variable column.The STYLE=
option sets the font of the cells in the column Department to italic. The other style attributes for
the cells will match those that were established for the COLUMN location in the PROC
REPORT statement.
style(column)=[font_style=italic];

Produce a customized summary. The COMPUTE statement begins a compute block that
produces a customized summary at the end of the report. This STYLE= option species the style
element to use for the text that is created by the LINE statement in this compute block. This
style element switches the foreground and background colors that were specied for the LINES
location in the PROC REPORT statement. It also changes the font style, the font weight, and
the font size.
compute after manager
/ style=[font_style=roman font_size=3 font_weight=bold
background=white foreground=black];

Specify the text for the customized summary. The LINE statement places the quoted text
and the values of Manager and Sales.sum (with the formats $MGRFMT. and DOLLAR7.2) in
the summary. An ENDCOMP statement must end the compute block.
line Subtotal for manager $mgrfmt. is
sales.sum dollar7.2 .;
endcomp;

Produce a customized background for the analysis column. This compute block species
a background color and a bold font for all cells in the Sales column that contain values of 100 or
greater and that are not summary lines.
compute sales;
if sales.sum>100 and _break_= then
call define(_col_, "style",
"style=[background=yellow
font_face=helvetica
font_weight=bold]");
endcomp;

998

Program

Chapter 42

Produce a customized end-of-report summary. This COMPUTE statement begins a

Select the observations to process. The WHERE statement selects for the report only the
observations for stores in the southeast sector.
where sector=se;

Specify the title.

title Sales for the Southeast Sector;
run;

Close the ODS destinations.

ods html close;
ods pdf close;
ods rtf close;

The REPORT Procedure

HTML Body File

999

1000

PDF Output

Chapter 42

PDF Output

The REPORT Procedure

RTF Output

1001

1002

1003

CHAPTER

43
The SORT Procedure
Overview: SORT Procedure 1003
What Does the SORT Procedure Do? 1003
Sorting SAS Data Sets 1004
Syntax: SORT Procedure 1005
PROC SORT Statement 1005
BY Statement 1012
Concepts: SORT Procedure 1013
Multi-threaded Sorting 1013
Using PROC SORT with a DBMS 1013
Sorting Orders for Numeric Variables 1013
Sorting Orders for Character Variables 1014
Default Collating Sequence 1014
EBCDIC Order 1014
ASCII Order 1014
Specifying Sorting Orders for Character Variables 1015
Stored Sort Information 1015
Integrity Constraints: SORT Procedure 1015
Results: SORT Procedure 1016
Procedure Output 1016
Output Data Set 1016
Examples: SORT Procedure 1016
Example 1: Sorting by the Values of Multiple Variables 1017
Example 2: Sorting in Descending Order 1019
Example 3: Maintaining the Relative Order of Observations in Each BY Group
Example 4: Retaining the First Observation of Each BY Group 1023

1020

Overview: SORT Procedure

What Does the SORT Procedure Do?
The SORT procedure orders SAS data set observations by the values of one or more
character or numeric variables. The SORT procedure either replaces the original data
set or creates a new data set. PROC SORT produces only an output data set. For more
information, see Procedure Output on page 1016.
Operating Environment Information: The sorting capabilities that are described in this
chapter are available for all operating environments. In addition, if you use the HOST
value of the SAS system option SORTPGM=, you might be able to use other sorting
options that are available only for your operating environment. Refer to the SAS

1004

Sorting SAS Data Sets

Chapter 43

documentation for your operating environment for information about other sorting
capabilities 4

Sorting SAS Data Sets

In the following example, the original data set was in alphabetical order by last
name. PROC SORT replaces the original data set with a data set that is sorted by
employee identication number. Output 43.1 shows the log that results from running
this PROC SORT step. Output 43.2 shows the results of the PROC PRINT step. The
statements that produce the output follow:
proc sort data=employee;
by idnumber;
run;
proc print data=employee;
run;

Output 43.1

SAS Log Generated by PROC SORT

NOTE: There were 6 observations read from the data set WORK.EMPLOYEE.
NOTE: The data set WORK.EMPLOYEE has 6 observations and 3 variables.
NOTE: PROCEDURE SORT used:
real time
0.01 seconds
cpu time
0.01 seconds

Output 43.2

Observations Sorted by the Values of One Variable

The SAS System
Obs
1
2
3
4
5
6

Name
Belloit
Wesley
Lemeux
Arnsbarger
Pierce
Capshaw

IDnumber
1988
2092
4210
5466
5779
7338

The following output shows the results of a more complicated sort by three variables.
The businesses in this example are sorted by town, then by debt from highest amount
to lowest amount, then by account number. For an explanation of the program that
produces this output, see Example 2 on page 1019.

The SORT Procedure

Output 43.3

PROC SORT Statement

1005

Observations Sorted by the Values of Three Variables

Customers with Past-Due Accounts
Listed by Town, Amount, Account Number

Obs
1
2
3
4
5
6
7
8
9
10
11
12
13
14

Company

Town

Debt

Pauls Pizza
Peters Auto Parts
Watson Tabor Travel
Tinas Pet Shop
Apex Catering
Deluxe Hardware
Boyd & Sons Accounting
World Wide Electronics
Elway Piano and Organ
Ice Cream Delight
Tims Burger Stand
Strickland Industries
Paulines Antiques
Bobs Beds

Apex
Apex
Apex
Apex
Apex
Garner
Garner
Garner
Garner
Holly Springs
Holly Springs
Morrisville
Morrisville
Morrisville

83.00
65.79
37.95
37.95
37.95
467.12
312.49
119.95
65.79
299.98
119.95
657.22
302.05
119.95

Account
Number
1019
7288
3131
5108
9923
8941
4762
1122
5217
2310
6335
1675
9112
4998

Syntax: SORT Procedure

BY statement
Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See
Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.
See: SORT Procedure in the documentation for your operating environment.
Requirements:

PROC SORT <collating-sequence-option> < other option(s)>;

BY <DESCENDING> variable-1 << DESCENDING> variable-n>;

PROC SORT Statement

PROC SORT <collating-sequence-option> < other option(s)>;

To do this

Use this option

Specify the collating sequence

Specify ASCII

ASCII

Specify EBCDIC

EBCDIC

1006

PROC SORT Statement

Chapter 43

To do this

Use this option

Specify Danish

DANISH

Specify Finnish

FINNISH

Specify Norwegian

NORWEGIAN

Specify Swedish

SWEDISH

Specify a customized sequence

NATIONAL

Specify any of these collating sequences: ASCII,

EBCDIC, DANISH, FINNISH, ITALIAN,
NORWEGIAN, SPANISH, SWEDISH

SORTSEQ=

Specify the input data set

DATA=

Sort a SAS data set without changing the created and

modied dates

DATECOPY

Create output data sets

Species the output data set

OUT=

Species the output data set to which duplicate

observations are written

DUPOUT=

Specify the output order

Reverse the collation order for character
variables

REVERSE

Maintain relative order within BY groups

EQUALS

Do not maintain relative order within BY groups

NOEQUALS

Eliminate duplicate observations

Delete observations with duplicate BY values

NODUPKEY

Delete duplicate observations

NODUPRECS

Delete the input data set before the replacement

output data set is populated

OVERWRITE

Specify the available memory

SORTSIZE=

Force redundant sorting

FORCE

Reduce temporary disk usage

TAGSORT

Override SAS system option THREADS

Enable multi-threaded sorting

THREADS

Prevent multi-threaded sorting

NOTHREADS

Options
Options can include one collating-sequence-option and multiple other options. The
order of the two types of options does not matter and both types are not necessary in
the same PROC SORT step.

The SORT Procedure

PROC SORT Statement

1007

Collating-Sequence-Options

Operating Environment Information: For information about behavior specic to

your operating environment for the DANISH, FINNISH, NORWEGIAN, or
SWEDISH collating-sequence-option, see the SAS documentation for your operating
environment. 4
Restriction: You can specify only one collating-sequence-option in a PROC SORT

step.
ASCII

sorts character variables using the ASCII collating sequence. You need this option
only when you sort by ASCII on a system where EBCDIC is the native collating
sequence.
See also: Sorting Orders for Character Variables on page 1014
DANISH
NORWEGIAN

sorts characters according to the Danish and Norwegian national standard.

The Danish and Norwegian collating sequence is shown in Figure 43.1 on page
1008.
EBCDIC

sorts character variables using the EBCDIC collating sequence. You need this option
only when you sort by EBCDIC on a system where ASCII is the native collating
sequence.
See also: Sorting Orders for Character Variables on page 1014
FINNISH
SWEDISH

sorts characters according to the Finnish and Swedish national standard. The
Finnish and Swedish collating sequence is shown in Figure 43.1 on page 1008.
NATIONAL

sorts character variables using an alternate collating sequence, as dened by your

installation, to reect a countrys National Use Differences. To use this option, your
site must have a customized national sort sequence dened. Check with the SAS
Installation Representative at your site to determine if a customized national sort
sequence is available.
NORWEGIAN

See DANISH.
SORTSEQ=collating-sequence

species the collating sequence. The value of collating-sequence can be any one of the
collating-sequence-options in the PROC SORT statement, or the value can be the
name of a translation table, either a default translation table or one that you have
created in the TRANTAB procedure. For an example of using PROC TRANTAB and
PROC SORT with SORTSEQ=, see Using Different Translation Tables for Sorting in
SAS National Language Support (NLS): Users Guide. The available translation
tables are
Danish
Finnish
Italian
Norwegian
Spanish

1008

PROC SORT Statement

Chapter 43

Swedish
The following gure shows how the alphanumeric characters in each language will
sort.

Figure 43.1

National Collating Sequences of Alphanumeric Characters

CAUTION:

If you use a host sort utility to sort your data, then specifying the SORTSEQ= option
might corrupt the character BY variables. For more information, see the PROC SORT
documentation for your operating environment. 4
SWEDISH

See FINNISH.

Other Options
DATA=SAS-data-set

identies the input SAS data set.

Main discussion: Input Data Sets on page 19
DATECOPY

copies the SAS internal date and time when the SAS data set was created and the
date and time when it was last modied prior to the sort to the resulting sorted data
set. Note that the operating environment date and time are not preserved.
Restriction: DATECOPY can be used only when the resulting data set uses the V8
or V9 engine.
Tip: You can alter the le creation date and time with the DTC= option in the
MODIFY statement in PROC DATASETS. For more information, see MODIFY
Statement on page 348.
DUPOUT= SAS-data-set

species the output data set to which duplicate observations are written.
EQUALS | NOEQUALS

species the order of the observations in the output data set. For observations with
identical BY-variable values, EQUALS maintains the relative order of the
observations within the input data set in the output data set. NOEQUALS does not
necessarily preserve this order in the output data set.
Default: EQUALS
Interaction: When you use NODUPRECS or NODUPKEY to remove observations
in the output data set, the choice of EQUALS or NOEQUALS can affect which
observations are removed.

The SORT Procedure

PROC SORT Statement

1009

Interaction: The EQUALS | NOEQUALS procedure option overrides the default

sort stability behavior that is established with the SORTEQUALS |

NOSORTEQUALS system option.
Interaction: The EQUALS option is supported by the multi-threaded sort.

However, I/O performance may be reduced when using the EQUALS option with
the multi-threaded sort because partitioned data sets will be processed as if they
are non-partitioned data sets.
Interaction: The NOEQUALS option is supported by the multi-threaded sort. The

order of observations within BY groups that are returned by the multi-threaded

sort might not be consistent between runs. Therefore, using the NOEQUALS
option can produce inconsistent results in your output data sets.
Tip:

Using NOEQUALS can save CPU time and memory.

FORCE

sorts and replaces an indexed data set when the OUT= option is not specied.
Without the FORCE option, PROC SORT does not sort and replace an indexed data
set because sorting destroys user-created indexes for the data set. When you specify
FORCE, PROC SORT sorts and replaces the data set and destroys all user-created
indexes for the data set. Indexes that were created or required by integrity
constraints are preserved.
PROC SORT checks for the sort information before it sorts a data set so that
data is not re-sorted unnecessarily. By default, PROC SORT does not sort a data
set if the sort information matches the requested sort. You can use FORCE to
override this behavior. You might need to use FORCE if SAS cannot verify the sort
specication in the data set option SORTEDBY=. For more information about
SORTEDBY=, see the chapter on SAS data set options in SAS Language
Reference: Dictionary.

Tip:

Restriction: If you use PROC SORT with the FORCE option on data sets that were

created with the Version 5 compatibility engine or with a sequential engine such
as a tape format engine, you must also specify the OUT= option.
NODUPKEY

checks for and eliminates observations with duplicate BY values. If you specify this
option, then PROC SORT compares all BY values for each observation to those for
the previous observation that is written to the output data set. If an exact match is
found, then the observation is not written to the output data set.
Operating Environment Information: If you use the VMS operating environment
sort, then the observation that is written to the output data set is not always the
rst observation of the BY group. 4
Note: See NODUPRECS for information about eliminating duplicate
observations. 4
Interaction: When you are removing observations with duplicate BY values with

NODUPKEY, the choice of EQUALS or NOEQUALS can have an effect on which

observations are removed.
Use the EQUALS option with the NODUPKEY option for consistent results in
your output data sets.

Tip:

Featured in:

Example 4 on page 1023

NODUPRECS

checks for and eliminates duplicate observations. If you specify this option, then
PROC SORT compares all variable values for each observation to those for the
previous observation that was written to the output data set. If an exact match is
found, then the observation is not written to the output data set.

1010

PROC SORT Statement

Chapter 43

Note: See NODUPKEY for information about eliminating observations with

duplicate BY values. 4
Alias : NODUP
Interaction: When you are removing consecutive duplicate observations in the
output data set with NODUPRECS, the choice of EQUALS or NOEQUALS can
have an effect on which observations are removed.
Tip: Use the EQUALS option with the NODUPRECS option for consistent results
in your output data sets.
Interaction: The action of NODUPRECS is directly related to the setting of the
SORTDUP= system option. When SORTDUP= is set to LOGICAL, NODUPRECS
removes duplicate observations based on the examination of the variables that
remain after a DROP or KEEP operation on the input data set. Setting
SORTDUP=LOGICAL increases the number of duplicate observations that are
removed, because it eliminates variables before observation comparisons take
place. Also, setting SORTDUP=LOGICAL can improve performance, because
dropping variables before sorting reduces the amount of memory required to
perform the sort. When SORTDUP= is set to PHYSICAL, NODUPRECS examines
all variables in the data set, regardless of whether they have been kept or
dropped. For more information about SORTDUP=, see the chapter on SAS system
options in SAS Language Reference: Dictionary.
Tip: Because NODUPRECS checks only consecutive observations, some
nonconsecutive duplicate observations might remain in the output data set. You
can remove all duplicates with this option by sorting on all variables.
NOEQUALS

See EQUALS | NOEQUALS.

NOTHREADS

See THREADS|NOTHREADS.
OUT=SAS-data-set

names the output data set. If SAS-data-set does not exist, then PROC SORT creates
it.
CAUTION:

Use care when you use PROC SORT without OUT=. Without OUT=, data could be lost
if your system failed during execution of PROC SORT. 4
Default: Without OUT=, PROC SORT overwrites the original data set.
Tip : You can use data set options with OUT=.
Featured in:

Example 1 on page 1017

OVERWRITE

enables the input data set to be deleted before the replacement output data set is
populated with observations.
Restriction: The OVERWRITE option has no effect if you also specify the
TAGSORT option. You cannot overwrite the input data set because TAGSORT
must reread the input data set while populating the output data set.
Restriction: The OVERWRITE option is supported by the SAS sort and SAS
multi-threaded sort only. The option has no effect if you are using a host sort.
Tip: Using the OVERWRITE option can reduce disk space requirements.
CAUTION:

Use the OVERWRITE option only with a data set that is backed up or with a data set that
you can reconstruct. Because the input data set is deleted, data will be lost if a
failure occurs while the output data set is being written. 4

The SORT Procedure

PROC SORT Statement

1011

REVERSE

sorts character variables using a collating sequence that is reversed from the normal
collating sequence.
Operating Environment Information: For information about the normal collating
sequence for your operating environment, see EBCDIC Order on page 1014, ASCII
Order on page 1014, and the SAS documentation for your operating environment. 4
Interaction: Using REVERSE with the DESCENDING option in the BY statement

restores the sequence to the normal order.

Restriction: The REVERSE option cannot be used with a collating-sequence-option.

You can specify either a collating-sequence-option or the REVERSE option in a

PROC SORT, but you cannot specify both.
See also: The DESCENDING option in the BY statement. The difference is that the
DESCENDING option can be used with both character and numeric variables.
SORTSIZE=memory-specication

species the maximum amount of memory that is available to PROC SORT. Valid
values for memory-specication are as follows:
MAX
species that all available memory can be used.
n
species the amount of memory in bytes, where n is a real number.
nK
species the amount of memory in kilobytes, where n is a real number.
nM
species the amount of memory in megabytes, where n is a real number.
nG
species the amount of memory in gigabytes, where n is a real number.
Specifying the SORTSIZE= option in the PROC SORT statement temporarily
overrides the SAS system option SORTSIZE=. For more information about
SORTSIZE=, see the chapter on SAS system options in SAS Language Reference:
Dictionary.
Operating Environment Information: Some system sort utilities may treat this
option differently. Refer to the SAS documentation for your operating environment.

Default: the value of the SAS system option SORTSIZE=

Setting the SORTSIZE= option in the PROC SORT statement to MAX or 0, or

not setting the SORTSIZE= option, limits the PROC SORT to the available
physical memory based on the settings of the SAS system options that relate to
memory and information regarding available memory that is gathered from the
operating environment.

Tip:

Operating Environment Information: For information about the SAS system

options that relate to memory, see the SAS documentation for your operating
environment. 4
TAGSORT

stores only the BY variables and the observation numbers in temporary les. The BY
variables and the observation numbers are called tags. At the completion of the
sorting process, PROC SORT uses the tags to retrieve records from the input data set
in sorted order.
Restriction: The TAGSORT option is not compatible with the OVERWRITE option.
Interaction: The TAGSORT option is not supported by the multi-threaded sort.

1012

BY Statement

Chapter 43

When the total length of BY variables is small compared with the record
length, TAGSORT reduces temporary disk usage considerably. However,
processing time may be much higher.

Tip:

THREADS | NOTHREADS

enables or prevents the activation of multi-threaded sorting.

Default: the value of the SAS system option THREADS
Interaction: THREADS|NOTHREADS overrides the value of the SAS system

option THREADS. For more information about THREADS, see the chapter on SAS
system options in SAS Language Reference: Dictionary.
Interaction: The THREADS option is honored if the value of the SAS system option

CPUCOUNT is greater than 1.

Interaction: The TAGSORT option is not supported by the multi-threaded sort.

If THREADS is specied either as a SAS system option or in PROC SORT,

and another program has the input SAS data set open for reading, writing, or
updating using the SPDE engine, then the procedure might fail. In this case,
PROC SORT stops processing and writes a message to the SAS log.

Note:

Example 1 on page 1017, Example 2 on page 1019, and Example 4 on page

1023

BY < DESCENDING> variable-1 <<DESCENDING> variable-n>;

Required Arguments
variable

species the variable by which PROC SORT sorts the observations. PROC SORT
rst arranges the data set by the values in ascending order, by default, of the rst
BY variable. PROC SORT then arranges any observations that have the same value
of the rst BY variable by the values of the second BY variable in ascending order.
This sorting continues for every specied BY variable.

Option
DESCENDING

reverses the sort order for the variable that immediately follows in the statement so
that observations are sorted from the largest value to the smallest value.
Featured in:

Example 2 on page 1019

The SORT Procedure

Sorting Orders for Numeric Variables

1013

Concepts: SORT Procedure

Multi-threaded Sorting
The SAS system option THREADS activates multi-threaded sorting, which is new
with SAS System 9. Multi-threaded sorting achieves a degree of parallelism in the
sorting operations. This parallelism is intended to reduce the real-time to completion
for a given operation at the possible cost of additional CPU resources. For more
information, see the section on Support for Parallel Processing in SAS Language
Reference: Concepts.
The performance of the multi-threaded sort will be affected by the value of the SAS
system option CPUCOUNT=. CPUCOUNT= suggests how many system CPUs are
available for use by the multi-threaded sort.
The multi-threaded sort supports concurrent input from the partitions of a
partitioned data set.
Note: These partitioned data sets should not be confused with partitioned data sets
on z/OS. 4
Operating Environment Information: For information about the support of partitioned
data sets in your operating environment, see the SAS documentation for your operating
environment. 4
For more information about THREADS and CPUCOUNT=, see the chapter on SAS
system options in SAS Language Reference: Dictionary.

Using PROC SORT with a DBMS

When you use a DBMS data source, the observation ordering that is produced by
PROC SORT depends on whether the DBMS or SAS performs the sorting. If you use
the BEST value of the SAS system option SORTPGM=, then either the DBMS or SAS
will perform the sort. If the DBMS performs the sort, then the conguration and
characteristics of the DBMS sorting program will affect the resulting data order. Most
database management systems do not guarantee sort stability, and the sort might be
performed by the DBMS regardless of the state of the SORTEQUALS/
NOSORTEQUALS system option and EQUALS/NOEQUALS procedure option.
If you set the SAS system option SORTPGM= to SAS, then unordered data is
delivered from the DBMS to SAS and SAS performs the sorting. However, consistency
in the delivery order of observations from a DBMS is not guaranteed. Therefore, even
though SAS can perform a stable sort on the DBMS data, SAS cannot guarantee that
the ordering of observations within output BY groups will be the same, run after run.
To achieve consistency in the ordering of observations within BY groups, rst populate
a SAS data set with the DBMS data, then use the EQUALS or SORTEQUALS option to
perform a stable sort.

Sorting Orders for Numeric Variables

For numeric variables, the smallest-to-largest comparison sequence is
1 SAS missing values (shown as a period or special missing value)
2 negative numeric values

1014

Sorting Orders for Character Variables

Chapter 43

3 zero
4 positive numeric values.

Sorting Orders for Character Variables

Default Collating Sequence

By default, PROC SORT uses either the EBCDIC or the ASCII collating sequence
when it compares character values, depending on the environment under which the
procedure is running.

EBCDIC Order
The z/OS operating environment uses the EBCDIC collating sequence.
The sorting order of the English-language EBCDIC sequence is
blank . < ( + | & ! $ * );

- / , % _ > ?: # @ = "

abcdefghijklmnopqr~stuvwxyz
{ A B C D E F G H I } J K L M N O P Q R \S T
UVWXYZ
0123456789

The main features of the EBCDIC sequence are that lowercase letters are sorted
before uppercase letters, and uppercase letters are sorted before digits. Note also that
some special characters interrupt the alphabetic sequences. The blank is the smallest
character that you can display.

ASCII Order
The operating environments that use the ASCII collating sequence include
3 UNIX and its derivatives
3 OpenVMS
3 Windows.
From the smallest to the largest character that you can display, the English-language
ASCII sequence is
blank ! " # $ % & ( )* + , - . /0 1 2 3 4 5 6 7 8 9 : ; < = > ? @
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z[ \] _
abcdefghijklmnopqrstuvwxyz{}~

The main features of the ASCII sequence are that digits are sorted before uppercase
letters, and uppercase letters are sorted before lowercase letters. The blank is the
smallest character that you can display.

The SORT Procedure

Integrity Constraints: SORT Procedure

1015

Specifying Sorting Orders for Character Variables

The options EBCDIC, ASCII, NATIONAL, DANISH, SWEDISH, and REVERSE
specify collating sequences that are stored in the HOST catalog.
If you want to provide your own collating sequences or change a collating sequence
provided for you, then use the TRANTAB procedure to create or modify translation
tables. For complete details, see the TRANTAB procedure in SAS National Language
Support (NLS): Users Guide. When you create your own translation tables, they are
stored in your PROFILE catalog, and they override any translation tables that have the
same name in the HOST catalog.
Note: System managers can modify the HOST catalog by copying newly created
tables from the PROFILE catalog to the HOST catalog. Then all users can access the
new or modied translation table. 4

Stored Sort Information

PROC SORT records the BY variables, collating sequence, and character set that it
uses to sort the data set. This information is stored with the data set to help avoid
unnecessary sorts.
Before PROC SORT sorts a data set, it checks the stored sort information. If you try
to sort a data set the way that it is currently sorted, then PROC SORT does not
perform the sort and writes a message to the log to that effect. To override this
behavior, use the FORCE option. If you try to sort a data set the way that it is
currently sorted and you specify an OUT= data set, then PROC SORT simply makes a
copy of the DATA= data set.
To override the sort information that PROC SORT stores, use the _NULL_ value
with the SORTEDBY= data set option. For more information about SORTEDBY=, see
the chapter on SAS data set options in SAS Language Reference: Dictionary.
If you want to change the sort information for an existing data set, then use the
SORTEDBY= data set option in the MODIFY statement in the DATASETS procedure.
For more information, see MODIFY Statement on page 348.
To access the sort information that is stored with a data set, use the CONTENTS
statement in PROC DATASETS. For more information, see CONTENTS Statement on
page 323.

Integrity Constraints: SORT Procedure

Sorting the input data set and replacing it with the sorted data set preserves both
referential and general integrity constraints, as well as any indexes that they may
require. A sort that creates a new data set will not preserve any integrity constraints or
indexes. For more information about implicit replacement, explicit replacement, and no
replacement with and without the OUT= option, see Output Data Set on page 1016.
For more information about integrity constraints, see the chapter on SAS data les in
SAS Language Reference: Concepts.

1016

Results: SORT Procedure

Chapter 43

Results: SORT Procedure

Procedure Output
PROC SORT produces only an output data set. To see the output data set, you can
use PROC PRINT, PROC REPORT, or another of the many available methods of
printing in SAS.

Output Data Set

Without the OUT= option, PROC SORT replaces the original data set with the sorted
observations when the procedure executes without errors. When you specify the OUT=
option using a new data set name, PROC SORT creates a new data set that contains
the sorted observations.
To do this

Use this statement

implicit replacement of input data set

proc sort data=names;

explicit replacement of input data set

proc sort data=names out=names;

no replacement of input data set

proc sort data=names out=namesbyid;

With all three replacement options (implicit replacement, explicit replacement, and no
replacement) there must be at least enough space in the output data library for a copy
of the original data set.
You can also sort compressed data sets. If you specify a compressed data set as the
input data set and omit the OUT= option, then the input data set is sorted and remains
compressed. If you specify an OUT= data set, then the resulting data set is compressed
only if you choose a compression method with the COMPRESS= data set option. For
more information about COMPRESS=, see the chapter on SAS data set options in SAS
Language Reference: Dictionary.
Note: If the SAS system option NOREPLACE is in effect, then you cannot replace
an original permanent data set with a sorted version. You must either use the OUT=
option or specify the SAS system option REPLACE in an OPTIONS statement. The
SAS system option NOREPLACE does not affect temporary SAS data sets. 4

Examples: SORT Procedure

The SORT Procedure

Program

1017

Example 1: Sorting by the Values of Multiple Variables

Procedure features:

PROC SORT statement option:

OUT=
BY statement
Other features:

PROC PRINT

This example

3 sorts the observations by the values of two variables

3 creates an output data set for the sorted observations
3 prints the results.

Program

Create the input data set ACCOUNT. ACCOUNT contains the name of each business that
owes money, the amount of money that it owes on its account, the account number, and the town
where the business is located.
data account;
input Company $ 1-22
Town $ 39-51;
datalines;
Pauls Pizza
World Wide Electronics
Strickland Industries
Ice Cream Delight
Watson Tabor Travel
Boyd & Sons Accounting
Bobs Beds
Tinas Pet Shop
Elway Piano and Organ
Tims Burger Stand
Peters Auto Parts
Deluxe Hardware
Paulines Antiques
Apex Catering
;

Debt 25-30 AccountNumber 33-36

83.00
119.95
657.22
299.98
37.95
312.49
119.95
37.95
65.79
119.95
65.79
467.12
302.05
37.95

1019
1122
1675
2310
3131
4762
4998
5108
5217
6335
7288
8941
9112
9923

Apex
Garner
Morrisville
Holly Springs
Apex
Garner
Morrisville
Apex
Garner
Holly Springs
Apex
Garner
Morrisville
Apex

1018

Output

Chapter 43

Create the output data set BYTOWN. OUT= creates a new data set for the sorted
observations.
proc sort data=account out=bytown;

Sort by two variables. The BY statement species that the observations should be rst
ordered alphabetically by town and then by company.
by town company;
run;

Print the output data set BYTOWN. PROC PRINT prints the data set BYTOWN.
proc print data=bytown;

Specify the variables to print. The VAR statement species the variables to print and their
column order in the output.
var company town debt accountnumber;

Specify the titles.

title Customers with Past-Due Accounts;
title2 Listed Alphabetically within Town;
run;

Output

Customers with Past-Due Accounts

Listed Alphabetically within Town

Obs
1
2
3
4
5
6
7
8
9
10
11
12
13
14

Company

Town

Apex Catering
Pauls Pizza
Peters Auto Parts
Tinas Pet Shop
Watson Tabor Travel
Boyd & Sons Accounting
Deluxe Hardware
Elway Piano and Organ
World Wide Electronics
Ice Cream Delight
Tims Burger Stand
Bobs Beds
Paulines Antiques
Strickland Industries

Apex
Apex
Apex
Apex
Apex
Garner
Garner
Garner
Garner
Holly Springs
Holly Springs
Morrisville
Morrisville
Morrisville

Debt

Account
Number

37.95
83.00
65.79
37.95
37.95
312.49
467.12
65.79
119.95
299.98
119.95
119.95
302.05
657.22

9923
1019
7288
5108
3131
4762
8941
5217
1122
2310
6335
4998
9112
1675

The SORT Procedure

Program

1019

Example 2: Sorting in Descending Order

Procedure features:

This example BY statement option:

DESCENDING
Other features

PROC PRINT
Data set:

ACCOUNT on page 1017

3 sorts the observations by the values of three variables

3 sorts one of the variables in descending order
3 prints the results.

Program

Create the output data set SORTED. OUT= creates a new data set for the sorted
observations.
proc sort data=account out=sorted;

Sort by three variables with one in descending order. The BY statement species that
observations should be rst ordered alphabetically by town, then by descending value of amount
owed, then by ascending value of the account number.
by town descending debt accountnumber;
run;

Print the output data set SORTED. PROC PRINT prints the data set SORTED.
proc print data=sorted;

Specify the variables to print. The VAR statement species the variables to print and their
column order in the output.
var company town debt accountnumber;

1020

Output

Chapter 43

Specify the titles.

title Customers with Past-Due Accounts;
title2 Listed by Town, Amount, Account Number;
run;

Output

Note that sorting last by AccountNumber puts the businesses in Apex with a debt of $37.95 in
order of account number.

Customers with Past-Due Accounts

Listed by Town, Amount, Account Number

Obs
1
2
3
4
5
6
7
8
9
10
11
12
13
14

Company

Town

Apex
Apex
Apex
Apex
Apex
Garner
Garner
Garner
Garner
Holly Springs
Holly Springs
Morrisville
Morrisville
Morrisville

Debt

Account
Number

83.00
65.79
37.95
37.95
37.95
467.12
312.49
119.95
65.79
299.98
119.95
657.22
302.05
119.95

1019
7288
3131
5108
9923
8941
4762
1122
5217
2310
6335
1675
9112
4998

Example 3: Maintaining the Relative Order of Observations in Each BY Group

Procedure features:

PROC SORT statement option:

EQUALS|NOEQUALS
Other features: PROC PRINT

This example

3 sorts the observations by the value of the rst variable

3 maintains the relative order with the EQUALS option
3 does not maintain the relative order with the NOEQUALS option.

The SORT Procedure

Program

1021

Program

Create the input data set INSURANCE. INSURANCE contains the number of years worked
by all insured employees and their insurance ids.
data insurance;
input YearsWorked 1 InsuranceID 3-5;
datalines;
5 421
5 336
1 209
1 564
3 711
3 343
4 212
4 616
;

Create the output data set BYYEARS1 with the EQUALS option. OUT= creates a new
data set for the sorted observations. The EQUALS option maintains the order of the
observations relative to each other.
proc sort data=insurance out=byyears1 equals;

Sort by the rst variable. The BY statement species that the observations should be ordered
numerically by the number of years worked.
by yearsworked;
run;

Print the output data set BYYEARS1. PROC PRINT prints the data set BYYEARS1.
proc print data=byyears1;

Specify the variables to print. The VAR statement species the variables to print and their
column order in the output.
var yearsworked insuranceid;

1022

Program

Chapter 43

Specify the title.

title Sort with EQUALS;
run;

Create the output data set BYYEARS2. OUT= creates a new data set for the sorted
observations. The NOEQUALS option will not maintain the order of the observations relative to
each other.
proc sort data=insurance out=byyears2 noequals;

Sort by the rst variable. The BY statement species that the observations should be ordered
numerically by the number of years worked.
by yearsworked;
run;

Print the output data set BYYEARS2. PROC PRINT prints the data set BYYEARS2.
proc print data=byyears2;

Specify the variables to print. The VAR statement species the variables to print and their
column order in the output.
var yearsworked insuranceid;

Specify the title.

title Sort with NOEQUALS;
run;

The SORT Procedure

Example 4: Retaining the First Observation of Each BY Group

1023

Output

Note that sorting with the EQUALS option versus sorting with the NOEQUALS option causes a
different sort order for the observations where YearsWorked=3.

Sort with EQUALS

Years
Worked

Obs
1
2
3
4
5
6
7
8

1
1
3
3
4
4
5
5

Insurance
ID
209
564
711
343
212
616
421
336

Sort with NOEQUALS

Obs
1
2
3
4
5
6
7
8

Years
Worked
1
1
3
3
4
4
5
5

Insurance
ID
209
564
343
711
212
616
421
336

Example 4: Retaining the First Observation of Each BY Group

Procedure features:

PROC SORT statement option:

NODUPKEY
BY statement
Other features:

PROC PRINT
Data set: ACCOUNT on page 1017
Interaction: The EQUALS option, which is the default, must be in effect to ensure that
the rst observation for each BY group is the one that is retained by the NODUPKEY
option. If the NOEQUALS option has been specied, then one observation for each BY
group will still be retained by the NODUPKEY option, but not necessarily the rst
observation.

In this example, PROC SORT creates an output data set that contains only the rst
observation of each BY group. The NODUPKEY option prevents an observation from

1024

Program

Chapter 43

being written to the output data set when its BY value is identical to the BY value of
the last observation written to the output data set. The resulting report contains one
observation for each town where the businesses are located.

Program

Create the output data set TOWNS but include only the rst observation of each BY
group. NODUPKEY writes only the rst observation of each BY group to the new data set
TOWNS.

Operating Environment Information: If you use the VMS operating environment sort,
then the observation that is written to the output data set is not always the rst
observation of the BY group. 4
proc sort data=account out=towns nodupkey;

Sort by one variable. The BY statement species that observations should be ordered by town.
by town;
run;

Print the output data set TOWNS. PROC PRINT prints the data set TOWNS.
proc print data=towns;

Specify the variables to print. The VAR statement species the variables to print and their
column order in the output.
var town company debt accountnumber;

Specify the title.

title Towns of Customers with Past-Due Accounts;
run;

The SORT Procedure

Output

The output data set contains only four observations, one for each town in the input data set.

Towns of Customers with Past-Due Accounts

Obs
1
2
3
4

Town

Company

Apex
Garner
Holly Springs
Morrisville

Pauls Pizza
World Wide Electronics
Ice Cream Delight
Strickland Industries

Debt
83.00
119.95
299.98
657.22

1
Account
Number
1019
1122
2310
1675

1025

1026

1027

CHAPTER

44
The SQL Procedure
Overview: SQL Procedure 1029
What Is the SQL Procedure? 1029
What Are PROC SQL Tables? 1029
What Are Views? 1029
SQL Procedure Coding Conventions 1030
Syntax: SQL Procedure 1031
PROC SQL Statement 1033
ALTER TABLE Statement 1038
CONNECT Statement 1042
CREATE INDEX Statement 1043
CREATE TABLE Statement 1045
CREATE VIEW Statement 1049
DELETE Statement 1051
DESCRIBE Statement 1052
DISCONNECT Statement 1053
DROP Statement 1054
EXECUTE Statement 1055
INSERT Statement 1056
RESET Statement 1058
SELECT Statement 1058
UPDATE Statement 1069
VALIDATE Statement 1070
SQL Procedure Component Dictionary 1071
BETWEEN condition 1071
BTRIM function 1072
CALCULATED 1073
CASE expression 1073
COALESCE Function 1075
column-denition 1075
column-modier 1076
column-name 1078
CONNECTION TO 1079
CONTAINS condition 1079
EXISTS condition 1080
IN condition 1080
IS condition 1081
joined-table 1082
LIKE condition 1091
LOWER function 1093
query-expression 1093
sql-expression 1099

1028

Contents

Chapter 44

SUBSTRING function 1106

summary-function 1107
table-expression 1113
UPPER function 1114
Concepts: SQL Procedure 1114
Using SAS Data Set Options with PROC SQL 1114
Connecting to a DBMS Using the SQL Procedure Pass-Through Facility 1115
What Is the Pass-Through Facility? 1115
Return Codes 1115
Connecting to a DBMS Using the LIBNAME Statement 1115
Using the DICTIONARY Tables 1116
What Are DICTIONARY Tables? 1116
Retrieving Information about DICTIONARY Tables and SASHELP Views 1117
Using DICTIONARY Tables 1118
DICTIONARY Tables and Performance 1118
Using Macro Variables Set by PROC SQL 1119
Updating PROC SQL and SAS/ACCESS Views 1121
PROC SQL and the ANSI Standard 1122
Compliance 1122
SQL Procedure Enhancements 1122
Reserved Words 1122
Column Modiers 1123
Alternate Collating Sequences 1123
ORDER BY Clause in a View Denition 1123
In-Line Views 1123
Outer Joins 1123
Arithmetic Operators 1123
Orthogonal Expressions 1123
Set Operators 1123
Statistical Functions 1124
SAS DATA Step Functions 1124
SQL Procedure Omissions 1124
COMMIT Statement 1124
ROLLBACK Statement 1124
Identiers and Naming Conventions 1124
Granting User Privileges 1124
Three-Valued Logic 1124
Embedded SQL 1125
Examples: SQL Procedure 1125
Example 1: Creating a Table and Inserting Data into It 1125
Example 2: Creating a Table from a Querys Result 1127
Example 3: Updating Data in a PROC SQL Table 1129
Example 4: Joining Two Tables 1131
Example 5: Combining Two Tables 1134
Example 6: Reporting from DICTIONARY Tables 1136
Example 7: Performing an Outer Join 1138
Example 8: Creating a View from a Querys Result 1143
Example 9: Joining Three Tables 1145
Example 10: Querying an In-Line View 1148
Example 11: Retrieving Values with the SOUNDS-LIKE Operator 1150
Example 12: Joining Two Tables and Calculating a New Value 1152
Example 13: Producing All the Possible Combinations of the Values in a Column 1154
Example 14: Matching Case Rows and Control Rows 1158
Example 15: Counting Missing Values with a SAS Macro 1160

The SQL Procedure

What Are Views?

1029

Overview: SQL Procedure

What Is the SQL Procedure?
The SQL procedure implements Structured Query Language (SQL) for SAS. SQL is a
standardized, widely used language that retrieves data from and updates data in tables
and the views that are based on those tables.
The SAS SQL procedure enables you to

3
3
3
3

retrieve and manipulate data that is stored in tables or views.

create tables, views, and indexes on columns in tables.
create SAS macro variables that contain values from rows in a querys result.
add or modify the data values in a tables columns or insert and delete rows. You
can also modify the table itself by adding, modifying, or dropping columns.

3 send DBMS-specic SQL statements to a database management system (DBMS)

and retrieve DBMS data.
The following gure summarizes the variety of source material that you can use with
PROC SQL and what the procedure can produce.

Figure 44.1

PROC SQL Input and Output

PROC SQL tables

(SAS data files)
SAS data views
(PROC SQL views)
(DATA step views)
(SAS/ACCESS views)

reports
PROC SQL tables
(SAS data files)
PROC
SQL

DBMS tables

macro variables
DBMS tables
PROC SQL views

What Are PROC SQL Tables?

A PROC SQL table is synonymous with a SAS data le and has a member type of
DATA. You can use PROC SQL tables as input into DATA steps and procedures.
You create PROC SQL tables from SAS data les, from SAS data views, or from
DBMS tables by using PROC SQLs Pass-Through Facility or the SAS/ACCESS
LIBNAME statement. The Pass-Through Facility is described in Connecting to a
DBMS Using the SQL Procedure Pass-Through Facility on page 1115. The
SAS/ACCESS LIBNAME statement is described in Connecting to a DBMS Using the
LIBNAME Statement on page 1115.
In PROC SQL terminology, a row in a table is the same as an observation in a SAS
data le. A column is the same as a variable.

What Are Views?

A SAS data view denes a virtual data set that is named and stored for later use. A
view contains no data but describes or denes data that is stored elsewhere. There are
three types of SAS data views:

1030

SQL Procedure Coding Conventions

Chapter 44

3 PROC SQL views

3 SAS/ACCESS views
3 DATA step views.
You can refer to views in queries as if they were tables. The view derives its data
from the tables or views that are listed in its FROM clause. The data that is accessed
by a view is a subset or superset of the data that is in its underlying table(s) or view(s).
A PROC SQL view is a SAS data set of type VIEW that is created by PROC SQL. A
PROC SQL view contains no data. It is a stored query expression that reads data
values from its underlying les, which can include SAS data les, SAS/ACCESS views,
DATA step views, other PROC SQL views, or DBMS data. When executed, a PROC
SQL views output can be a subset or superset of one or more underlying les.
SAS/ACCESS views and DATA step views are similar to PROC SQL views in that
they are both stored programs of member type VIEW. SAS/ACCESS views describe data
in DBMS tables from other software vendors. DATA step views are stored DATA step
programs.
Note: Starting in SAS System 9, PROC SQL views, the Pass-Through Facility, and
the SAS/ACCESS LIBNAME statement are the preferred ways to access relational
DBMS data; SAS/ACCESS views are no longer recommended. You can convert existing
SAS/ACCESS views to PROC SQL views by using the CV2VIEW procedure. See The
CV2VIEW Procedure in SAS/ACCESS for Relational Databases: Reference for more
information. 4
You can update data through a PROC SQL or SAS/ACCESS view with certain
restrictions. See Updating PROC SQL and SAS/ACCESS Views on page 1121.
You can use all types of views as input to DATA steps and procedures.
Note: In this chapter, the term view collectively refers to PROC SQL views, DATA
step views, and SAS/ACCESS views, unless otherwise noted. 4
Note: When the contents of an SQL view are processed (by a DATA step or a
procedure), the referenced data set must be opened to retrieve information about the
variables that is not stored in the view. If that data set has a libref associated with it
that is not dened in the current SAS code, then an error will result. You can avoid this
error by specifying a USING clause in the CREATE VIEW statement. See CREATE
VIEW Statement on page 1049 for details. 4

SQL Procedure Coding Conventions

Because PROC SQL implements Structured Query Language, it works somewhat
differently from other base SAS procedures, as described here:
3 When a PROC SQL statement is executed, PROC SQL continues to run until a
QUIT statement, a DATA step, or another SAS procedure is executed. Therefore,
you do not need to repeat the PROC SQL statement with each SQL statement.
You need to repeat the PROC SQL statement only if you execute a QUIT
statement, a DATA step, or another SAS procedure between SQL statements.
3 SQL procedure statements are divided into clauses. For example, the most basic
SELECT statement contains the SELECT and FROM clauses. Items within
clauses are separated with commas in SQL, not with blanks as in other SAS code.
For example, if you list three columns in the SELECT clause, then the columns
are separated with commas.
3 The SELECT statement, which is used to retrieve data, also automatically writes
the output data to the Output window unless you specify the NOPRINT option in
the PROC SQL statement. Therefore, you can display your output or send it to a
list le without specifying the PRINT procedure.

The SQL Procedure

Syntax: SQL Procedure

1031

3 The ORDER BY clause sorts data by columns. In addition, tables do not need to
be presorted by a variable for use with PROC SQL. Therefore, you do not need to
use the SORT procedure with your PROC SQL programs.
3 A PROC SQL statement runs when you submit it; you do not have to specify a
RUN statement. If you follow a PROC SQL statement with a RUN statement,
then SAS ignores the RUN statement and submits the statements as usual.

Syntax: SQL Procedure

Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Name: SQL_Results
Reminder: You can use any global statements. See Chapter 2, Fundamental Concepts
for Using Base SAS Procedures, on page 15 for a list.
Reminder: You can use data set options any time a table name or view name is specied.
See Using SAS Data Set Options with PROC SQL on page 1114 for details.
Tip:

Note:

Regular type indicates the name of a component that is described in SQL Procedure
Component Dictionary on page 1071.
view-name indicates a SAS data view of any type.

PROC SQL <option(s)>;

ALTER TABLE table-name
<ADD <CONSTRAINT> constraint-clause< , constraint-clause>>
<ADD column-denition<, column-denition>>
<DROP CONSTRAINT constraint-name <, constraint-name>>
<DROP column< , column>>
<DROP FOREIGN KEY constraint-name>
<DROP PRIMARY KEY>
<MODIFY column-denition< , column-denition>>
;
CREATE <UNIQUE> INDEX index-name
ON table-name ( column <, column>);
CREATE TABLE table-name
(column-specication<, column-specication | constraint-specication>)
;
CREATE TABLE table-name LIKE table-name2;
CREATE TABLE table-name AS query-expression
<ORDER BY order-by-item< , order-by-item>>;
CREATE VIEW proc-sql-view AS query-expression
<ORDER BY order-by-item< , order-by-item>>
<USING libname-clause<, libname-clause>> ;
DELETE
FROM table-name|proc-sql-view |sas/access-view <AS alias>
<WHERE sql-expression>;
DESCRIBE TABLEtable-name <, table-name>;

1032

Syntax: SQL Procedure

Chapter 44

DESCRIBE VIEW proc-sql-view <, proc-sql-view>;

DESCRIBE TABLE CONSTRAINTS table-name <, table-name>;
DROP INDEX index-name <, index-name>
FROM table-name;
DROP TABLE table-name <, table-name>;
DROP VIEW view-name <, view-name>;
INSERT INTO table-name|sas/access-view|proc-sql-view <(column<, column>)>
SET column=sql-expression
<, column=sql-expression>
<SET column=sql-expression
<, column=sql-expression>>;
INSERT INTO table-name|sas/access-view|proc-sql-view < (column<, column>)>
VALUES (value <, value>)
< VALUES (value <, value>)>;
INSERT INTO table-name|sas/access-view|proc-sql-view
<(column< , column>)> query-expression;
RESET < option(s)>;
SELECT <DISTINCT> object-item <, object-item>
<INTO macro-variable-specication
< , macro-variable-specication>>
FROM from-list
<WHERE sql-expression>
<GROUP BY group-by-item
< , group-by-item>>
<HAVING sql-expression>
<ORDER BY order-by-item
< , order-by-item>>;
UPDATE table-name|sas/access-view|proc-sql-view <AS alias>
SET column=sql-expression
<, column=sql-expression>
<SET column=sql-expression
<, column=sql-expression>>
<WHERE sql-expression>;
VALIDATE query-expression;

To connect to a DBMS and send it a DBMS-specic nonquery SQL statement, use

this form:
PROC SQL;
CONNECT TO dbms-name < AS alias>
<(connect-statement-argument-1=value <
connect-statement-argument-n=value>)>
<(database-connection-argument-1=value <
database-connection-argument-n=value>)>;
EXECUTE (dbms-SQL-statement)
BY dbms-name|alias;
<DISCONNECT FROM dbms-name|alias;>
<QUIT;>
To connect to a DBMS and query the DBMS data, use this form:

The SQL Procedure

PROC SQL Statement

PROC SQL;
CONNECT TO dbms-name <AS alias>
< (connect-statement-argument-1=value <
connect-statement-argument-n=value>)>
< (database-connection-argument-1=value <
database-connection-argument-n=value>)>;
SELECT column-list
FROM CONNECTION TO dbms-name|alias
(dbms-query)
optional PROC SQL clauses;
<DISCONNECT FROM dbms-name|alias;>
<QUIT;>

To do this

Use this statement

Modify, add, or drop columns

ALTER TABLE

Establish a connection with a DBMS

CONNECT TO

Create an index on a column

CREATE INDEX

Create a PROC SQL table

CREATE TABLE

Create a PROC SQL view

CREATE VIEW

Delete rows

DELETE

Display a denition of a table or view

DESCRIBE

Terminate the connection with a DBMS

DISCONNECT FROM

Delete tables, views, or indexes

DROP

Send a DBMS-specic nonquery SQL statement to a

DBMS

EXECUTE

Add rows

INSERT

Reset options that affect the procedure environment

without restarting the procedure

RESET

Select and execute rows

SELECT

Query a DBMS

CONNECTION TO

Modify values

UPDATE

Verify the accuracy of your query

VALIDATE

PROC SQL Statement

PROC SQL <option(s)>;

1033

1034

PROC SQL Statement

Chapter 44

To do this

Use this option

Control output
Double-space the report

DOUBLE|NODOUBLE

Write a statement to the SAS log that

expands the query

FEEDBACK|NOFEEDBACK

Flow characters within a column

FLOW|NOFLOW

Include a column of row numbers

NUMBER|NONUMBER

Specify whether PROC SQL prints the

querys result

PRINT|NOPRINT

Specify whether PROC SQL should display

sorting information

SORTMSG|NOSORTMSG

Specify a collating sequence

SORTSEQ=

Control execution
Allow PROC SQL to use names other than
SAS names

DQUOTE=

Specify whether PROC SQL should stop

executing after an error

ERRORSTOP|NOERRORSTOP

Specify whether PROC SQL should execute

statements

EXEC|NOEXEC

Restrict the number of input rows

INOBS=

Restrict the number of output rows

OUTOBS=

Restrict the number of loops

LOOPS=

Specify whether PROC SQL prompts you

when a limit is reached with the INOBS=,
OUTOBS=, or LOOPS= options

PROMPT|NOPROMPT

Specify whether PROC SQL writes timing

information for each statement to the SAS log

STIMER|NOSTIMER

Override the SAS system option

THREADS|NOTHREADS

Specify how PROC SQL handles updates

when there is an interruption

UNDO_POLICY=

Options
DOUBLE|NODOUBLE

double-spaces the report.

Default: NODOUBLE
Featured in:

Example 5 on page 1134

DQUOTE=ANSI|SAS

species whether PROC SQL treats values within double quotation marks (" ") as
variables or strings. With DQUOTE=ANSI, PROC SQL treats a quoted value as a

The SQL Procedure

PROC SQL Statement

1035

variable. This feature enables you to use the following as table names, column
names, or aliases:

3 reserved words such as AS, JOIN, GROUP, and so on

3 DBMS names and other names that are not normally permissible in SAS.
The quoted value can contain any character.
With DQUOTE=SAS, values within double quotation marks are treated as strings.
Default: SAS
ERRORSTOP|NOERRORSTOP

species whether PROC SQL stops executing if it encounters an error. In a batch or

noninteractive session, ERRORSTOP instructs PROC SQL to stop executing the
statements but to continue checking the syntax after it has encountered an error.
NOERRORSTOP instructs PROC SQL to execute the statements and to continue
checking the syntax after an error occurs.
Default: NOERRORSTOP in an interactive SAS session; ERRORSTOP in a batch

or noninteractive session
Interaction: This option is useful only when the EXEC option is in effect.

ERRORSTOP has an effect only when SAS is running in the batch or

noninteractive execution mode.

Tip:

NOERRORSTOP is useful if you want a batch job to continue executing SQL

procedure statements after an error is encountered.

Tip:

EXEC|NOEXEC

species whether a statement should be executed after its syntax is checked for
accuracy.
Default: EXEC

NOEXEC is useful if you want to check the syntax of your SQL statements
without executing the statements.

Tip:

3 Any PROC SQL view is expanded into the underlying query.

3 Macro variables are resolved.
3 Parentheses are shown around all expressions to further indicate their order of
evaluation.

3 Comments are removed.

Default: NOFEEDBACK
FLOW<=n <m>>|NOFLOW

species that character columns longer than n are owed to multiple lines. PROC
SQL sets the column width at n and species that character columns longer than n
are owed to multiple lines. When you specify FLOW=n m, PROC SQL oats the
width of the columns between these limits to achieve a balanced layout. Specifying
FLOW without arguments is equivalent to specifying FLOW=12 200.
Default: NOFLOW

1036

PROC SQL Statement

Chapter 44

INOBS=n

restricts the number of rows (observations) that PROC SQL retrieves from any single
source.
Tip:

This option is useful for debugging queries on large tables.

LOOPS=n

restricts PROC SQL to n iterations through its inner loop. You use the number of
iterations reported in the SQLOOPS macro variable (after each SQL statement is
executed) to discover the number of loops. Set a limit to prevent queries from
consuming excessive computer resources. For example, joining three large tables
without meeting the join-matching conditions could create a huge internal table that
would be inefcient to execute.
See also: Using Macro Variables Set by PROC SQL on page 1119
NODOUBLE

See DOUBLE|NODOUBLE on page 1034.

NOERRORSTOP

See ERRORSTOP|NOERRORSTOP on page 1035.

NOEXEC

See EXEC|NOEXEC on page 1035.

NOFEEDBACK

See FEEDBACK|NOFEEDBACK on page 1035.

NOFLOW

See FLOW|NOFLOW on page 1035.

NONUMBER

See NUMBER|NONUMBER on page 1036.

NOPRINT

See PRINT|NOPRINT on page 1036.

NOPROMPT

See PROMPT|NOPROMPT on page 1037.

NOSORTMSG

See SORTMSG|NOSORTMSG on page 1037.

NOSTIMER

See STIMER|NOSTIMER on page 1037.

NOTHREADS

See THREADS|NOTHREADS.
NUMBER|NONUMBER

species whether the SELECT statement includes a column called ROW, which is the
row (or observation) number of the data as the rows are retrieved.
Default: NONUMBER
Featured in:

Example 4 on page 1131

OUTOBS=n

restricts the number of rows (observations) in the output. For example, if you specify
OUTOBS=10 and insert values into a table using a query-expression, then the SQL
procedure inserts a maximum of 10 rows. Likewise, OUTOBS=10 limits the output to
10 rows.
PRINT|NOPRINT

species whether the output from a SELECT statement is printed.

The SQL Procedure

PROC SQL Statement

1037

Default: PRINT

NOPRINT is useful when you are selecting values from a table into macro
variables and do not want anything to be displayed.

Tip:

Interaction: NOPRINT affects the value of the SQLOBS automatic macro variable.

See Using Macro Variables Set by PROC SQL on page 1119 for details.
PROMPT|NOPROMPT

modies the effect of the INOBS=, OUTOBS=, and LOOPS= options. If you specify
the PROMPT option and reach the limit specied by INOBS=, OUTOBS=, or
LOOPS=, then PROC SQL prompts you to stop or continue. The prompting repeats if
the same limit is reached again.
Default: NOPROMPT
SORTMSG|NOSORTMSG

Certain operations, such as ORDER BY, may sort tables internally using PROC
SORT. Specifying SORTMSG requests information from PROC SORT about the sort
and displays the information in the log.
Default: NOSORTMSG
SORTSEQ=sort-table

species the collating sequence to use when a query contains an ORDER BY clause.
Use this option only if you want a collating sequence other than your systems or
installations default collating sequence.
See also: SORTSEQ= option in SAS National Language Support (NLS): Users

Guide.
STIMER|NOSTIMER

species whether PROC SQL writes timing information to the SAS log for each
statement, rather than as a cumulative value for the entire procedure. For this
option to work, you must also specify the SAS system option STIMER. Some
operating environments require that you specify this system option when you invoke
SAS. If you use the system option alone, then you receive timing information for the
entire SQL procedure, not on a statement-by-statement basis.
Default: NOSTIMER
THREADS|NOTHREADS

overrides the SAS system option THREADS|NOTHREADS for a particular

invocation of PROC SQL. THREADS|NOTHREADS can also be specied in a
RESET statement for use in particular queries. When THREADS is specied, PROC
SQL uses parallel processing in order to increase the performance of sorting
operations that involve large amounts of data. For more information about parallel
processing, see SAS Language Reference: Concepts.
Default: value of SAS system option THREADS|NOTHREADS.

Note: When THREADS|NOTHREADS has been specied in a PROC SQL

statement or a RESET statement, there is no way to reset the option to its default
(that is, the value of the SAS system option THREADS|NOTHREADS) for that
invocation of PROC SQL. 4
UNDO_POLICY=NONE|OPTIONAL|REQUIRED

species how PROC SQL handles updated data if errors occur while you are
updating data. You can use UNDO_POLICY= to control whether your changes will
be permanent:
NONE
keeps any updates or inserts.

1038

ALTER TABLE Statement

Chapter 44

OPTIONAL
reverses any updates or inserts that it can reverse reliably.
REQUIRED
reverses all inserts or updates that have been done to the point of the error. In
some cases, the UNDO operation cannot be done reliably. For example, when a
program uses a SAS/ACCESS view, it may not be able to reverse the effects of the
INSERT and UPDATE statements without reversing the effects of other changes
at the same time. In that case, PROC SQL issues an error message and does not
execute the statement. Also, when a SAS data set is accessed through a
SAS/SHARE server and is opened with the data set option CNTLLEV=RECORD,
you cannot reliably reverse your changes.
This option may enable other users to update newly inserted rows. If an error
occurs during the insert, then PROC SQL can delete a record that another user
updated. In that case, the statement is not executed, and an error message is
issued.
Default: REQUIRED

Note: Options can be added, removed, or changed between PROC SQL statements
with the RESET statement. 4

ALTER TABLE Statement

Adds columns to, drops columns from, and changes column attributes in an existing table. Adds,
modies, and drops integrity constraints from an existing table.
Restriction:

You cannot use any type of view in an ALTER TABLE statement.

Restriction: You cannot use ALTER TABLE on a table that is accessed by an engine that
does not support UPDATE processing.
Restriction: You must use at least one ADD, DROP, or MODIFY clause in the ALTER
TABLE statement.
Featured in:

Example 3 on page 1129

ALTER TABLE table-name

<ADD CONSTRAINT constraint-name constraint-clause<, constraint-name
constraint-clause>>
<ADD constraint-specication< , constraint-specication>>
<ADD column-denition< , column-denition>>
<DROP CONSTRAINT constraint-name <, constraint-name>>
<DROP column<, column>>
<DROP FOREIGN KEY constraint-name>
<DROP PRIMARY KEY>
<MODIFY column-denition<, column-denition>>
;

The SQL Procedure

ALTER TABLE Statement

1039

Arguments
<ADD CONSTRAINT constraint-name constraint-specication< , constraint-name
constraint-specication>>
adds the integrity constraint that is specied in constraint-specication and assigns
constraint-name to it.
<ADD constraint-specication< , constraint-specication>>
adds the integrity constraint that is specied in constraint-specication and assigns a
default name to it. The default constraint name has the form that is shown in the
following table:
Default Name

Constraint Type

_NMxxxx_

Not null

_UNxxxx_

Unique

_CKxxxx_

Check

_PKxxxx_

Primary key

_FKxxxx_

Foreign key

In these default names, xxxx is a counter that begins at 0001.

<ADD column-denition< , column-denition>>
adds the column(s) that are specied in each column-denition.
column

names a column in table-name.

column-denition

See column-denition on page 1075.

constraint

is one of the following integrity constraints:

CHECK (WHERE-clause)
species that all rows in table-name satisfy the WHERE-clause.
DISTINCT (column<, column>)
species that the values of each column must be unique. This constraint is
identical to UNIQUE.
FOREIGN KEY (column< , column>)
REFERENCES table-name
<ON DELETE referential-action > <ON UPDATE referential-action>
species a foreign key, that is, a set of columns whose values are linked to the
values of the primary key variable in another table (the table-name that is specied
for REFERENCES). The referential-actions are performed when the values of a
primary key column that is referenced by the foreign key are updated or deleted.
Restriction: When dening overlapping primary key and foreign key constraints,

which means that variables in a data le are part of both a primary key and a
foreign key denition,

3 if you use the exact same variables, then the variables must be dened in
a different order.

1040

ALTER TABLE Statement

Chapter 44

3 the foreign keys update and delete referential actions must both be
RESTRICT.
NOT NULL (column)
species that column does not contain a null or missing value, including special
missing values.
PRIMARY KEY (column< , column>)
species one or more primary key columns, that is, columns that do not contain
missing values and whose values are unique.
Restriction: When you are dening overlapping primary key and foreign key

constraints, which means that variables in a data le are part of both a primary
key denition and a foreign key denition, if you use the exact same variables,
then the variables must be dened in a different order.
UNIQUE (column< , column>)
species that the values of each column must be unique. This constraint is
identical to DISTINCT.
constraint-name

species a name for the constraint that is being specied. The name must be a valid
SAS name.
Note: The names PRIMARY, FOREIGN, MESSAGE, UNIQUE, DISTINCT,
CHECK, and NOT cannot be used as values for constraint-name. 4
constraint-specication

consists of
constraint <MESSAGE=message-string <MSGTYPE=message-type>>
<DROP column< , column>>
deletes each column from the table.
<DROP CONSTRAINT constraint-name < , constraint-name>>
deletes the integrity constraint that is referenced by each constraint-name. To nd
the name of an integrity constraint, use the DESCRIBE TABLE CONSTRAINTS
clause (see DESCRIBE Statement on page 1052).
<DROP FOREIGN KEY constraint-name>
Removes the foreign key constraint that is referenced by constraint-name.
Note:

The DROP FOREIGN KEY clause is a DB2 extension.

<DROP PRIMARY KEY>

Removes the primary key constraint from table-name.
Note:

The DROP PRIMARY KEY clause is a DB2 extension.

message-string

species the text of an error message that is written to the log when the integrity
constraint is not met. The maximum length of message-string is 250 characters.
message-type

species how the error message is displayed in the SAS log when an integrity
constraint is not met.

The SQL Procedure

ALTER TABLE Statement

1041

NEWLINE
the text that is specied for MESSAGE= is displayed as well as the default error
message for that integrity constraint.
USER
only the text that is specied for MESSAGE= is displayed.
<MODIFY column-denition<, column-denition>>
changes one or more attributes of the column that is specied in each
column-denition.
referential-action

species the type of action to be performed on all matching foreign key values.
CASCADE
allows primary key data values to be updated, and updates matching values in the
foreign key to the same values. This referential action is currently supported for
updates only.
RESTRICT
prevents the update or deletion of primary key data values if a matching value
exists in the foreign key. This referential action is the default.
SET NULL
allows primary key data values to be updated, and sets all matching foreign key
values to NULL.
table-name

3 in the ALTER TABLE statement, refers to the name of the table that is to be
altered.

3 in the REFERENCES clause, refers to the name of table that contains the
primary key that is referenced by the foreign key.
table-name can be a one-level name, a two-level libref.table name, or a physical
pathname that is enclosed in single quotation marks.
WHERE-clause

species a SAS WHERE clause. Do not include the WHERE keyword in the WHERE
clause.

Specifying Initial Values of New Columns

When the ALTER TABLE statement adds a column to the table, it initializes the
columns values to missing in all rows of the table. Use the UPDATE statement to add
values to the new column(s).

Changing Column Attributes

If a column is already in the table, then you can change the following column
attributes by using the MODIFY clause: length, informat, format, and label. The values
in a table are either truncated or padded with blanks (if character data) as necessary to
meet the specied length attribute.
You cannot change a character column to numeric and vice versa. To change a
columns data type, drop the column and then add it (and its data) again, or use the
DATA step.
Note: You cannot change the length of a numeric column with the ALTER TABLE
statement. Use the DATA step instead. 4

1042

CONNECT Statement

Chapter 44

Renaming Columns
To change a columns name, you must use the SAS data set option RENAME=. You
cannot change this attribute with the ALTER TABLE statement. RENAME= is
described in the section on SAS data set options in SAS Language Reference: Dictionary.

Indexes on Altered Columns

When you alter the attributes of a column and an index has been dened for that
column, the values in the altered column continue to have the index dened for them. If
you drop a column with the ALTER TABLE statement, then all the indexes (simple and
composite) in which the column participates are also dropped. See CREATE INDEX
Statement on page 1043 for more information about creating and using indexes.

Integrity Constraints
Use ALTER TABLE to modify integrity constraints for existing tables. Use the
CREATE TABLE statement to attach integrity constraints to new tables. For more
information on integrity constraints, see the section on SAS les in SAS Language
Reference: Concepts.

CONNECT Statement
Establishes a connection with a DBMS that is supported by SAS/ACCESS software.
Requirement: SAS/ACCESS software is required. For more information about this
statement, refer to your SAS/ACCESS documentation.
See also: Connecting to a DBMS Using the SQL Procedure Pass-Through Facility on
page 1115

CONNECT TO dbms-name <AS alias>

<(connect-statement-argument-1=value <
connect-statement-argument-n=value>)>
<(database-connection-argument-1=value <
database-connection-argument-n=value>)>;

Arguments
alias

species an alias that has 1 to 32 characters. The keyword AS must precede alias.
Some DBMSs allow more than one connection. The optional AS clause enables you to
name the connections so that you can refer to them later.
connect-statement-argument=value

species values for arguments that indicate whether you can make multiple
connections, shared or unique connections, and so on, to the database. These
arguments are optional, but if they are included, then they must be enclosed in
parentheses. See SAS/ACCESS for Relational Databases: Reference for more
information about these arguments.

The SQL Procedure

CREATE INDEX Statement

1043

database-connection-argument=value

species values for the DBMS-specic arguments that are needed by PROC SQL in
order to connect to the DBMS. These arguments are optional for most databases, but
if they are included, then they must be enclosed in parentheses. For more
information, see the SAS/ACCESS documentation for your DBMS.
dbms-name

identies the DBMS that you want to connect to (for example, ORACLE or DB2).

CREATE INDEX Statement

Creates indexes on columns in tables.
Restriction: You cannot use CREATE INDEX on a table that is accessed with an engine
that does not support UPDATE processing.

CREATE <UNIQUE> INDEX index-name

ON table-name ( column <, column>);

Arguments
column

species a column in table-name.

index-name

names the index that you are creating. If you are creating an index on one column
only, then index-name must be the same as column. If you are creating an index on
more than one column, then index-name cannot be the same as any column in the
table.
table-name

species a PROC SQL table.

1044

CREATE INDEX Statement

Chapter 44

Indexes in PROC SQL

An index stores both the values of a tables columns and a system of directions that
enable access to rows in that table by index value. Dening an index on a column or set
of columns enables SAS, under certain circumstances, to locate rows in a table more
quickly and efciently. Indexes enable PROC SQL to execute the following classes of
queries more efciently:

3 comparisons against a column that is indexed

3 an IN subquery where the column in the inner subquery is indexed
3 correlated subqueries, where the column being compared with the correlated
reference is indexed

3 join-queries, where the join-expression is an equals comparison and all the

columns in the join-expression are indexed in one of the tables being joined.
SAS maintains indexes for all changes to the table, whether the changes originate
from PROC SQL or from some other source. Therefore, if you alter a columns denition
or update its values, then the same index continues to be dened for it. However, if an
indexed column in a table is dropped, then the index on it is also dropped.
You can create simple or composite indexes. A simple index is created on one column
in a table. A simple index must have the same name as that column. A composite index
is one index name that is dened for two or more columns. The columns can be
specied in any order, and they can have different data types. A composite index name
cannot match the name of any column in the table. If you drop a composite index, then
the index is dropped for all the columns named in that composite index.

UNIQUE Keyword
The UNIQUE keyword causes SAS to reject any change to a table that would cause
more than one row to have the same index value. Unique indexes guarantee that data
in one column, or in a composite group of columns, remain unique for every row in a
table. For this reason, a unique index cannot be dened for a column that includes
NULL or missing values.

The SQL Procedure

CREATE TABLE Statement

1045

Managing Indexes
You can use the CONTENTS statement in the DATASETS procedure to display a
tables index names and the columns for which they are dened. You can also use the
DICTIONARY tables INDEXES, TABLES, and COLUMNS to list information about
indexes. For more information, see Using the DICTIONARY Tables on page 1116.
See the section on SAS les in SAS Language Reference: Dictionary for a further
description of when to use indexes and how they affect SAS statements that handle
BY-group processing.

CREATE TABLE Statement

Creates PROC SQL tables.
Featured in: Example 1 on page 1125 and Example 2 on page 1127

u CREATE TABLE table-name

(column-specication<, column-specication | constraint-specication>)
;
v CREATE TABLE table-name LIKE table-name2;
w CREATE TABLE table-name AS query-expression
<ORDER BY order-by-item<, order-by-item>>;

Arguments
column-constraint

is one of the following:

CHECK (WHERE-clause)
species that all rows in table-name satisfy the WHERE-clause.
DISTINCT
species that the values of the column must be unique. This constraint is identical
to UNIQUE.

1046

CREATE TABLE Statement

Chapter 44

NOT NULL
species that the column does not contain a null or missing value, including
special missing values.
PRIMARY KEY
species that the column is a primary key column, that is, a column that does not
contain missing values and whose values are unique.
Restriction: When dening overlapping primary key and foreign key constraints,

which means that variables in a data le are part of both a primary key and a
foreign key denition, if you use the exact same variables, then the variables
must be dened in a different order.
REFERENCES table-name
<ON DELETE referential-action > <ON UPDATE referential-action>
species that the column is a foreign key, that is, a column whose values are
linked to the values of the primary key variable in another table (the table-name
that is specied for REFERENCES). The referential-actions are performed when
the values of a primary key column that is referenced by the foreign key are
updated or deleted.
Restriction: When you are dening overlapping primary key and foreign key

constraints, which means that variables in a data le are part of both a primary
key denition and a foreign key denition,

3 if you use the exact same variables, then the variables must be dened in
a different order

3 the foreign keys update and delete referential actions must both be
RESTRICT.
UNIQUE
species that the values of the column must be unique. This constraint is identical
to DISTINCT.
Note: If you specify column-constraint, then SAS automatically assigns a name to
the constraint. The constraint name has the form
Default name

Constraint type

_CKxxxx_

Check

_FKxxxx_

Foreign key

_NMxxxx_

Not Null

_PKxxxx_

Primary key

_UNxxxx_

Unique

where xxxx is a counter that begins at 0001.

column-denition

See column-denition on page 1075.

column-specication

consists of
column-denition <column-constraint>
constraint

is one of the following:

The SQL Procedure

CREATE TABLE Statement

1047

CHECK (WHERE-clause)
species that all rows in table-name satisfy the WHERE-clause.
DISTINCT (column<, column>)
species that the values of each column must be unique. This constraint is
identical to UNIQUE.
FOREIGN KEY (column< , column>)
REFERENCES table-name
<ON DELETE referential-action > <ON UPDATE referential-action>
species a foreign key, that is, a set of columns whose values are linked to the
values of the primary key variable in another table (the table-name that is specied
for REFERENCES). The referential-actions are performed when the values of a
primary key column that is referenced by the foreign key are updated or deleted.
Restriction: When you are dening overlapping primary key and foreign key

constraints, which means that variables in a data le are part of both a primary
key denition and a foreign key denition,

3 if you use the exact same variables, then the variables must be dened in
a different order

3 the foreign keys update and delete referential actions must both be
RESTRICT.
NOT NULL (column)
species that column does not contain a null or missing value, including special
missing values.
PRIMARY KEY (column< , column>)
species one or more primary key columns, that is, columns that do not contain
missing values and whose values are unique.
Restriction: When dening overlapping primary key and foreign ke constraints,

which means that variables in a data le are part of both a primary key and a
foreign key denition, if you use the exact same variables, then the variables
must be dened in a different order.
UNIQUE (column< , column>)
species that the values of each column must be unique. This constraint is
identical to DISTINCT.
constraint-name

consists of
CONSTRAINT constraint-name constraint <MESSAGE=message-string
< MSGTYPE=message-type>>
message-string

species the text of an error message that is written to the log when the integrity
constraint is not met. The maximum length of message-string is 250 characters.
message-type

species how the error message is displayed in the SAS log when an integrity
constraint is not met.

1048

CREATE TABLE Statement

Chapter 44

NEWLINE
the text that is specied for MESSAGE= is displayed as well as the default error
message for that integrity constraint.
USER
only the text that is specied for MESSAGE= is displayed.
ORDER BY order-by-item
sorts the rows in table-name by the values of each order-by-item. See ORDER BY
Clause on page 1067.
query-expression

creates table-name from the results of a query. See query-expression on page 1093.
referential-action

species the type of action to be performed on all matching foreign key values.
CASCADE
allows primary key data values to be updated, and updates matching values in the
foreign key to the same values. This referential action is currently supported for
updates only.
RESTRICT
occurs only if there are matching foreign key values. This referential action is the
default.
SET NULL
sets all matching foreign key values to NULL.
table-name

3 in the CREATE TABLE statement, refers to the name of the table that is to be
created. You can use data set options by placing them in parentheses
immediately after table-name. See Using SAS Data Set Options with PROC
SQL on page 1114 for details.

3 in the REFERENCES clause, refers to the name of table that contains the
primary key that is referenced by the foreign key.
table-name2

creates table-name with the same column names and column attributes as
table-name2, but with no rows.
WHERE-clause

species a SAS WHERE clause. Do not include the WHERE keyword in the WHERE
clause.

Creating a Table without Rows

u The rst form of the CREATE TABLE statement creates tables that automatically
map SQL data types to those that are supported by SAS. Use this form when you
want to create a new table with columns that are not present in existing tables. It
is also useful if you are running SQL statements from an SQL application in
another SQL-based database.
v The second form uses a LIKE clause to create a table that has the same column
names and column attributes as another table. To drop any columns in the new
table, you can specify the DROP= data set option in the CREATE TABLE
statement. The specied columns are dropped when the table is created. Indexes
are not copied to the new table.

The SQL Procedure

CREATE VIEW Statement

1049

Both of these forms create a table without rows. You can use an INSERT
statement to add rows. Use an ALTER TABLE statement to modify column
attributes or to add or drop columns.

Creating a Table from a Query Expression

w The third form of the CREATE TABLE statement stores the results of any
query-expression in a table and does not display the output. It is a convenient way
to create temporary tables that are subsets or supersets of other tables.
When you use this form, a table is physically created as the statement is
executed. The newly created table does not reect subsequent changes in the
underlying tables (in the query-expression). If you want to continually access the
most current data, then create a view from the query expression instead of a table.
See CREATE VIEW Statement on page 1049.
CAUTION:

Recursive table references can cause data integrity problems. While it is possible to
recursively reference the target table of a CREATE TABLE AS statement, doing
so can cause data integrity problems and incorrect results. Constructions such
as the following should be avoided:
proc sql;
create table a as
select var1, var2
from a;

Integrity Constraints
You can attach integrity constraints when you create a new table. To modify integrity
constraints, use the ALTER TABLE statement. For more information on integrity
constraints, see the section on SAS les in SAS Language Reference: Concepts.

CREATE VIEW Statement

Creates a PROC SQL view from a query-expression.
See also: What Are Views? on page 1029
Featured in: Example 8 on page 1143

CREATE VIEW proc-sql-view <(column-name-list)> AS query-expression

<ORDER BY order-by-item<, order-by-item>>
<USING libname-clause<, libname-clause>> ;

1050

CREATE VIEW Statement

Chapter 44

Arguments
column-name-list

is a comma-separated list of column names for the view, to be used in place of the
column names or aliases that are specied in the SELECT clause. The names in this
list are assigned to columns in the order in which they are specied in the SELECT
clause. If the number of column names in this list does not equal the number of
columns in the SELECT clause, then a warning is written to the SAS log.
query-expression

See query-expression on page 1093.

libname-clause

is one of the following:

LIBNAME libref < engine> SAS-data-library <option(s)> <engine-host-option(s)>
LIBNAME libref SAS/ACCESS-engine-name
<SAS/ACCESS-engine-connection-option(s)>
<SAS/ACCESS-engine-LIBNAME-option(s)>
See SAS Language Reference: Dictionary for information about the base SAS
LIBNAME statement. See SAS/ACCESS for Relational Databases: Reference for
information about the LIBNAME statement for relational databases.
order-by-item

See ORDER BY Clause on page 1067.

proc-sql-view

species the name for the PROC SQL view that you are creating. See What Are
Views? on page 1029 for a denition of a PROC SQL view.

Sorting Data Retrieved by Views

PROC SQL enables you to specify the ORDER BY clause in the CREATE VIEW
statement. When a view with an ORDER BY clause is accessed, and the ORDER BY
clause directly affects the order of the results, its data is sorted and displayed as
specied by the ORDER BY clause. However, if the ORDER BY clause does not directly
affect the order of the results (for instance, if the view is specied as part of a join),
then PROC SQL ignores the ORDER BY clause in order to enhance performance.
Note: If you specify the NUMBER option in the PROC SQL statement when you
create your view, then the ROW column appears in the output. However, you cannot
order by the ROW column in subsequent queries. See the description of
NUMBER|NONUMBER on page 1036. 4

Librefs and Stored Views

You can refer to a table name alone (without the libref) in the FROM clause of a
CREATE VIEW statement if the table and view reside in the same SAS data library, as
in this example:
create view proclib.view1 as
select *
from invoice
where invqty>10;

In this view, VIEW1 and INVOICE are stored permanently in the SAS data library
referenced by PROCLIB. Specifying a libref for INVOICE is optional.

The SQL Procedure

DELETE Statement

1051

Updating Views
You can update a views underlying data with some restrictions. See Updating
PROC SQL and SAS/ACCESS Views on page 1121.

Embedded LIBNAME Statements

The USING clause enables you to store DBMS connection information in a view by
embedding the SAS/ACCESS LIBNAME statement inside the view. When PROC SQL
executes the view, the stored query assigns the libref and establishes the DBMS
connection using the information in the LIBNAME statement. The scope of the libref is
local to the view, and will not conict with any identically named librefs in the SAS
session. When the query nishes, the connection to the DBMS is terminated and the
libref is deassigned.
The USING clause must be the last clause in the CREATE VIEW statement. Multiple
LIBNAME statements can be specied, separated by commas. In the following example,
a connection is made and the libref ACCREC is assigned to an ORACLE database.
create view proclib.view1 as
select *
from accrec.invoices as invoices
using libname accrec oracle
user=username pass=password
path=dbms-path;

For more information on the SAS/ACCESS LIBNAME statement, see the SAS/ACCESS
documentation for your DBMS.
Note: Starting in SAS System 9, PROC SQL views, the Pass-Through Facility, and
the SAS/ACCESS LIBNAME statement are the preferred ways to access relational
DBMS data; SAS/ACCESS views are no longer recommended. You can convert existing
SAS/ACCESS views to PROC SQL views by using the CV2VIEW procedure. See The
CV2VIEW Procedure in SAS/ACCESS for Relational Databases: Reference for more
information. 4
You can also embed a SAS LIBNAME statement in a view with the USING clause.
This enables you to store SAS libref information in the view. Just as in the embedded
SAS/ACCESS LIBNAME statement, the scope of the libref is local to the view, and it
will not conict with an identically named libref in the SAS session.
create view work.tableview as
select * from proclib.invoices
using libname proclib sas-data-library;

DELETE Statement
Removes one or more rows from a table or view that is specied in the FROM clause.
Restriction: You cannot use DELETE FROM on a table that is accessed by an engine
that does not support UPDATE processing.
Featured in: Example 5 on page 1134

1052

DESCRIBE Statement

Chapter 44

DELETE
FROM table-name|sas/access-view|proc-sql-view <AS alias>
<WHERE sql-expression>;

Arguments
alias

assigns an alias to table-name, sas/access-view, or proc-sql-view.

sas/access-view

species a SAS/ACCESS view that you are deleting rows from.

proc-sql-view

species a PROC SQL view that you are deleting rows from. proc-sql-view can be a
one-level name, a two-level libref.view name, or a physical pathname that is enclosed
in single quotation marks.
sql-expression

See sql-expression on page 1099.

table-name

species the table that you are deleting rows from. table-name can be a one-level
name, a two-level libref.table name, or a physical pathname that is enclosed in single
quotation marks.
CAUTION:

Recursive table references can cause data integrity problems. While it is possible to
recursively reference the target table of a DELETE statement, doing so can cause
data integrity problems and incorrect results. Constructions such as the following
should be avoided:
proc sql;
delete from a
where var1 > (select min(var2) from a);

Deleting Rows through Views

You can delete one or more rows from a views underlying table, with some
restrictions. See Updating PROC SQL and SAS/ACCESS Views on page 1121.
CAUTION:

If you omit a WHERE clause, then the DELETE statement deletes all the rows from the
specied table or the table that is described by a view. 4

DESCRIBE Statement
Displays a PROC SQL denition in the SAS log.
Restriction:

PROC SQL views are the only type of view allowed in a DESCRIBE VIEW

statement.
Featured in:

Example 6 on page 1136

The SQL Procedure

DISCONNECT Statement

1053

DESCRIBE TABLE table-name <, table-name>;

DESCRIBE VIEW proc-sql-view <, proc-sql-view>;
DESCRIBE TABLE CONSTRAINTS table-name <, table-name>;

Arguments
table-name

species a PROC SQL table. table-name can be a one-level name, a two-level

libref.table name, or a physical pathname that is enclosed in single quotation marks.
proc-sql-view

species a PROC SQL view. proc-sql-view can be a one-level name, a two-level

libref.view name, or a physical pathname that is enclosed in single quotation marks.

Details
3 The DESCRIBE TABLE statement writes a CREATE TABLE statement to the
SAS log for the table specied in the DESCRIBE TABLE statement, regardless of
how the table was originally created (for example, with a DATA step). If
applicable, SAS data set options are included with the table denition. If indexes
are dened on columns in the table, then CREATE INDEX statements for those
indexes are also written to the SAS log.
When you are transferring a table to a DBMS that is supported by
SAS/ACCESS software, it is helpful to know how it is dened. To nd out more
information about a table, use the FEEDBACK option or the CONTENTS
statement in the DATASETS procedure.

3 The DESCRIBE VIEW statement writes a view denition to the SAS log. If you
use a PROC SQL view in the DESCRIBE VIEW statement that is based on or
derived from another view, then you might want to use the FEEDBACK option in
the PROC SQL statement. This option displays in the SAS log how the underlying
view is dened and expands any expressions that are used in this view denition.
The CONTENTS statement in DATASETS procedure can also be used with a view
to nd out more information.

3 The DESCRIBE TABLE CONSTRAINTS statement lists the integrity constraints

that are dened for the specied table(s).

DISCONNECT Statement
Ends the connection with a DBMS that is supported by a SAS/ACCESS interface.
Requirement: SAS/ACCESS software is required. For more information on this
statement, refer to your SAS/ACCESS documentation.
See also: Connecting to a DBMS Using the SQL Procedure Pass-Through Facility on
page 1115

DISCONNECT FROM dbms-name|alias;

1054

DROP Statement

Chapter 44

Arguments

alias

species the alias that is dened in the CONNECT statement.

dbms-name

species the DBMS from which you want to end the connection (for example, DB2 or
ORACLE). The name you specify should match the name that is specied in the
CONNECT statement.

Details
3 An implicit COMMIT is performed before the DISCONNECT statement ends the
DBMS connection. If a DISCONNECT statement is not submitted, then implicit
DISCONNECT and COMMIT actions are performed and the connection to the
DBMS is broken when PROC SQL terminates.

3 PROC SQL continues executing until you submit a QUIT statement, another SAS
procedure, or a DATA step.

DROP Statement
Deletes tables, views, or indexes.
Restriction: You cannot use DROP TABLE or DROP INDEX on a table that is accessed
by an engine that does not support UPDATE processing.

DROP TABLE table-name <, table-name>;

DROP VIEW view-name < , view-name>;
DROP INDEX index-name <, index-name>
FROM table-name;

Arguments

index-name

species an index that exists on table-name.

table-name

species a PROC SQL table. table-name can be a one-level name, a two-level

libref.table name, or a physical pathname that is enclosed in single quotation marks.
view-name

species a SAS data view of any type: PROC SQL view, SAS/ACCESS view, or DATA
step view. view-name can be a one-level name, a two-level libref.view name, or a
physical pathname that is enclosed in single quotation marks.

The SQL Procedure

EXECUTE Statement

1055

Details
3 If you drop a table that is referenced in a view denition and try to execute the
view, then an error message is written to the SAS log that states that the table
does not exist. Therefore, remove references in queries and views to any table(s)
and view(s) that you drop.

3 If you drop a table with indexed columns, then all the indexes are automatically
dropped. If you drop a composite index, then the index is dropped for all the
columns that are named in that index.

3 You can use the DROP statement to drop a table or view in an external database
that is accessed with the Pass-Through Facility or SAS/ACCESS LIBNAME
statement, but not for an external database table or view that is described by a
SAS/ACCESS view.

EXECUTE Statement
Sends a DBMS-specic SQL statement to a DBMS that is supported by a SAS/ACCESS interface.
Requirement: SAS/ACCESS software is required. For more information on this
statement, refer to your SAS/ACCESS documentation.
See also: Connecting to a DBMS Using the SQL Procedure Pass-Through Facility on
page 1115 and the SQL documentation for your DBMS.

EXECUTE (dbms-SQL-statement)
BY dbms-name|alias;

Arguments
alias

species an optional alias that is dened in the CONNECT statement. Note that
alias must be preceded by the keyword BY.
dbms-name

identies the DBMS to which you want to direct the DBMS statement (for example,
ORACLE or DB2).
dbms-SQL-statement

is any DBMS-specic SQL statement, except the SELECT statement, that can be
executed by the DBMS-specic dynamic SQL.

Details
3 If your DBMS supports multiple connections, then you can use the alias that is
dened in the CONNECT statement. This alias directs the EXECUTE statements
to a specic DBMS connection.

3 Any return code or message that is generated by the DBMS is available in the
macro variables SQLXRC and SQLXMSG after the statement completes.

1056

INSERT Statement

Chapter 44

INSERT Statement
Adds rows to a new or existing table or view.
Restriction: You cannot use INSERT INTO on a table that is accessed with an engine
that does not support UPDATE processing.
Featured in:

Example 1 on page 1125

u INSERT INTO table-name|sas/access-view|proc-sql-view <(column< , column>)>

SET column=sql-expression
<, column=sql-expression>
<SET column=sql-expression
<, column=sql-expression>>;
v INSERT INTO table-name|sas/access-view|proc-sql-view <(column<, column>)>
VALUES (value <, value>)
< VALUES (value <, value>)>;
w INSERT INTO table-name|sas/access-view|proc-sql-view
<(column< , column>)> query-expression;

Arguments
column

species the column into which you are inserting rows.

proc-sql-view

species a PROC SQL view into which you are inserting rows. proc-sql-view can be a
one-level name, a two-level libref.view name, or a physical pathname that is enclosed
in single quotation marks.
query-expression

See query-expression on page 1093.

sas/access-view

species a SAS/ACCESS view into which you are inserting rows.

sql-expression

See sql-expression on page 1099.

table-name

species a PROC SQL table into which you are inserting rows. table-name can be a
one-level name, a two-level libref.table name, or a physical pathname that is enclosed
in single quotation marks.
value

is a data value.
CAUTION:

Recursive table references can cause data integrity problems. While it is possible to
recursively reference the target table of an INSERT statement, doing so can cause
data integrity problems and incorrect results. Constructions such as the following
should be avoided:

The SQL Procedure

INSERT Statement

1057

proc sql;
insert into a
select var1, var2
from a
where var1 > 0;

Methods for Inserting Values

u The rst form of the INSERT statement uses the SET clause, which species or
alters the values of a column. You can use more than one SET clause per INSERT
statement, and each SET clause can set the values in more than one column.
Multiple SET clauses are not separated by commas. If you specify an optional list
of columns, then you can set a value only for a column that is specied in the list
of columns to be inserted.
v The second form of the INSERT statement uses the VALUES clause. This clause
can be used to insert lists of values into a table. You can either give a value for
each column in the table or give values just for the columns specied in the list of
column names. One row is inserted for each VALUES clause. Multiple VALUES
clauses are not separated by commas. The order of the values in the VALUES
clause matches the order of the column names in the INSERT column list or, if no
list was specied, the order of the columns in the table.
w The third form of the INSERT statement inserts the results of a query-expression
into a table. The order of the values in the query-expression matches the order of
the column names in the INSERT column list or, if no list was specied, the order
of the columns in the table.

Note: If the INSERT statement includes an optional list of column names, then only
those columns are given values by the statement. Columns that are in the table but not
listed are given missing values. 4

Inserting Rows through Views

You can insert one or more rows into a table through a view, with some restrictions.
See Updating PROC SQL and SAS/ACCESS Views on page 1121.

Adding Values to an Indexed Column

If an index is dened on a column and you insert a new row into the table, then that
value is added to the index. You can display information about indexes with
3 the CONTENTS statement in the DATASETS procedure. See CONTENTS
Statement on page 323.

3 the DICTIONARY.INDEXES table. See Using the DICTIONARY Tables on page

1116 for more information.
For more information on creating and using indexes, see CREATE INDEX
Statement on page 1043.

1058

RESET Statement

Chapter 44

RESET Statement
Resets PROC SQL options without restarting the procedure.
Featured in:

Example 5 on page 1134

RESET <option(s)>;
The RESET statement enables you to add, drop, or change the options in PROC SQL
without restarting the procedure. See PROC SQL Statement on page 1033 for a
description of the options.

SELECT Statement
Selects columns and rows of data from tables and views.
Restriction:

The clauses in the SELECT statement must appear in the order shown.

SELECT <DISTINCT> object-item <, object-item>

<INTO macro-variable-specication
<, macro-variable-specication>>
FROM from-list
<WHERE sql-expression>
<GROUP BY group-by-item
<, group-by-item>>
<HAVING sql-expression>
<ORDER BY order-by-item
<, order-by-item>>;

SELECT Clause
Lists the columns that will appear in the output.
See Also: column-denition on page 1075
Featured in:

Example 1 on page 1125 and Example 2 on page 1127

SELECT <DISTINCT> object-item <, object-item>

The SQL Procedure

SELECT Clause

1059

Arguments
alias

assigns a temporary, alternate name to the column.

DISTINCT

eliminates duplicate rows.

Featured in: Example 13 on page 1154
object-item

is one of the following:

*
represents all columns in the tables or views that are listed in the FROM clause.
case-expression <AS alias>
derives a column from a CASE expression. See CASE expression on page 1073.
column-name <<AS> alias>
<column-modier < column-modier>>
names a single column. See column-name on page 1078 and column-modier
on page 1076.
sql-expression <AS alias>
<column-modier < column-modier>>
derives a column from an sql-expression. See sql-expression on page 1099 and
column-modier on page 1076.
table-name.*
species all columns in the PROC SQL table that is specied in table-name.
table-alias.*
species all columns in the PROC SQL table that has the alias that is specied in
table-alias.
view-name.*
species all columns in the SAS data view that is specied in view-name.
view-alias.*
species all columns in the SAS data view that has the alias that is specied in
view-alias.

Asterisk (*) Notation

The asterisk (*) represents all columns of the table(s) listed in the FROM clause.
When an asterisk is not prexed with a table name, all the columns from all tables in
the FROM clause are included; when it is prexed (for example, table-name.* or
table-alias.*), all the columns from that table only are included.

Column Aliases
A column alias is a temporary, alternate name for a column. Aliases are specied in
the SELECT clause to name or rename columns so that the result table is clearer or
easier to read. Aliases are often used to name a column that is the result of an
arithmetic expression or summary function. An alias is one word only. If you need a
longer column name, then use the LABEL= column-modier, as described in
column-modier on page 1076. The keyword AS is not required with a column alias.
Column aliases are optional, and each column name in the SELECT clause can have
an alias. After you assign an alias to a column, you can use the alias to refer to that
column in other clauses.

1060

INTO Clause

Chapter 44

If you use a column alias when creating a PROC SQL view, then the alias becomes
the permanent name of the column for each execution of the view.

INTO Clause
Stores the value of one or more columns for use later in another PROC SQL query or SAS
statement.
An INTO clause cannot be used in a CREATE TABLE statement.
See also: Using Macro Variables Set by PROC SQL on page 1119
Restriction:

INTO macro-variable-specication
<, macro-variable-specication>

Arguments
macro-variable

species a SAS macro variable that stores the values of the rows that are returned.
macro-variable-specication

is one of the following:

:macro-variable <SEPARATED BY character(s) < NOTRIM>>
stores the values that are returned into a single macro variable.
:macro-variable-1 :macro-variable-n <NOTRIM>
stores the values that are returned into a range of macro variables.
NOTRIM

protects the leading and trailing blanks from being deleted from values that are
stored in a range of macro variables or multiple values that are stored in a single
macro variable.
SEPARATED BY character

species a character that separates the values of the rows.

Details
3 Use the INTO clause only in the outer query of a SELECT statement and not in a
subquery.

3 When storing a single value into a macro variable, PROC SQL preserves leading
or trailing blanks. However, when storing values into a range of macro variables,
or when using the SEPARATED BY option to store multiple values in one macro
variable, PROC SQL trims leading or trailing blanks unless you use the NOTRIM
option.

3 You can put multiple rows of the output into macro variables. You can check the
PROC SQL macro variable SQLOBS to see the number of rows that are produced
by a query-expression. See Using Macro Variables Set by PROC SQL on page
1119 for more information on SQLOBS.

The SQL Procedure

INTO Clause

1061

Examples
These examples use the PROCLIB.HOUSES table:

The SAS System

Style
SqFeet
-----------------CONDO
900
CONDO
1000
RANCH
1200
RANCH
1400
SPLIT
1600
SPLIT
1800
TWOSTORY
2100
TWOSTORY
3000
TWOSTORY
1940
TWOSTORY
1860

With the macro-variable-specication, you can do the following:

3 You can create macro variables based on the rst row of the result.
proc sql noprint;
select style, sqfeet
into :style, :sqfeet
from proclib.houses;
%put &style &sqfeet;

The results are written to the SAS log:

1
proc sql noprint;
2
select style, sqfeet
3
into :style, :sqfeet
4
from proclib.houses;
5
6
%put &style &sqfeet;
CONDO
900

3 You can create one new macro variable per row in the result of the SELECT
statement. This example shows how you can request more values for one column
than for another. The hyphen (-) is used in the INTO clause to imply a range of
macro variables. You can use either of the keywords THROUGH or THRU instead
of a hyphen.
The following PROC SQL step puts the values from the rst four rows of the
PROCLIB.HOUSES table into macro variables:
proc sql noprint;
select distinct Style, SqFeet
into :style1 - :style3, :sqfeet1 - :sqfeet4
from proclib.houses;
%put
%put
%put
%put

&style1 &sqfeet1;
&style2 &sqfeet2;
&style3 &sqfeet3;
&sqfeet4;

1062

INTO Clause

Chapter 44

The %PUT statements write the results to the SAS log:

1
proc sql noprint;
2
select distinct style, sqfeet
3
into :style1 - :style3, :sqfeet1 - :sqfeet4
4
from proclib.houses;
5
6
%put &style1 &sqfeet1;
CONDO 900
7
%put &style2 &sqfeet2;
CONDO 1000
8
%put &style3 &sqfeet3;
RANCH 1200
9
%put &sqfeet4;
1400

3 You can concatenate the values of one column into one macro variable. This form
is useful for building up a list of variables or constants.
proc sql noprint;
select distinct style
into :s1 separated by ,
from proclib.houses;
%put &s1;

The results are written to the SAS log:

3
4
5
6
7
8

proc sql noprint;

select distinct style
into :s1 separated by ,
from proclib.houses;
%put &s1

CONDO,RANCH,SPLIT,TWOSTORY

3 You can use leading zeros in order to create a range of macro variable names, as
shown in the following example:
proc sql noprint;
select SqFeet
into :sqfeet01 - :sqfeet10
from proclib.houses;
%put &sqfeet01 &sqfeet02 &sqfeet03 &sqfeet04 &sqfeet05;
%put &sqfeet06 &sqfeet07 &sqfeet08 &sqfeet09 &sqfeet10;

The results are written to the SAS log:

11
12
13
14

proc sql noprint;

select sqfeet
into :sqfeet01 - :sqfeet10
from proclib.houses;

15
%put &sqfeet01 &sqfeet02 &sqfeet03 &sqfeet04 &sqfeet05;
900 1000 1200 1400 1600
16
%put &sqfeet06 &sqfeet07 &sqfeet08 &sqfeet09 &sqfeet10;
1800 2100 3000 1940 1860

The SQL Procedure

FROM Clause

1063

3 You can prevent leading and trailing blanks from being trimmed from values that
are stored in macro variables. By default, when storing values in a range of macro
variables or when storing multiple values in one macro variable (with the
SEPARATED BY option), PROC SQL trims the leading and trailing blanks from
the values before creating the macro variables. If you do not want the blanks to be
trimmed, then add the NOTRIM option, as shown in the following example:
proc sql noprint;
select style, sqfeet
into :style1 - :style4 notrim,
:sqfeet separated by , notrim
from proclib.houses;
%put
%put
%put
%put

*&style1*
*&style2*
*&style3*
*&style4*

*&sqfeet*;
*&sqfeet*;
*&sqfeet*;
*&sqfeet*;

The results are written to the SAS log, as shown in the following output:
3
proc sql noprint;
4
select style, sqfeet
5
into :style1 - :style4 notrim,
6
:sqfeet separated by , notrim
7
from proclib.houses;
8
9
%put *&style1* *&sqfeet*;
*CONDO
* *
900,
1000,
1200,
1400,
3000,
1940,
1860*
10 %put *&style2* *&sqfeet*;
*CONDO
* *
900,
1000,
1200,
1400,
3000,
1940,
1860**
11 %put *&style3* *&sqfeet*;
*RANCH
* *
900,
1000,
1200,
1400,
3000,
1940,
1860**
12 %put *&style4* *&sqfeet*;
*RANCH
* *
900,
1000,
1200,
1400,
3000,
1940,
1860**

1600,

1800,

2100,

1600,

1800,

2100,

1600,

1800,

2100,

1600,

1800,

2100,

FROM Clause
Species source tables or views.
Featured in: Example 1 on page 1125, Example 4 on page 1131, Example 9 on page 1145,
and Example 10 on page 1148

FROM from-list

Arguments
alias

species a temporary, alternate name for a table, view, or in-line view that is
specied in the FROM clause.

1064

FROM Clause

Chapter 44

column

names the column that appears in the output. The column names that you specify
are matched by position to the columns in the output.
from-list

is one of the following:

table-name <<AS> alias>
names a single PROC SQL table. table-name can be a one-level name, a two-level
libref.table name, or a physical pathname that is enclosed in single quotation
marks.
view-name <<AS> alias>
names a single SAS data view. view-name can be a one-level name, a two-level
libref.view name, or a physical pathname that is enclosed in single quotation
marks.
joined-table
species a join. See joined-table on page 1082.
(query-expression) <AS alias>
<(column <, column>)>
species an in-line view. See query-expression on page 1093.
CONNECTION TO
species a DBMS table. See CONNECTION TO on page 1079.
Note: With table-name and view-name, you can use data set options by placing
them in parentheses immediately after table-name or view-name. See Using SAS
Data Set Options with PROC SQL on page 1114 for details. 4

Table Aliases
A table alias is a temporary, alternate name for a table that is specied in the FROM
clause. Table aliases are prexed to column names to distinguish between columns that
are common to multiple tables. Column names in reexive joins (joining a table with
itself) must be prexed with a table alias in order to distinguish which copy of the table
the column comes from. Column names in other kinds of joins must be prexed with
table aliases or table names unless the column names are unique to those tables.
The optional keyword AS is often used to distinguish a table alias from other table
names.

In-Line Views
The FROM clause can itself contain a query-expression that takes an optional table
alias. This kind of nested query-expression is called an in-line view. An in-line view is
any query-expression that would be valid in a CREATE VIEW statement. PROC SQL
can support many levels of nesting, but it is limited to 32 tables in any one query. The
32-table limit includes underlying tables that may contribute to views that are specied
in the FROM clause.
An in-line view saves you a programming step. Rather than creating a view and
referring to it in another query, you can specify the view in-line in the FROM clause.
Characteristics of in-line views include the following:

3 An in-line view is not assigned a permanent name, although it can take an alias.
3 An in-line view can be referred to only in the query in which it is dened. It
cannot be referenced in another query.

3 You cannot use an ORDER BY clause in an in-line view.

The SQL Procedure

GROUP BY Clause

1065

3 The names of columns in an in-line view can be assigned in the object-item list of
that view or with a parenthesized list of names following the alias. This syntax
can be useful for renaming columns. See Example 10 on page 1148 for an example.
3 In order to visually separate an in-line view from the rest of the query, you can
enclose the in-line view in any number of pairs of parentheses. Note that if you
specify an alias for the in-line view, the alias specication must appear outside the
outermost pair of parentheses for that in-line view.

WHERE Clause
Subsets the output based on specied conditions.
Featured in: Example 4 on page 1131 and Example 9 on page 1145

WHERE sql-expression

Argument
sql-expression

See sql-expression on page 1099.

Details
3 When a condition is met (that is, the condition resolves to true), those rows are
displayed in the result table; otherwise, no rows are displayed.
3 You cannot use summary functions that specify only one column. For example:
where max(measure1) > 50;

However, this WHERE clause will work:

where max(measure1,measure2) > 50;

GROUP BY Clause
Species how to group the data for summarizing.
Featured in: Example 8 on page 1143 and Example 12 on page 1152

GROUP BY group-by-item <, , group-by-item>

1066

GROUP BY Clause

Chapter 44

Arguments
group-by-item

is one of the following:

integer
is a positive integer that equates to a columns position.
column-name
is the name of a column or a column alias. See column-name on page 1078.
sql-expression
See sql-expression on page 1099.

Details
3 You can specify more than one group-by-item to get more detailed reports. Both
the grouping of multiple items and the BY statement of a PROC step are
evaluated in similar ways. If more than one group-by-item is specied, then the
rst one determines the major grouping.

3 Integers can be substituted for column names (that is, SELECT object-items) in
the GROUP BY clause. For example, if the group-by-item is 2, then the results are
grouped by the values in the second column of the SELECT clause list. Using
integers can shorten your coding and enable you to group by the value of an
unnamed expression in the SELECT list. Note that if you use a oating-point
value (for example, 2.3), then PROC SQL ignores the decimal portion.

3 The data does not have to be sorted in the order of the group-by values because
PROC SQL handles sorting automatically. You can use the ORDER BY clause to
specify the order in which rows are displayed in the result table.

3 If you specify a GROUP BY clause in a query that does not contain a summary
function, then your clause is transformed into an ORDER BY clause and a
message to that effect is written to the SAS log.

3 You can group the output by the values that are returned by an expression. For
example, if X is a numeric variable, then the output of the following is grouped by
the integer portion of values of X:
select x, sum(y)
from table1
group by int(x);

Similarly, if Y is a character variable, then the output of the following is grouped

by the second character of values of Y:
select sum(x), y
from table1
group by substring(y from 2 for 1);

Note that an expression that contains only numeric literals (and functions of
numeric literals) or only character literals (and functions of character literals) is
ignored.
An expression in a GROUP BY clause cannot be a summary function. For
example, the following GROUP BY clause is not valid:
group by sum(x)

The SQL Procedure

ORDER BY Clause

1067

HAVING Clause
Subsets grouped data based on specied conditions.
Featured in: Example 8 on page 1143 and Example 12 on page 1152

HAVING sql-expression

Argument
sql-expression

See sql-expression on page 1099.

Subsetting Grouped Data

The HAVING clause is used with at least one summary function and an optional
GROUP BY clause to summarize groups of data in a table. A HAVING clause is any
valid SQL expression that is evaluated as either true or false for each group in a query.
Alternatively, if the query involves remerged data, then the HAVING expression is
evaluated for each row that participates in each group. The query must include one or
more summary functions.
Typically, the GROUP BY clause is used with the HAVING expression and denes
the group(s) to be evaluated. If you omit the GROUP BY clause, then the summary
function and the HAVING clause treat the table as one group.
The following PROC SQL step uses the PROCLIB.PAYROLL table (shown in
Example 2 on page 1127) and groups the rows by Gender to determine the oldest
employee of each gender. In SAS, dates are stored as integers. The lower the birth date
as an integer, the greater the age. The expression birth=min(birth)is evaluated for
each row in the table. When the minimum birth date is found, the expression becomes
true and the row is included in the output.
proc sql;
title Oldest Employee of Each Gender;
select *
from proclib.payroll
group by gender
having birth=min(birth);

Note: This query involves remerged data because the values returned by a
summary function are compared to values of a column that is not in the GROUP BY
clause. See Remerging Data on page 1110 for more information about summary
functions and remerging data. 4

ORDER BY Clause
Species the order in which rows are displayed in a result table.
See also: query-expression on page 1093
Featured in: Example 11 on page 1150

1068

ORDER BY Clause

Chapter 44

ORDER BY order-by-item <ASC|DESC><, order-by-item <ASC|DESC>>;

Arguments
order-by-item

is one of the following:

integer
equates to a columns position.
column-name
is the name of a column or a column alias. See column-name on page 1078.
sql-expression
See sql-expression on page 1099.
ASC

orders the data in ascending order. This is the default order; if neither ASC nor
DESC is specied, the data is ordered in ascending order.
DESC

orders the data in descending order.

Details
3 The ORDER BY clause sorts the result of a query expression according to the

3
3

order specied in that query. When this clause is used, the default ordering
sequence is ascending, from the lowest value to the highest. You can use the
SORTSEQ= option to change the collating sequence for your output. See PROC
SQL Statement on page 1033.
If an ORDER BY clause is omitted, then a particular order to the output rows,
such as the order in which the rows are encountered in the queried table, cannot
be guaranteed. Without an ORDER BY clause, the order of the output rows is
determined by the internal processing of PROC SQL, the default collating
sequence of SAS, and your operating environment. Therefore, if you want your
result table to appear in a particular order, then use the ORDER BY clause.
If more than one order-by-item is specied (separated by commas), then the rst
one determines the major sort order.
Integers can be substituted for column names (that is, SELECT object-items) in
the ORDER BY clause. For example, if the order-by-item is 2 (an integer), then the
results are ordered by the values of the second column. If a query-expression
includes a set operator (for example, UNION), then use integers to specify the
order. Doing so avoids ambiguous references to columns in the table expressions.
Note that if you use a oating-point value (for example, 2.3) instead of an integer,
then PROC SQL ignores the decimal portion.
In the ORDER BY clause, you can specify any column of a table or view that is
specied in the FROM clause of a query-expression, regardless of whether that
column has been included in the querys SELECT clause. For example, this query
produces a report ordered by the descending values of the population change for
each country from 1990 to 1995:
proc sql;
select country

The SQL Procedure

UPDATE Statement

1069

from census
order by pop95-pop90 desc;

NOTE: The query as specified involves

ordering by an item that
doesnt appear in its SELECT clause.

3 You can order the output by the values that are returned by an expression. For
example, if X is a numeric variable, then the output of the following is ordered by
the integer portion of values of X:
select x, y
from table1
order by int(x);

Similarly, if Y is a character variable, then the output of the following is ordered

by the second character of values of Y:
select x, y
from table1
order by substring(y from 2 for 1);

Note that an expression that contains only numeric literals (and functions of
numeric literals) or only character literals (and functions of character literals) is
ignored.

UPDATE Statement
Modies a columns values in existing rows of a table or view.
Restriction: You cannot use UPDATE on a table that is accessed by an engine that does
not support UPDATE processing.
Featured in: Example 3 on page 1129

UPDATE table-name|sas/access-view|proc-sql-view <AS alias>

SET column=sql-expression
<, column=sql-expression>
<SET column=sql-expression
<, column=sql-expression>>
<WHERE sql-expression>;

Arguments
alias

assigns an alias to table-name, sas/access-view, or proc-sql-view.

column

species a column in table-name, sas/access-view, or proc-sql-view.

sas/access-view

species a SAS/ACCESS view.

1070

VALIDATE Statement

Chapter 44

sql-expression

See sql-expression on page 1099.

table-name

species a PROC SQL table. table-name can be a one-level name, a two-level

libref.table name, or a physical pathname that is enclosed in single quotation marks.
proc-sql-view

species a PROC SQL view. proc-sql-view can be a one-level name, a two-level

libref.view name, or a physical pathname that is enclosed in single quotation marks.

Updating Tables through Views

You can update one or more rows of a table through a view, with some restrictions.
See Updating PROC SQL and SAS/ACCESS Views on page 1121.

Details
3 Any column that is not modied retains its original values, except in certain
queries using the CASE expression. See CASE expression on page 1073 for a
description of CASE expressions.

3 To add, drop, or modify a columns denition or attributes, use the ALTER TABLE
statement, described in ALTER TABLE Statement on page 1038.

3 In the SET clause, a column reference on the left side of the equal sign can also
appear as part of the expression on the right side of the equal sign. For example,
you could use this expression to give employees a $1,000 holiday bonus:
set salary=salary + 1000

3 If you omit the WHERE clause, then all the rows are updated. When you use a
WHERE clause, only the rows that meet the WHERE condition are updated.

3 When you update a column and an index has been dened for that column, the
values in the updated column continue to have the index dened for them.

VALIDATE Statement
Checks the accuracy of a query-expressions syntax and semantics without executing the
expression.
VALIDATE query-expression;

Argument
query-expression

See query-expression on page 1093.

The SQL Procedure

BETWEEN condition

1071

Details
3 The VALIDATE statement writes a message in the SAS log that states that the
query is valid. If there are errors, then VALIDATE writes error messages to the
SAS log.
3 The VALIDATE statement can also be included in applications that use the macro
facility. When used in such an application, VALIDATE returns a value that
indicates the query-expressions validity. The value is returned through the macro
variable SQLRC (a short form for SQL return code). For example, if a SELECT
statement is valid, then the macro variable SQLRC returns a value of 0. See
Using Macro Variables Set by PROC SQL on page 1119 for more information.

SQL Procedure Component Dictionary

This section describes the components that are used in SQL procedure statements.
Components are the items in PROC SQL syntax that appear in roman type.
Most components are contained in clauses within the statements. For example, the
basic SELECT statement is composed of the SELECT and FROM clauses, where each
clause contains one or more components. Components can also contain other
components.
For easy reference, components appear in alphabetical order, and some terms are
referred to before they are dened. Use the index or the See Also references to refer
to other statement or component descriptions that may be helpful.

BETWEEN condition
Selects rows where column values are within a range of values.
sql-expression <NOT> BETWEEN sql-expression
AND sql-expression

Argument
sql-expression

is described in sql-expression on page 1099.

Details
3 The sql-expressions must be of compatible data types. They must be either all
numeric or all character types.
3 Because a BETWEEN condition evaluates the boundary values as a range, it is
not necessary to specify the smaller quantity rst.
3 You can use the NOT logical operator to exclude a range of numbers, for example,
to eliminate customer numbers between 1 and 15 (inclusive) so that you can
retrieve data on more recently acquired customers.

1072

BTRIM function

Chapter 44

3 PROC SQL supports the same comparison operators that the DATA step supports.
For example:
x between 1 and 3
x between 3 and 1
1<=x<=3
x>=1 and x<=3

BTRIM function
Removes blanks or specied characters from the beginning, the end, or both the beginning and
end of a character string.
BTRIM (<< btrim-specication> <btrim-character FROM>> sql-expression)

Arguments
btrim-specication

is one of the following:

LEADING
removes the blanks or specied characters from the beginning of the character
string.
TRAILING
removes the blanks or specied characters from the end of the character string.
BOTH
removes the blanks or specied characters from both the beginning and the end of
the character string.
Default: BOTH
btrim-character

is a single character that is to be removed from the character string. The default
character is a blank.
sql-expression

must resolve to a character string or character variable and is described in

sql-expression on page 1099.

Details
The BTRIM function operates on character strings. BTRIM removes one or more
instances of a single character (the value of btrim-character) from the beginning, the
end, or both the beginning and end of a string, depending whether LEADING,
TRAILING, or BOTH is specied. If btrim-specication is not specied, then BOTH is
used. If btrim-character is omitted, then blanks are removed.
Note: SAS adds trailing blanks to character values that are shorter than the length
of the variable. Suppose you have a character variable Z, with length 10, and a value
xxabcxx. SAS stores the value with three blanks after the last x (for a total length of
10). If you attempt to remove all the x characters with
btrim(both x from z)

The SQL Procedure

CASE expression

1073

then the result is abcxx because PROC SQL sees the trailing characters as blanks, not
the x character. In order to remove all the x characters, use
btrim(both x from btrim(z))

The inner BTRIM function removes the trailing blanks before passing the value to the
outer BTRIM function. 4

CALCULATED
Refers to columns already calculated in the SELECT clause.
CALCULATED column-alias

Argument
column-alias

is the name that is assigned to the column in the SELECT clause.

Referencing a CALCULATED Column

CALCULATED enables you to use the results of an expression in the same SELECT
clause or in the WHERE clause. It is valid only when used to refer to columns that are
calculated in the immediate query expression.

CASE expression
Selects result values that satisfy specied conditions.
Featured in: Example 3 on page 1129 and Example 13 on page 1154

CASE <case-operand>
WHEN when-condition THEN result-expression
<WHEN when-condition THEN result-expression>
<ELSE result-expression>
END

1074

CASE expression

Chapter 44

Arguments
case-operand

is a valid sql-expression that resolves to a table column whose values are compared
to all the when-conditions. See sql-expression on page 1099.
when-condition

3 When case-operand is specied, when-condition is a shortened sql-expression

that assumes case-operand as one of its operands and that resolves to true or
false.
3 When case-operand is not specied, when-condition is an sql-expression that
resolves to true or false.
result-expression

is an sql-expression that resolves to a value.

Details
The CASE expression selects values if certain conditions are met. A CASE expression
returns a single value that is conditionally evaluated for each row of a table (or view).
Use the WHEN-THEN clauses when you want to execute a CASE expression for some
but not all of the rows in the table that is being queried or created. An optional ELSE
expression gives an alternative action if no THEN expression is executed.
When you omit case-operand, when-condition is evaluated as a Boolean (true or false)
value. If when-condition returns a nonzero, nonmissing result, then the WHEN clause
is true. If case-operand is specied, then it is compared with when-condition for
equality. If case-operand equals when-condition, then the WHEN clause is true.
If the when-condition is true for the row that is being executed, then the
result-expression that follows THEN is executed. If when-condition is false, then PROC
SQL evaluates the next when-condition until they are all evaluated. If every
when-condition is false, then PROC SQL executes the ELSE expression, and its result
becomes the CASE expressions result. If no ELSE expression is present and every
when-condition is false, then the result of the CASE expression is a missing value.
You can use a CASE expression as an item in the SELECT clause and as either
operand in an sql-expression.

Example
The following two PROC SQL steps show two equivalent CASE expressions that
create a character column with the strings in the THEN clause. The CASE expression
in the second PROC SQL step is a shorthand method that is useful when all the
comparisons are with the same column.
proc sql;
select Name, case
when Continent = North America then Continental U.S.
when Continent = Oceania then Pacific Islands
else None
end as Region
from states;
proc sql;
select Name, case Continent
when North America then Continental U.S.
when Oceania then Pacific Islands

The SQL Procedure

column-denition

1075

else None
end as Region
from states;

Note: When you use the shorthand method, the conditions must all be equality
tests. That is, they cannot use comparison operators or other types of operators. 4

COALESCE Function
Returns the rst nonmissing value from a list of columns.
Featured in: Example 7 on page 1138

COALESCE (column-name <, column-name>)

Arguments
column-name

is described in column-name on page 1078.

Details
COALESCE accepts one or more column names of the same data type. The
COALESCE function checks the value of each column in the order in which they are
listed and returns the rst nonmissing value. If only one column is listed, the
COALESCE function returns the value of that column. If all the values of all
arguments are missing, the COALESCE function returns a missing value.
In some SQL DBMSs, the COALESCE function is called the IFNULL function. See
PROC SQL and the ANSI Standard on page 1122 for more information.
Note: If your query contains a large number of COALESCE function calls, it might
be more efcient to use a natural join instead. See Natural Joins on page 1088. 4

column-denition
Denes PROC SQLs data types and dates.
See also: column-modier on page 1076
Featured in: Example 1 on page 1125

column data-type < column-modier < column-modier>>

1076

column-modier

Chapter 44

Arguments
column

is a column name.
column-modier

is described in column-modier on page 1076.

data-type

is one of the following data types:

CHARACTER|VARCHAR <(width)>
indicates a character column with a column width of width. The default column
width is eight characters.
INTEGER|SMALLINT
indicates an integer column.
DECIMAL|NUMERIC|FLOAT <(width<, ndec>)>
indicates a oating-point column with a column width of width and ndec decimal
places.
REAL|DOUBLE PRECISION
indicates a oating-point column.
DATE
indicates a date column.

Details
3 SAS supports many but not all of the data types that SQL-based databases
support.

3 For all the numeric data types (INTEGER, SMALLINT, DECIMAL, NUMERIC,
FLOAT, REAL, DOUBLE PRECISION, and DATE), the SQL procedure defaults to
the SAS data type NUMERIC. The width and ndec arguments are ignored; PROC
SQL creates all numeric columns with the maximum precision allowed by SAS. If
you want to create numeric columns that use less storage space, then use the
LENGTH statement in the DATA step. The various numeric data type names,
along with the width and ndec arguments, are included for compatibility with
other SQL software.
3 For the character data types (CHARACTER and VARCHAR), the SQL procedure
defaults to the SAS data type CHARACTER. The width argument is honored.
3 The CHARACTER, INTEGER, and DECIMAL data types can be abbreviated to
CHAR, INT, and DEC, respectively.
3 A column that is declared with DATE is a SAS numeric variable with a date
informat or format. You can use any of the column-modiers to set the appropriate
attributes for the column that is being dened. See SAS Language Reference:
Dictionary for more information on dates.

column-modier
Sets column attributes.
See also: column-denition on page 1075 and SELECT Clause on page 1058
Featured in:

Example 1 on page 1125 and Example 2 on page 1127

The SQL Procedure

column-modier

1077

column-modier

Arguments
column-modier

is one of the following:

INFORMAT=informatw.d
species a SAS informat to be used when SAS accesses data from a table or view.
You can change one permanent informat to another by using the ALTER
statement. PROC SQL stores informats in its table denitions so that other SAS
procedures and the DATA step can use this information when they reference tables
created by PROC SQL.
See SAS Language Reference: Dictionary for more information about informats.
FORMAT=formatw.d
species a SAS format for determining how character and numeric values in a
column are displayed by the query-expression. If the FORMAT= modier is used
in the ALTER, CREATE TABLE, or CREATE VIEW statements, then it species
the permanent format to be used when SAS displays data from that table or view.
You can change one permanent format to another by using the ALTER statement.
See SAS Language Reference: Dictionary for more information about formats.
LABEL=label
species a column label. If the LABEL= modier is used in the ALTER, CREATE
TABLE, or CREATE VIEW statements, then it species the permanent label to be
used when displaying that column. You can change one permanent label to
another by using the ALTER statement.
A label can begin with the following characters: a through z, A through Z, 0
through 9, an underscore (_), or a blank space. If you begin a label with any other
character, such as pound sign (#), then that character is used as a split character
and it splits the label onto the next line wherever it appears. For example:
select dropout label=
#Percentage of#Students Who#Dropped Out
from educ(obs=5);

If a special character must appear as the rst character in the output, then
precede it with a space or a forward slash (/).
You can omit the LABEL= part of the column-modier and still specify a label.
Be sure to enclose the label in quotation marks, as in this example:
select empname "Names of Employees"
from sql.employees;

If an apostrophe must appear in the label, then type it twice so that SAS reads
the apostrophe as a literal. Alternatively, you can use single and double quotation
marks alternately (for example, Date Recd).
LENGTH=length
species the length of the column. This column modier is valid only in the
context of a SELECT statement.
TRANSCODE=YES|NO
for character columns, species whether values can be transcoded. Use
TRANSCODE=NO to suppress transcoding. Note that when you create a table by
using the CREATE TABLE AS statement, the transcoding attribute for a given
character column in the created table is the same as it is in the source table unless

1078

column-name

Chapter 44

you change it with the TRANSCODE= column modier. For more information
about transcoding, see SAS National Language Support (NLS): Users Guide.
Default: YES
Restriction: Suppression of transcoding is not supported for the V6TAPE engine.
Interaction: If the TRANSCODE= attribute is set to NO for any character
variable in a table, then PROC CONTENTS prints a transcode column that
contains the TRANSCODE= value for each variable in the data set. If all
variables in the table are set to the default TRANSCODE= value (YES), then no
transcode column is printed.

Details
If you refer to a labeled column in the ORDER BY or GROUP BY clause, then you
must use either the column name (not its label), the columns alias, or its ordering
integer (for example, ORDER BY 2). See the section on SAS statements in SAS
Language Reference: Dictionary for more information about labels.

column-name
Species the column to select.
See also: column-modier on page 1076 and SELECT Clause on page 1058

column-name

is one of the following:

column
is the name of a column.
table-name.column
is the name of a column in the table table-name.
table-alias.column
is the name of a column in the table that is referenced by table-alias.
view-name.column
is the name of a column in the view view-name.
view-alias.column
is the name of a column in the view that is referenced by view-alias.

Details
A column can be referred to by its name alone if it is the only column by that name
in all the tables or views listed in the current query-expression. If the same column
name exists in more than one table or view in the query-expression, then you must
qualify each use of the column name by prexing a reference to the table that contains
it. Consider the following examples:

The SQL Procedure

SALARY
EMP.SALARY
E.SALARY

CONTAINS condition

1079

/* name of the column */

/* EMP is the table or view name */
/* E is an alias for the table
or view that contains the
SALARY column */

CONNECTION TO
Retrieves and uses DBMS data in a PROC SQL query or view.
You can use CONNECTION TO in the SELECT statements FROM clause as part
of the from-list.

Tip:

See also: Connecting to a DBMS Using the SQL Procedure Pass-Through Facility on
page 1115 and your SAS/ACCESS documentation.

CONNECTION TO dbms-name (dbms-query)

CONNECTION TO alias (dbms-query)

Arguments
alias

species an alias, if one was dened in the CONNECT statement.

dbms-name

identies the DBMS that you are using.

dbms-query

species the query to send to a DBMS. The query uses the DBMSs dynamic SQL.
You can use any SQL syntax that the DBMS understands, even if that is not valid for
PROC SQL. However, your DBMS query cannot contain a semicolon because that
represents the end of a statement to SAS.
The number of tables that you can join with dbms-query is determined by the
DBMS. Each CONNECTION TO component counts as one table toward the 32-table
PROC SQL limit for joins.
See SAS/ACCESS for Relational Databases: Reference for more information about
DBMS queries.

CONTAINS condition
Tests whether a string is part of a columns value.
Alias:

Restriction:

The CONTAINS condition is used only with character operands.

Featured in: Example 7 on page 1138

1080

EXISTS condition

Chapter 44

sql-expression <NOT> CONTAINS sql-expression

Argument
sql-expression

is described in sql-expression on page 1099.

EXISTS condition
Tests if a subquery returns one or more rows.
See also: Query Expressions (Subqueries) on page 1102

<NOT> EXISTS (query-expression)

Argument

query-expression

is described in query-expression on page 1093.

Details
The EXISTS condition is an operator whose right operand is a subquery. The result
of an EXISTS condition is true if the subquery resolves to at least one row. The result
of a NOT EXISTS condition is true if the subquery evaluates to zero rows. For example,
the following query subsets PROCLIB.PAYROLL (which is shown in Example 2 on page
1127) based on the criteria in the subquery. If the value for STAFF.IDNUM is on the
same row as the value CT in PROCLIB.STAFF (which is shown in Example 4 on page
1131), then the matching IDNUM in PROCLIB.PAYROLL is included in the output.
Thus, the query returns all the employees from PROCLIB.PAYROLL who live in CT.
proc sql;
select *
from proclib.payroll p
where exists (select *
from proclib.staff s
where p.idnumber=s.idnum
and state=CT);

IN condition
Tests set membership.
Featured in:

Example 4 on page 1131

The SQL Procedure

IS condition

1081

sql-expression <NOT> IN (query-expression | constant <, constant>)

Arguments

constant

is a number or a quoted character string (or other special notation) that indicates a
xed value. Constants are also called literals.
query-expression

is described in query-expression on page 1093.

sql-expression

is described in sql-expression on page 1099.

Details
An IN condition tests if the column value that is returned by the sql-expression on
the left is a member of the set (of constants or values returned by the query-expression)
on the right. The IN condition is true if the value of the left-hand operand is in the set
of values that are dened by the right-hand operand.

IS condition
Tests for a missing value.
Featured in: Example 5 on page 1134

sql-expression IS <NOT> NULL | MISSING

Argument

sql-expression

is described in sql-expression on page 1099.

Details
IS NULL and IS MISSING are predicates that test for a missing value. IS NULL and
IS MISSING are used in the WHERE, ON, and HAVING expressions. Each predicate
resolves to true if the sql-expressions result is missing and false if it is not missing.
SAS stores a numeric missing value as a period (.) and a character missing value as
a blank space. Unlike missing values in some versions of SQL, missing values in SAS
always appear rst in the collating sequence. Therefore, in Boolean and comparison
operations, the following expressions resolve to true in a predicate:
3>null
-3>null
0>null

1082

joined-table

Chapter 44

The SAS way of evaluating missing values differs from that of the ANSI Standard for
SQL. According to the Standard, these expressions are NULL. See sql-expression on
page 1099 for more information on predicates and operators. See PROC SQL and the
ANSI Standard on page 1122 for more information on the ANSI Standard.

joined-table
Joins a table with itself or with other tables or views.
Restrictions:

Joins are limited to 32 tables.

See also: FROM Clause on page 1063 and query-expression on page 1093
Featured in: Example 4 on page 1131, Example 7 on page 1138, Example 9 on page 1145,
Example 13 on page 1154, and Example 14 on page 1158

u table-name << AS> alias>, table-name << AS> alias>

<, table-name < <AS> alias>>
v <(>table-name < INNER> JOIN table-name
ON sql-expression<)>
w <(>table-name LEFT JOIN | RIGHT JOIN | FULL JOIN
table-name ON sql-expression<)>
x <(>table-name CROSS JOIN table-name<)>
y <(>table-name UNION JOIN table-name<)>
U <(>table-name NATURAL
<INNER | FULL <OUTER> | LEFT <OUTER > | RIGHT <OUTER >>
JOIN table-name<)>

Arguments
alias

species an alias for table-name. The AS keyword is optional.

sql-expression

is described in sql-expression on page 1099.

table-name

can be one of the following:

3 the name of a PROC SQL table.

3 the name of a SAS data view or PROC SQL view.
3 a query-expression. A query-expression in the FROM clause is usually referred
to as an in-line view. See FROM Clause on page 1063 for more information
about in-line views.

3 a connection to a DBMS in the form of the CONNECTION TO component. See

CONNECTION TO on page 1079 for more information.
table-name can be a one-level name, a two-level libref.table name, or a physical
pathname that is enclosed in single quotation marks.

The SQL Procedure

joined-table

1083

Note: If you include parentheses, then be sure to include them in pairs.

Parentheses are not valid around comma joins (type u). 4

Types of Joins
uv Inner join. See Inner Joins on page 1084.
w Outer join. See Outer Joins on page 1086.
x Cross join. See Cross Joins on page 1087.
y Union join. See Union Joins on page 1088.
U Natural join. See Natural Joins on page 1088.

Joining Tables
When multiple tables, views, or query-expressions are listed in the FROM clause,
they are processed to form one table. The resulting table contains data from each
contributing table. These queries are referred to as joins.
Conceptually, when two tables are specied, each row of table A is matched with all
the rows of table B to produce an internal or intermediate table. The number of rows in
the intermediate table (Cartesian product) is equal to the product of the number of rows
in each of the source tables. The intermediate table becomes the input to the rest of the
query in which some of its rows may be eliminated by the WHERE clause or
summarized by a summary function.
A common type of join is an equijoin, in which the values from a column in the rst
table must equal the values of a column in the second table.

Table Limit
PROC SQL can process a maximum of 32 tables for a join. If you are using views in
a join, then the number of tables on which the views are based count toward the
32-table limit. Each CONNECTION TO component in the Pass-Through Facility counts
as one table.

Specifying the Rows to Be Returned

The WHERE clause or ON clause contains the conditions (sql-expression) under
which the rows in the Cartesian product are kept or eliminated in the result table.
WHERE is used to select rows from inner joins. ON is used to select rows from inner or
outer joins.
The expression is evaluated for each row from each table in the intermediate table
described earlier in Joining Tables on page 1083. The row is considered to be matching
if the result of the expression is true (a nonzero, nonmissing value) for that row.
Note: You can follow the ON clause with a WHERE clause to further subset the
query result. See Example 7 on page 1138 for an example. 4

Table Aliases
Table aliases are used in joins to distinguish the columns of one table from those in
the other table(s). A table name or alias must be prexed to a column name when you
are joining tables that have matching column names. See FROM Clause on page 1063
for more information on table aliases.

1084

joined-table

Chapter 44

Joining a Table with Itself

A single table can be joined with itself to produce more information. These joins are
sometimes called reexive joins. In these joins, the same table is listed twice in the
FROM clause. Each instance of the table must have a table alias or you will not be able
to distinguish between references to columns in either instance of the table. See
Example 13 on page 1154 and Example 14 on page 1158 for examples.

Inner Joins
An inner join returns a result table for all the rows in a table that have one or more
matching rows in the other table(s), as specied by the sql-expression. Inner joins can
be performed on up to 32 tables in the same query-expression.
You can perform an inner join by using a list of table-names separated by commas or
by using the INNER, JOIN, and ON keywords.
The LEFTTAB and RIGHTTAB tables are used to illustrate this type of join:

Left Table - LEFTTAB

Continent Export
Country
----------------------------NA
wheat
Canada
EUR
corn
France
EUR
rice
Italy
AFR
oil
Egypt

Right Table - RIGHTTAB

Continent Export
Country
----------------------------NA
sugar
USA
EUR
corn
Spain
EUR
beets
Belgium
ASIA
rice
Vietnam

The following example joins the LEFTTAB and RIGHTTAB tables to get the
Cartesian product of the two tables. The Cartesian product is the result of combining
every row from one table with every row from another table. You get the Cartesian
product when you join two tables and do not subset them with a WHERE clause or ON
clause.
proc sql;
title The Cartesian Product of;
title2 LEFTTAB and RIGHTTAB;
select *
from lefttab, righttab;

The SQL Procedure

joined-table

1085

The Cartesian Product of

LEFTTAB and RIGHTTAB
Continent Export
Country
Continent Export
Country
-----------------------------------------------------------NA
wheat
Canada
NA
sugar
USA
NA
wheat
Canada
EUR
corn
Spain
NA
wheat
Canada
EUR
beets
Belgium
NA
wheat
Canada
ASIA
rice
Vietnam
EUR
corn
France
NA
sugar
USA
EUR
corn
France
EUR
corn
Spain
EUR
corn
France
EUR
beets
Belgium
EUR
corn
France
ASIA
rice
Vietnam
EUR
rice
Italy
NA
sugar
USA
EUR
rice
Italy
EUR
corn
Spain
EUR
rice
Italy
EUR
beets
Belgium
EUR
rice
Italy
ASIA
rice
Vietnam
AFR
oil
Egypt
NA
sugar
USA
AFR
oil
Egypt
EUR
corn
Spain
AFR
oil
Egypt
EUR
beets
Belgium
AFR
oil
Egypt
ASIA
rice
Vietnam

The LEFTTAB and RIGHTTAB tables can be joined by listing the table names in the
FROM clause. The following query represents an equijoin because the values of
Continent from each table are matched. The column names are prexed with the table
aliases so that the correct columns can be selected.
proc sql;
title Inner Join;
select *
from lefttab as l, righttab as r
where l.continent=r.continent;

Inner Join
Continent Export
Country
Continent Export
Country
-----------------------------------------------------------NA
wheat
Canada
NA
sugar
USA
EUR
corn
France
EUR
corn
Spain
EUR
corn
France
EUR
beets
Belgium
EUR
rice
Italy
EUR
corn
Spain
EUR
rice
Italy
EUR
beets
Belgium

The following PROC SQL step is equivalent to the previous one and shows how to
write an equijoin using the INNER JOIN and ON keywords.
proc sql;
title Inner Join;
select *
from lefttab as l inner join
righttab as r
on l.continent=r.continent;

See Example 4 on page 1131, Example 13 on page 1154, and Example 14 on page
1158 for more examples.

1086

joined-table

Chapter 44

Outer Joins
Outer joins are inner joins that have been augmented with rows that did not match
with any row from the other table in the join. The three types of outer joins are left,
right, and full.
A left outer join, specied with the keywords LEFT JOIN and ON, has all the rows
from the Cartesian product of the two tables for which the sql-expression is true, plus
rows from the rst (LEFTTAB) table that do not match any row in the second
(RIGHTTAB) table.
proc sql;
title Left Outer Join;
select *
from lefttab as l left join
righttab as r
on l.continent=r.continent;

Left Outer Join

Continent Export
Country
Continent Export
Country
-----------------------------------------------------------AFR
oil
Egypt
EUR
rice
Italy
EUR
beets
Belgium
EUR
corn
France
EUR
beets
Belgium
EUR
rice
Italy
EUR
corn
Spain
EUR
corn
France
EUR
corn
Spain
NA
wheat
Canada
NA
sugar
USA

A right outer join, specied with the keywords RIGHT JOIN and ON, has all the
rows from the Cartesian product of the two tables for which the sql-expression is true,
plus rows from the second (RIGHTTAB) table that do not match any row in the rst
(LEFTTAB) table.
proc sql;
title Right Outer Join;
select *
from lefttab as l right join
righttab as r
on l.continent=r.continent;

Right Outer Join

Continent Export
Country
Continent Export
Country
-----------------------------------------------------------ASIA
rice
Vietnam
EUR
rice
Italy
EUR
beets
Belgium
EUR
rice
Italy
EUR
corn
Spain
EUR
corn
France
EUR
beets
Belgium
EUR
corn
France
EUR
corn
Spain
NA
wheat
Canada
NA
sugar
USA

A full outer join, specied with the keywords FULL JOIN and ON, has all the rows
from the Cartesian product of the two tables for which the sql-expression is true, plus
rows from each table that do not match any row in the other table.

The SQL Procedure

joined-table

proc sql;
title Full Outer Join;
select *
from lefttab as l full join
righttab as r
on l.continent=r.continent;

Full Outer Join

Continent Export
Country
Continent Export
Country
-----------------------------------------------------------AFR
oil
Egypt
ASIA
rice
Vietnam
EUR
rice
Italy
EUR
beets
Belgium
EUR
rice
Italy
EUR
corn
Spain
EUR
corn
France
EUR
beets
Belgium
EUR
corn
France
EUR
corn
Spain
NA
wheat
Canada
NA
sugar
USA

See Example 7 on page 1138 for another example.

Cross Joins
A cross join returns as its result table the product of the two tables.
Using the LEFTTAB and RIGHTTAB example tables, the following program
demonstrates the cross join:
proc sql;
title Cross Join;
select *
from lefttab as l cross join
righttab as r;

Cross Join
Continent Export
Country
Continent Export
Country
-----------------------------------------------------------NA
wheat
Canada
NA
sugar
USA
NA
wheat
Canada
EUR
corn
Spain
NA
wheat
Canada
EUR
beets
Belgium
NA
wheat
Canada
ASIA
rice
Vietnam
EUR
corn
France
NA
sugar
USA
EUR
corn
France
EUR
corn
Spain
EUR
corn
France
EUR
beets
Belgium
EUR
corn
France
ASIA
rice
Vietnam
EUR
rice
Italy
NA
sugar
USA
EUR
rice
Italy
EUR
corn
Spain
EUR
rice
Italy
EUR
beets
Belgium
EUR
rice
Italy
ASIA
rice
Vietnam
AFR
oil
Egypt
NA
sugar
USA
AFR
oil
Egypt
EUR
corn
Spain
AFR
oil
Egypt
EUR
beets
Belgium
AFR
oil
Egypt
ASIA
rice
Vietnam

1087

1088

joined-table

Chapter 44

The cross join is not functionally different from a Cartesian product join. You would
get the same result by submitting the following program:
proc sql;
select *
from lefttab, righttab;

Do not use an ON clause with a cross join. An ON clause will cause a cross join to
fail. However, you can use a WHERE clause to subset the output.

Union Joins
A union join returns a union of the columns of both tables. The union join places in
the results all rows with their respective column values from each input table. Columns
that do not exist in one table will have null (missing) values for those rows in the result
table. The following example demonstrates a union join.
proc sql;
title Union Join;
select *
from lefttab union join righttab;

Union Join
Continent Export
Country
Continent Export
Country
-----------------------------------------------------------NA
sugar
USA
EUR
corn
Spain
EUR
beets
Belgium
ASIA
rice
Vietnam
NA
wheat
Canada
EUR
corn
France
EUR
rice
Italy
AFR
oil
Egypt

Using a union join is similar to concatenating tables with the OUTER UNION set
operator. See query-expression on page 1093 for more information.
Do not use an ON clause with a union join. An ON clause will cause a union join to
fail.

Natural Joins
A natural join selects rows from two tables that have equal values in columns that
share the same name and the same type. An error results if two columns have the same
name but different types. If join-specication is omitted when specifying a natural join,
then INNER is implied. If no like columns are found, then a cross join is performed.
The following examples use these two tables:

table1
x
y
z
---------------------------1
2
3
2
1
8
6
5
4
2
5
6

The SQL Procedure

joined-table

1089

table2
x
b
z
---------------------------1
5
3
3
5
4
2
7
8
6
0
4

The following program demonstrates a natural inner join.

proc sql;
title Natural Inner Join;
select *
from table1 natural join table2;

Natural Inner Join

x
z
b
y
-------------------------------------1
3
5
2
2
8
7
1
6
4
0
5

The following program demonstrates a natural left outer join.

proc sql;
title Natural Left Outer Join;
select *
from table1 natural left join table2;

Natural Left Outer Join

x
z
b
y
-------------------------------------1
3
5
2
2
6
.
5
2
8
7
1
6
4
0
5

Do not use an ON clause with a natural join. An ON clause will cause a natural join
to fail. When using a natural join, an ON clause is implied, matching all like columns.

Joining More Than Two Tables

Inner joins are usually performed on two or three tables, but they can be performed
on up to 32 tables in PROC SQL. A join on three tables is described here to explain how
and why the relationships work among the tables.
In a three-way join, the sql-expression consists of two conditions: one relates the rst
table to the second table and the other relates the second table to the third table. It is
possible to break this example into stages, performing a two-way join into a temporary

1090

joined-table

Chapter 44

table and then joining that table with the third one for the same result. However,
PROC SQL can do it all in one step as shown in the next example.
The example shows the joining of three tables: COMM, PRICE, and AMOUNT. To
calculate the total revenue from exports for each country, you need to multiply the
amount exported (AMOUNT table) by the price of each unit (PRICE table), and you
must know the commodity that each country exports (COMM table).

COMM Table
Continent Export
Country
----------------------------NA
wheat
Canada
EUR
corn
France
EUR
rice
Italy
AFR
oil
Egypt

PRICE Table
Export
Price
-----------------rice
3.56
corn
3.45
oil
18
wheat
2.98

AMOUNT Table
Country
Quantity
-----------------Canada
16000
France
2400
Italy
500
Egypt
10000

proc sql;
title Total Export Revenue;
select c.Country, p.Export, p.Price,
a.Quantity,a.quantity*p.price
as Total
from comm c, price p, amount a
where c.export=p.export
and c.country=a.country;

Total Export Revenue

Country
Export
Price Quantity
Total
-----------------------------------------------Italy
rice
3.56
500
1780
France
corn
3.45
2400
8280
Egypt
oil
18
10000
180000
Canada
wheat
2.98
16000
47680

The SQL Procedure

LIKE condition

1091

See Example 9 on page 1145 for another example.

Comparison of Joins and Subqueries

You can often use a subquery or a join to get the same result. However, it is often
more efcient to use a join if the outer query and the subquery do not return duplicate
rows. For example, the following queries produce the same result. The second query is
more efcient:
proc sql;
select IDNumber, Birth
from proclib.payroll
where IDNumber in (select idnum
from proclib.staff
where lname like B%);
proc sql;
select p.IDNumber, p.Birth
from proclib.payroll p, proclib.staff s
where p.idnumber=s.idnum
and s.lname like B%;

Note:

PROCLIB.PAYROLL is shown in Example 2 on page 1127.

LIKE condition
Tests for a matching pattern.
sql-expression <NOT> LIKE sql-expression < ESCAPE character-expression>

Arguments

sql-expression

is described in sql-expression on page 1099.

character-expression

is an sql-expression that evaluates to a single character. The operands of

character-expression must be character or string literals; they cannot be column
names.
Note: If you use an ESCAPE clause, then the pattern-matching specication must
be a quoted string or quoted concatenated string; it cannot contain column names. 4

Details
The LIKE condition selects rows by comparing character strings with a
pattern-matching specication. It resolves to true and displays the matched string(s) if
the left operand matches the pattern specied by the right operand.
The ESCAPE clause is used to search for literal instances of the percent (%) and
underscore (_) characters, which are usually used for pattern matching.

1092

LIKE condition

Chapter 44

Patterns for Searching

Patterns are composed of three classes of characters:
underscore (_)
matches any single character.
percent sign (%)
matches any sequence of zero or more characters.
any other character
matches that character.
These patterns can appear before, after, or on both sides of characters that you want to
match. The LIKE condition is case-sensitive.
The following list uses these values: Smith, Smooth, Smothers, Smart, and Smuggle.
Sm%

matches Smith, Smooth, Smothers, Smart, Smuggle.

%th

matches Smith, Smooth.

S__gg%

matches Smuggle.
S_o

matches a three-letter word, so it has no matches here.

S_o%

matches Smooth, Smothers.

S%th

matches Smith, Smooth.

matches the single, uppercase character Z only, so it has no matches here.

Searching for Literal % and _

Because the % and _ characters have special meaning in the context of the LIKE
condition, you must use the ESCAPE clause to search for these character literals in the
input character string.
These example use the values app, a_%, a__, bbaa1, and ba_1.

3 The condition like a_% matches app, a_%, and a__, because the underscore (_)
in the search pattern matches any single character (including the underscore), and
the percent (%) in the search pattern matches zero or more characters, including
% and _.

3 The condition like a_^% escape ^ matches only a_%, because the escape
character (^) species that the pattern search for a literal %.

3 The condition like a_% escape _ matches none of the values, because the
escape character (_) species that the pattern search for an a followed by a literal
%, which does not apply to any of these values.

Searching for Mixed-Case Strings

To search for mixed-case strings, use the UPCASE function to make all the names
uppercase before entering the LIKE condition:
upcase(name) like SM%;

The SQL Procedure

query-expression

1093

Note: When you are using the % character, be aware of the effect of trailing blanks.
You may have to use the TRIM function to remove trailing blanks in order to match
values. 4

LOWER function
Converts the case of a character string to lowercase.
See also: UPPER function on page 1114

LOWER (sql-expression)

Argument
sql-expression

must resolve to a character string and is described in sql-expression on page 1099.

Details
The LOWER function operates on character strings. LOWER changes the case of its
argument to all lowercase.
Note: The LOWER function is provided for compatibility with the ANSI SQL
standard. You can also use the SAS function LOWCASE. 4

query-expression
Retrieves data from tables.
See also: table-expression on page 1113, Query Expressions (Subqueries) on page
1102, and In-Line Views on page 1064

table-expression <set-operator table-expression> <set-operator table-expression>

Arguments
table-expression

is described in table-expression on page 1113.

1094

query-expression

Chapter 44

set-operator

is one of the following:

INTERSECT <CORRESPONDING> < ALL>
OUTER UNION <CORRESPONDING>
UNION <CORRESPONDING> < ALL>
EXCEPT <CORRESPONDING> < ALL>

Query Expressions and Table Expressions

A query-expression is one or more table-expressions. Multiple table expressions are
linked by set operators. The following gure illustrates the relationship between
table-expressions and query-expressions.

tableexpression
queryexpression

SELECT clause
FROM clause
(more clauses)

set operator

tableexpression

SELECT clause
FROM clause
(more clauses)

Set Operators
PROC SQL provides these set operators:
OUTER UNION
concatenates the query results.
UNION
produces all unique rows from both queries.
EXCEPT
produces rows that are part of the rst query only.
INTERSECT
produces rows that are common to both query results.
A query-expression with set operators is evaluated as follows.

3 Each table-expression is evaluated to produce an (internal) intermediate result

table.

3 Each intermediate result table then becomes an operand linked with a set
operator to form an expression, for example, A UNION B.

3 If the query-expression involves more than two table-expressions, then the result
from the rst two becomes an operand for the next set operator and operand, such
as (A UNION B) EXCEPT C, ((A UNION B) EXCEPT C) INTERSECT D, and so on.

3 Evaluating a query-expression produces a single output table.

Set operators follow this order of precedence unless they are overridden by
parentheses in the expression(s): INTERSECT is evaluated rst. OUTER UNION,
UNION, and EXCEPT have the same level of precedence.

The SQL Procedure

query-expression

1095

PROC SQL performs set operations even if the tables or views that are referred to in
the table-expressions do not have the same number of columns. The reason for this
behavior is that the ANSI Standard for SQL requires that tables or views that are
involved in a set operation have the same number of columns and that the columns have
matching data types. If a set operation is performed on a table or view that has fewer
columns than the one(s) with which it is being linked, then PROC SQL extends the table
or view with fewer columns by creating columns with missing values of the appropriate
data type. This temporary alteration enables the set operation to be performed correctly.

CORRESPONDING (CORR) Keyword

The CORRESPONDING keyword is used only when a set operator is specied.
CORR causes PROC SQL to match the columns in table-expressions by name and not
by ordinal position. Columns that do not match by name are excluded from the result
table, except for the OUTER UNION operator. See OUTER UNION on page 1095.
For example, when performing a set operation on two table-expressions, PROC SQL
matches the rst specied column-name (listed in the SELECT clause) from one
table-expression with the rst specied column-name from the other. If CORR is
omitted, then PROC SQL matches the columns by ordinal position.

ALL Keyword
The set operators automatically eliminate duplicate rows from their output tables.
The optional ALL keyword preserves the duplicate rows, reduces the execution by one
step, and thereby improves the query-expressions performance. You use it when you
want to display all the rows resulting from the table-expressions, rather than just the
unique rows. The ALL keyword is used only when a set operator is also specied.

OUTER UNION
Performing an OUTER UNION is very similar to performing the SAS DATA step
with a SET statement. The OUTER UNION concatenates the intermediate results from
the table-expressions. Thus, the result table for the query-expression contains all the
rows produced by the rst table-expression followed by all the rows produced by the
second table-expression. Columns with the same name are in separate columns in the
result table.
For example, the following query expression concatenates the ME1 and ME2 tables
but does not overlay like-named columns. Output 44.1 shows the result.

ME1
IDnum
Jobcode
Salary
Bonus
-------------------------------------1400
ME1
29769
587
1403
ME1
28072
342
1120
ME1
28619
986
1120
ME1
28619
986

1096

query-expression

Chapter 44

ME2
IDnum
Jobcode
Salary
---------------------------1653
ME2
35108
1782
ME2
35345
1244
ME2
36925

proc sql;
title ME1 and ME2: OUTER UNION;
select *
from me1
outer union
select *
from me2;

Output 44.1

OUTER UNION of ME1 and ME2 Tables

ME1 and ME2: OUTER UNION

IDnum
Jobcode
Salary
Bonus IDnum
Jobcode
Salary
-------------------------------------------------------------------1400
ME1
29769
587
.
1403
ME1
28072
342
.
1120
ME1
28619
986
.
1120
ME1
28619
986
.
.
. 1653
ME2
35108
.
. 1782
ME2
35345
.
. 1244
ME2
36925

Concatenating tables with the OUTER UNION set operator is similar to performing
a union join. See Union Joins on page 1088 for more information.
To overlay columns with the same name, use the CORRESPONDING keyword.
proc sql;
title ME1 and ME2: OUTER UNION CORRESPONDING;
select *
from me1
outer union corr
select *
from me2;

The SQL Procedure

query-expression

1097

ME1 and ME2: OUTER UNION CORRESPONDING

IDnum
Jobcode
Salary
Bonus
-------------------------------------1400
ME1
29769
587
1403
ME1
28072
342
1120
ME1
28619
986
1120
ME1
28619
986
1653
ME2
35108
.
1782
ME2
35345
.
1244
ME2
36925
.

In the resulting concatenated table, notice the following:

3 OUTER UNION CORRESPONDING retains all nonmatching columns.

3 For columns with the same name, if a value is missing from the result of the rst
table-expression, then the value in that column from the second table-expression is
inserted.

3 The ALL keyword is not used with OUTER UNION because this operators default
action is to include all rows in a result table. Thus, both rows from the table ME1
where IDnum is 1120 appear in the output.

UNION
The UNION operator produces a table that contains all the unique rows that result
from both table-expressions. That is, the output table contains rows produced by the
rst table-expression, the second table-expression, or both.
Columns are appended by position in the tables, regardless of the column names.
However, the data type of the corresponding columns must match or the union will not
occur. PROC SQL issues a warning message and stops executing.
The names of the columns in the output table are the names of the columns from the
rst table-expression unless a column (such as an expression) has no name in the rst
table-expression. In such a case, the name of that column in the output table is the
name of the respective column in the second table-expression.
In the following example, PROC SQL combines the two tables:
proc sql;
title ME1 and ME2: UNION;
select *
from me1
union
select *
from me2;

ME1 and ME2: UNION

IDnum
Jobcode
Salary
Bonus
-------------------------------------1120
ME1
28619
986
1244
ME2
36925
.
1400
ME1
29769
587
1403
ME1
28072
342
1653
ME2
35108
.
1782
ME2
35345
.

1098

query-expression

Chapter 44

In the following example, ALL includes the duplicate row from ME1. In addition,
ALL changes the sorting by specifying that PROC SQL make one pass only. Thus, the
values from ME2 are simply appended to the values from ME1.
proc sql;
title ME1 and ME2: UNION ALL;
select *
from me1
union all
select *
from me2;

ME1 and ME2: UNION ALL

IDnum
Jobcode
Salary
Bonus
-------------------------------------1400
ME1
29769
587
1403
ME1
28072
342
1120
ME1
28619
986
1120
ME1
28619
986
1653
ME2
35108
.
1782
ME2
35345
.
1244
ME2
36925
.

See Example 5 on page 1134 for another example.

EXCEPT
The EXCEPT operator produces (from the rst table-expression) an output table that
has unique rows that are not in the second table-expression. If the intermediate result
from the rst table-expression has at least one occurrence of a row that is not in the
intermediate result of the second table-expression, then that row (from the rst
table-expression) is included in the result table.
In the following example, the IN_USA table contains ights to cities within and
outside the USA. The OUT_USA table contains ights only to cities outside the USA.
This example returns only the rows from IN_USA that are not also in OUT_USA:
proc sql;
title Flights from IN_USA Only;
select * from in_usa
except
select * from out_usa;

IN_USA
Flight
Dest
-----------------145
ORD
156
WAS
188
LAX
193
FRA
207
LON

The SQL Procedure

sql-expression

1099

OUT_USA
Flight
Dest
-----------------193
FRA
207
LON
311
SJA

Flights from IN_USA Only

Flight
Dest
-----------------145
ORD
156
WAS
188
LAX

INTERSECT
The INTERSECT operator produces an output table that has rows that are common
to both tables. For example, using the IN_USA and OUT_USA tables shown above, the
following example returns rows that are in both tables:
proc sql;
title Flights from Both IN_USA and OUT_USA;
select * from in_usa
intersect
select * from out_usa;

Flights from Both IN_USA and OUT_USA

Flight
Dest
-----------------193
FRA
207
LON

sql-expression
Produces a value from a sequence of operands and operators.
operand operator operand

Arguments

operand

is one of the following:

3 a constant, which is a number or a quoted character string (or other special
notation) that indicates a xed value. Constants are also called literals.
Constants are described in SAS Language Reference: Dictionary.

1100

sql-expression

Chapter 44

3 a column-name, which is described in column-name on page 1078.

3 a CASE expression, which is described in CASE expression on page 1073.
3 a SAS function, which is any SAS function except LAG, DIF, and SOUND.
Functions are described in SAS Language Reference: Dictionary.

3 the ANSI SQL functions COALESCE, BTRIM, LOWER, UPPER, and

SUBSTRING.

3 a summary-function, which is described in summary-function on page 1107.

3 a query-expression, which is described in query-expression on page 1093.
3 the USER literal, which references the userid of the person who submitted the
program. The userid that is returned is operating environment-dependent, but
PROC SQL uses the same value that the &SYSJOBID macro variable has on
the operating environment.
operator

is described in Operators and the Order of Evaluation on page 1100.

Note: SAS functions, including summary functions, can stand alone as SQL
expressions. For example
select min(x) from table;
select scan(y,4) from table;

SAS Functions
PROC SQL supports the same SAS functions as the DATA step, except for the
functions LAG, DIF, and SOUND. For example, the SCAN function is used in the
following query:
select style, scan(street,1) format=$15.
from houses;

See SAS Language Reference: Dictionary for complete documentation on SAS

functions. Summary functions are also SAS functions. See summary-function on page
1107 for more information.

USER Literal
USER can be specied in a view denition, for example, to create a view that restricts
access to those in the users department. Note that the USER literal value is stored in
uppercase, so it is advisable to use the UPCASE function when comparing to this value:
create view myemp as
select * from dept12.employees
where upcase(manager)=user;

This view produces a different set of employee information for each manager who
references it.

Operators and the Order of Evaluation

The order in which operations are evaluated is the same as in the DATA step with
this one exception: NOT is grouped with the logical operators AND and OR in PROC
SQL; in the DATA step, NOT is grouped with the unary plus and minus signs.

The SQL Procedure

sql-expression

1101

Unlike missing values in some versions of SQL, missing values in SAS always appear
rst in the collating sequence. Therefore, in Boolean and comparison operations, the
following expressions resolve to true in a predicate:
3>null
-3>null
0>null

You can use parentheses to group values or to nest mathematical expressions.

Parentheses make expressions easier to read and can also be used to change the order
of evaluation of the operators. Evaluating expressions with parentheses begins at the
deepest level of parentheses and moves outward. For example, SAS evaluates A+B*C
as A+(B*C), although you can add parentheses to make it evaluate as (A+B)*C for a
different result.
Higher priority operations are performed rst: that is, group 0 operators are
evaluated before group 5 operators. The following table shows the operators and their
order of evaluation, including their priority groups.
Table 44.1

Operators and Order of Evaluation

Group

Operator

Description

()

forces the expression enclosed to be evaluated rst

case-expression

selects result values that satisfy specied conditions

raises to a power

unary +, unary -

indicates a positive or negative number

multiplies

divides

adds

subtracts

concatenates

<NOT> BETWEEN condition

See BETWEEN condition on page 1071.

<NOT> CONTAINS condition

see CONTAINS condition on page 1079.

<NOT> EXISTS condition

See EXISTS condition on page 1080.

<NOT> IN condition

See IN condition on page 1080.

IS <NOT> condition

See IS condition on page 1081.

<NOT> LIKE condition

See LIKE condition on page 1091.

=, eq

equals

=, ^=, < >, ne

does not equal

>, gt

is greater than

<, lt

is less than

>=, ge

is greater than or equal to

<=, le

is less than or equal to

sounds like (use with character operands only). See Example 11 on

page 1150.

1102

Group

sql-expression

Chapter 44

Operator

Description

eqt

equal to truncated strings (use with character operands only). See

Truncated String Comparison Operators on page 1102.

gtt

greater than truncated strings

ltt

less than truncated strings

get

greater than or equal to truncated strings

let

less than or equal to truncated strings

net

not equal to truncated strings

&, AND

indicates logical AND

|, OR

indicates logical OR

, ^, NOT

indicates logical NOT

Symbols for operators might vary, depending on your operating environment. See
SAS Language Reference: Dictionary for more information on operators and expressions.

Truncated String Comparison Operators

PROC SQL supports truncated string comparison operators (see Group 7 in Table
44.1 on page 1101). In a truncated string comparison, the comparison is performed
after making the strings the same length by truncating the longer string to be the same
length as the shorter string. For example, the expression TWOSTORY eqt TWO is
true because the string TWOSTORY is reduced to TWO before the comparison is
performed. Note that the truncation is performed internally; neither operand is
permanently changed.
Note: Unlike the DATA step, PROC SQL does not support the colon operators (such
as =:, >:, and <=:) for truncated string comparisons. Use the alphabetic operators (such
as EQT, GTT, and LET). 4

Query Expressions (Subqueries)

A query-expression is called a subquery when it is used in a WHERE or HAVING
clause. A subquery is a query-expression that is nested as part of another
query-expression. A subquery selects one or more rows from a table based on values in
another table.
Depending on the clause that contains it, a subquery can return a single value or
multiple values. If more than one subquery is used in a query-expression, then the
innermost query is evaluated rst, then the next innermost query, and so on, moving
outward.
PROC SQL allows a subquery (contained in parentheses) at any point in an
expression where a simple column value or constant can be used. In this case, a
subquery must return a single value, that is, one row with only one column.
The following is an example of a subquery that returns one value. This PROC SQL
step subsets the PROCLIB.PAYROLL table based on information in the
PROCLIB.STAFF table. (PROCLIB.PAYROLL is shown in Example 2 on page 1127,
and PROCLIB.STAFF is shown in Example 4 on page 1131.) PROCLIB.PAYROLL
contains employee identication numbers (IdNumber) and their salaries (Salary) but
does not contain their names. If you want to return only the row from
PROCLIB.PAYROLL for one employee, then you can use a subquery that queries the
PROCLIB.STAFF table, which contains the employees identication numbers and their
names (Lname and Fname).

The SQL Procedure

sql-expression

1103

options ls=64 nodate nonumber;

proc sql;
title Information for Earl Bowden;
select *
from proclib.payroll
where idnumber=
(select idnum
from proclib.staff
where upcase(lname)=BOWDEN);

Information for Earl Bowden

Id
Number Gender Jobcode
Salary
Birth
Hired
--------------------------------------------------1403
M
ME1
28072 28JAN69 21DEC91

Subqueries can return multiple values. The following example uses the tables
PROCLIB.DELAY and PROCLIB.MARCH. These tables contain information about the
same ights and have the Flight column in common. The following subquery returns all
the values for Flight in PROCLIB.DELAY for international ights. The values from the
subquery complete the WHERE clause in the outer query. Thus, when the outer query
is executed, only the international ights from PROCLIB.MARCH are in the output.
options ls=64 nodate nonumber;
proc sql outobs=5;
title International Flights from;
title2 PROCLIB.MARCH;
select Flight, Date, Dest, Boarded
from proclib.march
where flight in
(select flight
from proclib.delay
where destype=International);

International Flights from

PROCLIB.MARCH
Flight
Date Dest
Boarded
------------------------------219
01MAR94 LON
198
622
01MAR94 FRA
207
132
01MAR94 YYZ
115
271
01MAR94 PAR
138
219
02MAR94 LON
147

Sometimes it is helpful to compare a value with a set of values returned by a

subquery. The keywords ANY or ALL can be specied before a subquery when the
subquery is the right-hand operand of a comparison. If ALL is specied, then the
comparison is true only if it is true for all values that are returned by the subquery. If a

1104

sql-expression

Chapter 44

subquery returns no rows, then the result of an ALL comparison is true for each row of
the outer query.
If ANY is specied, then the comparison is true if it is true for any one of the values
that are returned by the subquery. If a subquery returns no rows, then the result of an
ANY comparison is false for each row of the outer query.
The following example selects all those in PROCLIB.PAYROLL who earn more than
the highest paid ME3:
options ls=64 nodate nonumber ;
proc sql;
title Employees who Earn More than;
title2 All MEs;
select *
from proclib.payroll
where salary > all (select salary
from proclib.payroll
where jobcode=ME3);

Employees who Earn More than

All MEs
Id
Number Gender Jobcode
Salary
Birth
Hired
--------------------------------------------------1333
M
PT2
88606 30MAR61 10FEB81
1739
M
PT1
66517 25DEC64 27JAN91
1428
F
PT1
68767 04APR60 16NOV91
1404
M
PT2
91376 24FEB53 01JAN80
1935
F
NA2
51081 28MAR54 16OCT81
1905
M
PT1
65111 16APR72 29MAY92
1407
M
PT1
68096 23MAR69 18MAR90
1410
M
PT2
84685 03MAY67 07NOV86
1439
F
PT1
70736 06MAR64 10SEP90
1545
M
PT1
66130 12AUG59 29MAY90
1106
M
PT2
89632 06NOV57 16AUG84
1442
F
PT2
84536 05SEP66 12APR88
1417
M
NA2
52270 27JUN64 07MAR89
1478
M
PT2
84203 09AUG59 24OCT90
1556
M
PT1
71349 22JUN64 11DEC91
1352
M
NA2
53798 02DEC60 16OCT86
1890
M
PT2
91908 20JUL51 25NOV79
1107
M
PT2
89977 09JUN54 10FEB79
1830
F
PT2
84471 27MAY57 29JAN83
1928
M
PT2
89858 16SEP54 13JUL90
1076
M
PT1
66558 14OCT55 03OCT91

Note: See the rst item in Subqueries and Efciency on page 1105 for a note
about efciency when using ALL. 4
In order to visually separate a subquery from the rest of the query, you can enclose
the subquery in any number of pairs of parentheses.

Correlated Subqueries
In a correlated subquery, the WHERE expression in a subquery refers to values in a
table in the outer query. The correlated subquery is evaluated for each row in the outer

The SQL Procedure

sql-expression

1105

query. With correlated subqueries, PROC SQL executes the subquery and the outer
query together.
The following example uses the PROCLIB.DELAY and PROCLIB.MARCH tables. A
DATA step (PROCLIB.DELAY on page 1390) creates PROCLIB.DELAY.
PROCLIB.MARCH is shown in Example 13 on page 1154. PROCLIB.DELAY has the
Flight, Date, Orig, and Dest columns in common with PROCLIB.MARCH:
proc sql outobs=5;
title International Flights;
select *
from proclib.march
where International in
(select destype
from proclib.delay
where march.Flight=delay.Flight);

The subquery resolves by substituting every value for MARCH.Flight into the
subquerys WHERE clause, one row at a time. For example, when MARCH.Flight=219,
the subquery resolves as follows:
1 PROC SQL retrieves all the rows from DELAY where Flight=219 and passes their

DESTYPE values to the WHERE clause.

2 PROC SQL uses the DESTYPE values to complete the WHERE clause:
where International in
(International,International, ...)

3 The WHERE clause checks to see if International is in the list. Because it is, all

rows from MARCH that have a value of 219 for Flight become part of the output.
The following output contains the rows from MARCH for international ights only.

Output 44.2

International Flights for March

International Flights

Flight
Date Depart Orig Dest
Miles
Boarded Capacity
----------------------------------------------------------------219
01MAR94
9:31 LGA
LON
3442
198
250
622
01MAR94
12:19 LGA
FRA
3857
207
250
132
01MAR94
15:35 LGA
YYZ
366
115
178
271
01MAR94
13:17 LGA
PAR
3635
138
250
219
02MAR94
9:31 LGA
LON
3442
147
250

Subqueries and Efciency

3 Use the MAX function in a subquery instead of the ALL keyword before the
subquery. For example, the following queries produce the same result, but the
second query is more efcient:
proc sql;
select * from proclib.payroll
where salary> all(select salary
from proclib.payroll
where jobcode=ME3);

1106

SUBSTRING function

Chapter 44

proc sql;
select * from proclib.payroll
where salary> (select max(salary)
from proclib.payroll
where jobcode=ME3);

3 With subqueries, use IN instead of EXISTS when possible. For example, the
following queries produce the same result, but the second query is usually more
efcient:
proc sql;
select *
from proclib.payroll p
where exists (select *
from staff s
where p.idnum=s.idnum
and state=CT);
proc sql;
select *
from proclib.payroll
where idnum in (select idnum
from staff
where state=CT);

SUBSTRING function
Returns a part of a character expression.
SUBSTRING (sql-expression FROM start <FOR length>)

3 sql-expression must be a character string and is described in sql-expression on

page 1099.

3 start is a number (not a variable or column name) that species the position,
counting from the left end of the character string, at which to begin extracting the
substring.

3 length is a number (not a variable or column name) that species the length of the
substring that is to be extracted.

Details
The SUBSTRING function operates on character strings. SUBSTRING returns a
specied part of the input character string, beginning at the position that is specied by
start. If length is omitted, then the SUBSTRING function returns all characters from
start to the end of the input character string. The values of start and length must be
numbers (not variables) and can be positive, negative, or zero.
If start is greater than the length of the input character string, then the
SUBSTRING function returns a zero-length string.
If start is less than 1, then the SUBSTRING function begins extraction at the
beginning of the input character string.
If length is specied, then the sum of start and length cannot be less than start or an
error is returned. If the sum of start and length is greater than the length of the input

The SQL Procedure

summary-function

1107

character string, then the SUBSTRING function returns all characters from start to the
end of the input character string. If the sum of start and length is less than 1, then the
SUBSTRING function returns a zero-length string.
Note: The SUBSTRING function is provided for compatibility with the ANSI SQL
standard. You can also use the SAS function SUBSTR. 4

summary-function
Performs statistical summary calculations.
Restriction:

A summary function cannot appear in an ON clause or a WHERE clause.

See also: GROUP BY on page 1065, HAVING Clause on page 1067, SELECT Clause on
page 1058, and table-expression on page 1113
Featured in: Example 8 on page 1143, Example 12 on page 1152, and Example 15 on

page 1160

summary-function (<DISTINCT | ALL> sql-expression)

Arguments
summary-function

is one of the following:

AVG|MEAN
arithmetic mean or average of values
COUNT|FREQ|N
number of nonmissing values
CSS
corrected sum of squares
CV
coefcient of variation (percent)
MAX
largest value
MIN
smallest value
NMISS
number of missing values
PRT
probability of a greater absolute value of Students t
RANGE
range of values

1108

summary-function

Chapter 44

STD
standard deviation
STDERR
standard error of the mean
SUM
sum of values
SUMWGT
sum of the WEIGHT variable values*
T
Students t value for testing the hypothesis that the population mean is zero
USS
uncorrected sum of squares
VAR
variance
For a description and the formulas used for these statistics, see Appendix 1, SAS
Elementary Statistics Procedures, on page 1339.
DISTINCT

species that only the unique values of sql-expression be used in the calculation.
ALL

species that all values of sql-expression be used in the calculation. If neither

DISTINCT nor ALL is specied, then ALL is used.
sql-expression

is described in sql-expression on page 1099.

Summarizing Data
Summary functions produce a statistical summary of the entire table or view that is
listed in the FROM clause or for each group that is specied in a GROUP BY clause. If
GROUP BY is omitted, then all the rows in the table or view are considered to be a
single group. These functions reduce all the values in each row or column in a table to
one summarizing or aggregate value. For this reason, these functions are often called
aggregate functions. For example, the sum (one value) of a column results from the
addition of all the values in the column.

Counting Rows
The COUNT function counts rows. COUNT(*) returns the total number of rows in a
group or in a table. If you use a column name as an argument to COUNT, then the
result is the total number of rows in a group or in a table that have a nonmissing value
for that column. If you want to count the unique values in a column, then specify
COUNT(DISTINCT column).

* Currently, there is no way to designate a WEIGHT variable for a table in PROC SQL. Thus, each row (or observation) has a
weight of 1.

The SQL Procedure

summary-function

1109

If the SELECT clause of a table-expression contains one or more summary functions

and that table-expression resolves to no rows, then the summary function results are
missing values. The following are exceptions that return zeros:
COUNT(*)
COUNT(<DISTINCT> sql-expression)
NMISS(<DISTINCT> sql-expression)
See Example 8 on page 1143 and Example 15 on page 1160 for examples.

Calculating Statistics Based on the Number of Arguments

The number of arguments that is specied in a summary function affects how the
calculation is performed. If you specify a single argument, then the values in the
column are calculated. If you specify multiple arguments, then the arguments or
columns that are listed are calculated for each row. For example, consider calculations
on the following table.
proc sql;
title Summary Table;
select * from summary;

Summary Table
X
Y
Z
---------------------------1
3
4
2
4
5
8
9
4
4
5
4

If you use one argument in the function, then the calculation is performed on that
column only. If you use more than one argument, then the calculation is performed on
each row of the specied columns. In the following PROC SQL step, the MIN and MAX
functions return the minimum and maximum of the columns they are used with. The
SUM function returns the sum of each row of the columns specied as arguments:
proc sql;
select min(x) as Colmin_x,
min(y) as Colmin_y,
max(z) as Colmax_z,
sum(x,y,z) as Rowsum
from summary;

Summary Table
Colmin_x Colmin_y Colmax_z
Rowsum
-------------------------------------1
3
5
8
1
3
5
11
1
3
5
21
1
3
5
13

1110

summary-function

Chapter 44

Remerging Data
When you use a summary function in a SELECT clause or a HAVING clause, you
might see the following message in the SAS log:
NOTE: The query requires remerging summary
statistics back with the original
data.

The process of remerging involves two passes through the data. On the rst pass,
PROC SQL

3 calculates and returns the value of summary functions. It then uses the result to
calculate the arithmetic expressions in which the summary function participates.

3 groups data according to the GROUP BY clause.

On the second pass, PROC SQL retrieves any additional columns and rows that it
needs to show in the output.
The following examples use the PROCLIB.PAYROLL table (shown in Example 2 on
page 1127) to show when remerging of data is and is not necessary.
The rst query requires remerging. The rst pass through the data groups the data
by Jobcode and resolves the AVG function for each group. However, PROC SQL must
make a second pass in order to retrieve the values of IdNumber and Salary.
proc sql outobs=10;
title Salary Information;
title2 (First 10 Rows Only);
select IdNumber, Jobcode, Salary,
avg(salary) as AvgSalary
from proclib.payroll
group by jobcode;

Salary Information
(First 10 Rows Only)
Id
Number Jobcode
Salary AvgSalary
-----------------------------------1704
BCK
25465
25794.22
1677
BCK
26007
25794.22
1383
BCK
25823
25794.22
1845
BCK
25996
25794.22
1100
BCK
25004
25794.22
1663
BCK
26452
25794.22
1673
BCK
25477
25794.22
1389
BCK
25028
25794.22
1834
BCK
26896
25794.22
1132
FA1
22413
23039.36

You can change the previous query to return only the average salary for each
jobcode. The following query does not require remerging because the rst pass of the
data does the summarizing and the grouping. A second pass is not necessary.
proc sql outobs=10;
title Average Salary for Each Jobcode;
select Jobcode, avg(salary) as AvgSalary
from proclib.payroll
group by jobcode;

The SQL Procedure

summary-function

1111

Average Salary for Each Jobcode

Jobcode AvgSalary
-----------------BCK
25794.22
FA1
23039.36
FA2
27986.88
FA3
32933.86
ME1
28500.25
ME2
35576.86
ME3
42410.71
NA1
42032.2
NA2
52383
PT1
67908

When you use the HAVING clause, PROC SQL may have to remerge data to resolve
the HAVING expression.
First, consider a query that uses HAVING but that does not require remerging. The
query groups the data by values of Jobcode, and the result contains one row for each
value of Jobcode and summary information for people in each Jobcode. On the rst
pass, the summary functions provide values for the Number, Average Age, and Average
Salary columns. The rst pass provides everything that PROC SQL needs to resolve
the HAVING clause, so no remerging is necessary.
proc sql outobs=10;
title Summary Information for Each Jobcode;
title2 (First 10 Rows Only);
select Jobcode,
count(jobcode) as number
label=Number,
avg(int((today()-birth)/365.25))
as avgage format=2.
label=Average Age,
avg(salary) as avgsal format=dollar8.
label=Average Salary
from proclib.payroll
group by jobcode
having avgage ge 30;

Summary Information for Each Jobcode

(First 10 Rows Only)
Average
Average
Jobcode
Number
Age
Salary
-----------------------------------BCK
9
36
$25,794
FA1
11
33
$23,039
FA2
16
37
$27,987
FA3
7
39
$32,934
ME1
8
34
$28,500
ME2
14
39
$35,577
ME3
7
42
$42,411
NA1
5
30
$42,032
NA2
3
42
$52,383
PT1
8
38
$67,908

1112

summary-function

Chapter 44

In the following query, PROC SQL remerges the data because the HAVING clause
uses the SALARY column in the comparison and SALARY is not in the GROUP BY
clause.
proc sql outobs=10;
title Employees who Earn More than the;
title2 Average for Their Jobcode;
title3 (First 10 Rows Only);
select Jobcode, Salary,
avg(salary) as AvgSalary
from proclib.payroll
group by jobcode
having salary > AvgSalary;

Employees who Earn More than the

Average for Their Jobcode
(First 10 Rows Only)
Jobcode
Salary AvgSalary
---------------------------BCK
26007
25794.22
BCK
25823
25794.22
BCK
25996
25794.22
BCK
26452
25794.22
BCK
26896
25794.22
FA1
23177
23039.36
FA1
23738
23039.36
FA1
23979
23039.36
FA1
23916
23039.36
FA1
23644
23039.36

Keep in mind that PROC SQL remerges data when

3 the values returned by a summary function are used in a calculation. For

example, the following query returns the values of X and the percent of the total
for each row. On the rst pass, PROC SQL computes the sum of X, and on the
second pass PROC SQL computes the percentage of the total for each value of X:
proc sql;
title Percentage of the Total;
select X, (100*x/sum(X)) as Pct_Total
from summary;
Percentage of the Total
x Pct_Total
------------------32
14.81481
86
39.81481
49
22.68519
49
22.68519

3 the values returned by a summary function are compared to values of a column

that is not specied in the GROUP BY clause. For example, the following query
uses the PROCLIB.PAYROLL table. PROC SQL remerges data because the
column Salary is not specied in the GROUP BY clause:

The SQL Procedure

table-expression

1113

proc sql;
select

jobcode, salary,
avg(salary) as avsal
from proclib.payroll
group by jobcode
having salary > avsal;

3 a column from the input table is specied in the SELECT clause and is not
specied in the GROUP BY clause. This rule does not refer to columns used as
arguments to summary functions in the SELECT clause.
For example, in the following query, the presence of IdNumber in the SELECT
clause causes PROC SQL to remerge the data because IdNumber is not involved in
grouping or summarizing during the rst pass. In order for PROC SQL to retrieve
the values for IdNumber, it must make a second pass through the data.
proc sql;
select IdNumber, jobcode,
avg(salary) as avsal
from proclib.payroll
group by jobcode;

table-expression
Denes part or all of a query-expression.
See also: query-expression on page 1093

SELECT <DISTINCT> object-item<, object-item>

<INTO :macro-variable-specication
<, :macro-variable-specication>>
FROM from-list
<WHERE sql-expression>
<GROUP BY group-by-item < , group-by-item>>
<HAVING sql-expression>
See SELECT Statement on page 1058 for complete information on the SELECT
statement.

Details
A table-expression is a SELECT statement. It is the fundamental building block of
most SQL procedure statements. You can combine the results of multiple
table-expressions with set operators, which creates a query-expression. Use one
ORDER BY clause for an entire query-expression. Place a semicolon only at the end of
the entire query-expression. A query-expression is often only one SELECT statement or
table-expression.

1114

UPPER function

Chapter 44

UPPER function
Converts the case of a character string to uppercase.
See also: LOWER function on page 1093

UPPER (sql-expression)

3 sql-expression must be a character string and is described in sql-expression on

page 1099.

Details
The UPPER function operates on character strings. UPPER converts the case of its
argument to all uppercase.

Concepts: SQL Procedure

Using SAS Data Set Options with PROC SQL

In PROC SQL, you can apply most of the SAS data set options, such as KEEP= and
DROP=, to tables or SAS/ACCESS views any time that you specify a table or
SAS/ACCESS view. In the SQL procedure, SAS data set options that are separated by
spaces are enclosed in parentheses, and they follow immediately after the table or
SAS/ACCESS view name. In the following PROC SQL step, RENAME= renames
LNAME to LASTNAME for the STAFF1 table. OBS= restricts the number of rows
written to STAFF1 to 15:
proc sql;
create table
staff1(rename=(lname=lastname)) as
select *
from staff(obs=15);

SAS data set options can be combined with SQL statement arguments:
proc sql;
create table test
(a character, b numeric, pw=cat);
create index staffidx on
staff1 (lastname, alter=dog);

You cannot use SAS data set options with DICTIONARY tables because
DICTIONARY tables are read-only objects.
The only SAS data set options that you can use with PROC SQL views are those that
assign and provide SAS passwords: READ=, WRITE=, ALTER=, and PW=.
See SAS Language Reference: Dictionary for a description of SAS data set options.

The SQL Procedure

Connecting to a DBMS Using the LIBNAME Statement

1115

Connecting to a DBMS Using the SQL Procedure Pass-Through Facility

What Is the Pass-Through Facility?
The SQL Procedure Pass-Through Facility enables you to send DBMS-specic SQL
statements directly to a DBMS for execution. The Pass-Through Facility uses a SAS/
ACCESS interface engine to connect to the DBMS. Therefore, you must have SAS/
ACCESS software installed for your DBMS.
You submit SQL statements that are DBMS-specic. For example, you pass
Transact-SQL statements to a SYBASE database. The Pass-Through Facilitys basic
syntax is the same for all the DBMSs. Only the statements that are used to connect to
the DBMS and the SQL statements are DBMS-specic.
With the Pass-Through Facility, you can perform the following tasks:
3 establish a connection with the DBMS using a CONNECT statement and
terminate the connection with the DISCONNECT statement.
3 send nonquery DBMS-specic SQL statements to the DBMS using the EXECUTE
statement.
3 retrieve data from the DBMS to be used in a PROC SQL query with the
CONNECTION TO component in a SELECT statements FROM clause.
You can use the Pass-Through Facility statements in a query, or you can store them
in a PROC SQL view. When a view is stored, any options that are specied in the
corresponding CONNECT statement are also stored. Thus, when the PROC SQL view
is used in a SAS program, SAS can automatically establish the appropriate connection
to the DBMS.
See CONNECT Statement on page 1042, DISCONNECT Statement on page 1053,
EXECUTE Statement on page 1055, CONNECTION TO on page 1079, and The
Pass-Through Facility for Relational Databases in SAS/ACCESS for Relational
Databases: Reference.
Note: SAS procedures that do multipass processing cannot operate on PROC SQL
views that store Pass-Through Facility statements, because the Pass-Through Facility
does not allow reopening of a table after the rst record has been retrieved. To work
around this limitation, create a SAS data set from the view and use the SAS data set as
the input data set. 4

Return Codes
As you use PROC SQL statements that are available in the Pass-Through Facility,
any errors are written to the SAS log. The return codes and messages that are
generated by the Pass-Through Facility are available to you through the SQLXRC and
SQLXMSG macro variables. Both macro variables are described in Using Macro
Variables Set by PROC SQL on page 1119.

Connecting to a DBMS Using the LIBNAME Statement

For many DBMSs, you can directly access DBMS data by assigning a libref to the
DBMS using the SAS/ACCESS LIBNAME statement. Once you have associated a libref
with the DBMS, you can specify a DBMS table in a two-level SAS name and work with
the table like any SAS data set. You can also embed the LIBNAME statement in a
PROC SQL view (see CREATE VIEW Statement on page 1049).
PROC SQL will take advantage of the capabilities of a DBMS by passing it certain
operations whenever possible. For example, before implementing a join, PROC SQL

1116

Using the DICTIONARY Tables

Chapter 44

checks to see if the DBMS can do the join. If it can, then PROC SQL passes the join to
the DBMS. This enhances performance by reducing data movement and translation. If
the DBMS cannot do the join, then PROC SQL processes the join. Using the
SAS/ACCESS LIBNAME statement can often provide you with the performance
benets of the SQL Procedure Pass-Through Facility without having to write
DBMS-specic code.
To use the SAS/ACCESS LIBNAME statement, you must have SAS/ACCESS
software installed for your DBMS. For more information about the SAS/ACCESS
LIBNAME statement, refer to the SAS/ACCESS documentation for your DBMS.

Using the DICTIONARY Tables

What Are DICTIONARY Tables?

DICTIONARY tables are special, read-only SAS data views that contain information
about your SAS session. For example, the DICTIONARY.COLUMNS table contains
information, such as name, type, length, and format, about all columns in all tables
that are known to the current SAS session. DICTIONARY tables are accessed by using
the libref DICTIONARY in the FROM clause in a SELECT statement in PROC SQL.
Additionally, there are PROC SQL views, stored in the SASHELP library and known as
SASHELP views, that reference the DICTIONARY tables and that can be used in other
SAS procedures and in the DATA step.
Note:

You cannot use data set options with DICTIONARY tables.

For an example that demonstrates the use of a DICTIONARY table, see Example 6
on page 1136.
The following table describes the DICTIONARY tables that are available and shows
the associated SASHELP view(s) for each table.
Table 44.2

DICTIONARY Tables and Associated SASHELP Views

DICTIONARY table

SASHELP
view

Description

CATALOGS

VCATALG

Contains information about known SAS

catalogs.

CHECK_CONSTRAINTS

VCHKCON

Contains information about known check

constraints.

COLUMNS

VCOLUMN

Contains information about columns in all

known tables.

CONSTRAINT_COLUMN_USAGE

VCNCOLU

Contains information about columns that are

referred to by integrity constraints.

CONSTRAINT_TABLE_USAGE

VCNTABU

Contains information about tables that have

integrity constraints dened on them.

DICTIONARIES

VDCTNRY

Contains information about all DICTIONARY

tables.

ENGINES

VENGINE

Contains information about SAS engines.

The SQL Procedure

Using the DICTIONARY Tables

DICTIONARY table

SASHELP
view

Description

EXTFILES

VEXTFL

Contains information about known external

les.

FORMATS

VFORMAT

Contains information about currently

accessible formats and informats.

GOPTIONS

VGOPT
VALLOPT

Contains information about currently dened

graphics options (SAS/GRAPH software).
SASHELP.VALLOPT includes SAS system
options as well as graphics options.

INDEXES

VINDEX

Contains information about known indexes.

LIBNAMES

VLIBNAM

Contains information about currently dened

SAS data libraries.

MACROS

VMACRO

Contains information about currently dened

macros.

MEMBERS

VMEMBER

1117

Contains information about all objects that

are in currently dened SAS data libraries.
SASHELP.VMEMBER contains information
for all member types; the other SASHELP
views are specic to particular member types
(such as tables or views).

VSACCES
VSCATLG
VSLIB
VSTABLE
VSTABVW
VSVIEW
OPTIONS

VOPTION
VALLOPT

Contains information on SAS system options.

SASHELP.VALLOPT includes graphics
options as well as SAS system options.

REFERENTIAL_CONSTRAINTS

VREFCON

Contains information about referential

constraints.

STYLES

VSTYLE

Contains information about known ODS styles.

TABLE_CONSTRAINTS

VTABCON

Contains information about integrity

constraints in all known tables.

TABLES

VTABLE

Contains information about known tables.

TITLES

VTITLE

Contains information about currently dened

titles and footnotes.

VIEWS

VVIEW

Contains information about known data views.

Retrieving Information about DICTIONARY Tables and SASHELP Views

To see how each DICTIONARY table is dened, submit a DESCRIBE TABLE
statement. After you know how a table is dened, you can use its column names in a
subsetting WHERE clause in order to retrieve more specic information. For example:
proc sql;
describe table dictionary.indexes;

1118

Using the DICTIONARY Tables

Chapter 44

The results are written to the SAS log:

6
proc sql;
7
describe table dictionary.indexes;
NOTE: SQL table DICTIONARY.INDEXES was created like:
create table DICTIONARY.INDEXES
(
libname char(8) label=Library Name,
memname char(32) label=Member Name,
memtype char(8) label=Member Type,
name char(32) label=Column Name,
idxusage char(9) label=Column Index Type,
indxname char(32) label=Index Name,
indxpos num label=Position of Column in Concatenated Key,
nomiss char(3) label=Nomiss Option,
unique char(3) label=Unique Option
);

Use the DESCRIBE VIEW statement in PROC SQL to nd out how a SASHELP
view is dened. Heres an example:
proc sql;
describe view sashelp.vstabvw;

The results are written to the SAS log:

6
proc sql;
7
describe view sashelp.vstabvw;
NOTE: SQL view SASHELP.VSTABVW is defined as:
select
from
where
order by

libname, memname, memtype

DICTIONARY.MEMBERS
(memtype=VIEW) or (memtype=DATA)
libname asc, memname asc;

Using DICTIONARY Tables

DICTIONARY tables are commonly used to monitor and manage SAS sessions
because the data is more easily manipulated than the output from, for example, PROC
DATASETS. You can query DICTIONARY tables the same way that you query any other
table, including subsetting with a WHERE clause, ordering the results, and creating
PROC SQL views. Note that many character values in the DICTIONARY tables are
stored as all-uppercase characters; you should design your queries accordingly.
Because DICTIONARY tables are read-only objects, you cannot insert rows or
columns, alter column attributes, or add integrity constraints to them.
Note: For DICTIONARY.TABLES and SASHELP.VTABLE, if a table is
read-protected with a password, then the only information that is listed for that table is
the library name, member name, member type, and type of password protection; all
other information is set to missing. 4

DICTIONARY Tables and Performance

When querying a DICTIONARY table, SAS launches a discovery process that gathers
information that is pertinent to that table. Depending on the DICTIONARY table that
is being queried, this discovery process can search libraries, open tables, and execute
views. Unlike other SAS procedures and the DATA step, PROC SQL can mitigate this

The SQL Procedure

Using Macro Variables Set by PROC SQL

1119

process by optimizing the query before the discovery process is launched. Therefore,
although it is possible to access DICTIONARY table information with SAS procedures
or the DATA step by using the SASHELP views, it is often more efcient to use PROC
SQL instead.
For example, the following programs both produce the same result, but the PROC
SQL step runs much faster because the WHERE clause is processed prior to opening
the tables that are referenced by the SASHELP.VCOLUMN view:
data mytable;
set sashelp.vcolumn;
where libname=WORK and memname=SALES;
run;

proc sql;
create table mytable as
select * from sashelp.vcolumn
where libname=WORK and memname=SALES;
quit;

Note: SAS does not maintain DICTIONARY table information between queries.
Each query of a DICTIONARY table launches a new discovery process. 4
If you are querying the same DICTIONARY table several times in a row, then you
can get even faster performance by creating a temporary SAS data set (with the DATA
step SET statement or PROC SQL CREATE TABLE AS statement) with the
information that you want and running your query against that data set.

Using Macro Variables Set by PROC SQL

PROC SQL sets up macro variables with certain values after it executes each
statement. These macro variables can be tested inside a macro to determine whether to
continue executing the PROC SQL step. SAS/AF software users can also test them in a
program after an SQL SUBMIT block of code, using the SYMGET function.
After each PROC SQL statement has executed, the following macro variables are
updated with these values:
SQLOBS
contains the number of rows executed by an SQL procedure statement. For
example, it contains the number of rows formatted and displayed in SAS output by
a SELECT statement or the number of rows deleted by a DELETE statement.
When the NOPRINT option is specied, the value of the SQLOBS macro
variable depends on whether an output table, single macro variable, macro
variable range, or macro variable list is created:

3 If no output table, macro variable list, or macro variable range is created,

then SQLOBS contains the value 1.

3 If an output table is created, then SQLOBS contains the number of rows in

the output table.

3 If a single macro variable is created, then SQLOBS contains the value 1.

3 If a macro variable list or macro variable range is created, then SQLOBS
contains the number of rows that are processed to create the macro variable
list or range.

1120

Using Macro Variables Set by PROC SQL

Chapter 44

SQLRC
contains the following status values that indicate the success of the SQL procedure
statement:
0
PROC SQL statement completed successfully with no errors.
4
PROC SQL statement encountered a situation for which it issued a warning.
The statement continued to execute.
8
PROC SQL statement encountered an error. The statement stopped
execution at this point.
12
PROC SQL statement encountered an internal error, indicating a bug in
PROC SQL that should be reported to SAS Technical Support. These errors
can occur only during compile time.
16
PROC SQL statement encountered a user error. This error code is used, for
example, when a subquery (that can only return a single value) evaluates to
more than one row. These errors can only be detected during run time.
24
PROC SQL statement encountered a system error. This error is used, for
example, if the system cannot write to a PROC SQL table because the disk is
full. These errors can occur only during run time.
28
PROC SQL statement encountered an internal error, indicating a bug in
PROC SQL that should be reported to SAS Technical Support. These errors
can occur only during run time.
SQLOOPS
contains the number of iterations that the inner loop of PROC SQL executes. The
number of iterations increases proportionally with the complexity of the query. See
also the description of LOOPS= on page 1036.
SQLXRC
contains the DBMS-specic return code that is returned by the Pass-Through
Facility.
SQLXMSG
contains descriptive information and the DBMS-specic return code for the error
that is returned by the Pass-Through Facility.
Note: Because the value of the SQLXMSG macro variable can contain special
characters (such as &, %, /, *, and ;), use the %SUPERQ macro function when
printing the value:
%put %superq(sqlxmsg);

See SAS Macro Language: Reference for information about the %SUPERQ
function. 4
This example retrieves the data but does not display them in SAS output because of
the NOPRINT option in the PROC SQL statement. The %PUT macro statement
displays the macro variables values.
proc sql noprint;
select *

The SQL Procedure

Updating PROC SQL and SAS/ACCESS Views

1121

from proclib.payroll;
%put sqlobs=**&sqlobs**
sqloops=**&sqloops**
sqlrc=**&sqlrc**;

The message in Output 44.3 appears in the SAS log and gives you the macros values.

Output 44.3

PROC SQL Macro Variable Values

40
options ls=80;
41
42
proc sql noprint;
43
select *
44
from proclib.payroll;
45
46
%put sqlobs=**&sqlobs**
47
sqloops=**&sqloops**
48
sqlrc=**&sqlrc**;
sqlobs=**1**
sqloops=**11**

sqlrc=**0**

Macro variables that are generated by PROC SQL follow the scoping rules for %LET.
For more information about macro variable scoping, see SAS Macro Language:
Reference.

Updating PROC SQL and SAS/ACCESS Views

You can update PROC SQL and SAS/ACCESS views using the INSERT, DELETE,
and UPDATE statements, under the following conditions.

3 If the view accesses a DBMS table, then you must have been granted the
appropriate authorization by the external database management system (for
example, DB2). You must have installed the SAS/ACCESS software for your
DBMS. See the SAS/ACCESS interface guide for your DBMS for more information
on SAS/ACCESS views.
3 You can update only a single table through a view. The table cannot be joined to
another table or linked to another table with a set-operator. The view cannot
contain a subquery.
3 You can update a column in a view using the columns alias, but you cannot
update a derived column, that is, a column produced by an expression. In the
following example, you can update the column SS, but not WeeklySalary.
create view EmployeeSalaries as
select Employee, SSNumber as SS,
Salary/52 as WeeklySalary
from employees;

3 You cannot update a view containing an ORDER BY.

Note: Starting in SAS System 9, PROC SQL views, the Pass-Through Facility, and
the SAS/ACCESS LIBNAME statement are the preferred ways to access relational
DBMS data; SAS/ACCESS views are no longer recommended. You can convert existing
SAS/ACCESS views to PROC SQL views by using the CV2VIEW procedure. See The
CV2VIEW Procedure in SAS/ACCESS for Relational Databases: Reference for more
information. 4

1122

PROC SQL and the ANSI Standard

Chapter 44

PROC SQL and the ANSI Standard

Compliance
PROC SQL follows most of the guidelines set by the American National Standards
Institute (ANSI) in its implementation of SQL. However, it is not fully compliant with
the current ANSI Standard for SQL.*
The SQL research project at SAS has focused primarily on the expressive power of
SQL as a query language. Consequently, some of the database features of SQL have not
yet been implemented in PROC SQL.

SQL Procedure Enhancements

Reserved Words
PROC SQL reserves very few keywords and then only in certain contexts. The ANSI
Standard reserves all SQL keywords in all contexts. For example, according to the
Standard you cannot name a column GROUP because of the keywords GROUP BY.
The following words are reserved in PROC SQL:

3 The keyword CASE is always reserved; its use in the CASE expression (an SQL2
feature) precludes its use as a column name.
If you have a column named CASE in a table and you want to specify it in a
PROC SQL step, then you can use the SAS data set option RENAME= to rename
that column for the duration of the query. You can also surround CASE in double
quotation marks (CASE) and set the PROC SQL option DQUOTE=ANSI.

3 The keywords AS, ON, FULL, JOIN, LEFT, FROM, WHEN, WHERE, ORDER,
GROUP, RIGHT, INNER, OUTER, UNION, EXCEPT, HAVING, and INTERSECT
cannot normally be used for table aliases. These keywords all introduce clauses
that appear after a table name. Since the alias is optional, PROC SQL deals with
this ambiguity by assuming that any one of these words introduces the
corresponding clause and is not the alias. If you want to use one of these keywords
as an alias, then use the PROC SQL option DQUOTE=ANSI.

3 The keyword USER is reserved for the current userid. If you specify USER on a
SELECT statement in conjunction with a CREATE TABLE statement, then the
column is created in the table with a temporary column name that is similar to
_TEMA001. If you specify USER in a SELECT statement without using the
CREATE TABLE statement, then the column is written to the output without a
column heading. In either case, the value for the column varies by operating
environment, but is typically the userid of the user who is submitting the program
or the value of the &SYSJOBID automatic macro variable.
If you have a column named USER in a table and you want to specify it in a
PROC SQL step, then you can use the SAS data set option RENAME= to rename
that column for the duration of the query. You can also enclose USER with double
quotation marks (USER) and set the PROC SQL option DQUOTE=ANSI.
* International Organization for Standardization (ISO): Database SQL. Document ISO/IEC 9075:1992. Also available as
American National Standards Institute (ANSI) Document ANSI X3.135-1992.

The SQL Procedure

SQL Procedure Enhancements

1123

Column Modiers
PROC SQL supports the SAS INFORMAT=, FORMAT=, and LABEL= modiers for
expressions within the SELECT clause. These modiers control the format in which
output data are displayed and labeled.

Alternate Collating Sequences

PROC SQL allows you to specify an alternate collating (sorting) sequence to be used
when you specify the ORDER BY clause. See the description of the SORTSEQ= option
in PROC SQL Statement on page 1033 for more information.

ORDER BY Clause in a View Denition

PROC SQL permits you to specify an ORDER BY clause in a CREATE VIEW
statement. When the view is queried, its data are always sorted according to the
specied order unless a query against that view includes a different ORDER BY clause.
See CREATE VIEW Statement on page 1049 for more information.

In-Line Views
The ability to code nested query-expressions in the FROM clause is a requirement of
the ANSI Standard. PROC SQL supports such nested coding.

Outer Joins
The ability to include columns that both match and do not match in a join-expression
is a requirement of the ANSI Standard. PROC SQL supports this ability.

Arithmetic Operators
PROC SQL supports the SAS exponentiation (**) operator. PROC SQL uses the
notation <> to mean not equal.

Orthogonal Expressions
PROC SQL permits the combination of comparison, Boolean, and algebraic
expressions. For example, (X=3)*7 yields a value of 7 if X=3 is true because true is
dened to be 1. If X=3 is false, then it resolves to 0 and the entire expression yields a
value of 0.
PROC SQL permits a subquery in any expression. This feature is required by the
ANSI Standard. Therefore, you can have a subquery on the left side of a comparison
operator in the WHERE expression.
PROC SQL permits you to order and group data by any kind of mathematical
expression (except those including summary functions) using ORDER BY and GROUP
BY clauses. You can also group by an expression that appears on the SELECT clause
by using the integer that represents the expressions ordinal position in the SELECT
clause. You are not required to select the expression by which you are grouping or
ordering. See ORDER BY Clause on page 1067 and GROUP BY Clause on page 1065
for more information.

Set Operators
The set operators UNION, INTERSECT, and EXCEPT are required by the ANSI
Standard. PROC SQL provides these operators plus the OUTER UNION operator.

1124

SQL Procedure Omissions

Chapter 44

The ANSI Standard also requires that the tables being operated upon all have the
same number of columns with matching data types. The SQL procedure works on
tables that have the same number of columns, as well as on those that do not, by
creating virtual columns so that a query can evaluate correctly. See query-expression
on page 1093 for more information.

Statistical Functions
PROC SQL supports many more summary functions than required by the ANSI
Standard for SQL.
PROC SQL supports the remerging of summary function results into the tables
original data. For example, computing the percentage of total is achieved with 100*x/
SUM(x) in PROC SQL. See summary-function on page 1107 for more information on
the available summary functions and remerging data.

SAS DATA Step Functions

PROC SQL supports all the functions available to the SAS DATA step, except for
LAG, DIF, and SOUND. Other SQL databases support their own set of functions.

SQL Procedure Omissions

COMMIT Statement
The COMMIT statement is not supported.

ROLLBACK Statement
The ROLLBACK statement is not supported. The UNDO_POLICY= option in the
PROC SQL statement addresses rollback. See the description of the UNDO_POLICY=
option in PROC SQL Statement on page 1033 for more information.

Identiers and Naming Conventions

In SAS, table names, column names, and aliases are limited to 32 characters and can
contain mixed case. For more information on SAS naming conventions, see SAS
Language Reference: Dictionary. The ANSI Standard for SQL allows longer names.

Granting User Privileges

The GRANT statement, PRIVILEGES keyword, and authorization-identier features
of SQL are not supported. You might want to use operating environment-specic means
of security instead.

Three-Valued Logic
ANSI-compatible SQL has three-valued logic, that is, special cases for handling
comparisons involving NULL values. Any value compared with a NULL value
evaluates to NULL.
PROC SQL follows the SAS convention for handling missing values: when numeric
NULL values are compared to non-NULL numbers, the NULL values are less than or
smaller than all the non-NULL values; when character NULL values are compared to
non-NULL characters, the character NULL values are treated as a string of blanks.

The SQL Procedure

Program

1125

Embedded SQL
Currently there is no provision for embedding PROC SQL statements in other SAS
programming environments, such as the DATA step or SAS/IML software.

Examples: SQL Procedure

Example 1: Creating a Table and Inserting Data into It

Procedure features:

CREATE TABLE statement

column-modier
INSERT statement
VALUES clause
SELECT clause
FROM clause
Table: PROCLIB.PAYLIST

This example creates the table PROCLIB.PAYLIST and inserts data into it.

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Create the PROCLIB.PAYLIST table. The CREATE TABLE statement creates

PROCLIB.PAYLIST with six empty columns. Each column denition indicates whether the
column is character or numeric. The number in parentheses species the width of the column.
INFORMAT= and FORMAT= assign date informats and formats to the Birth and Hired columns.
proc sql;
create table proclib.paylist
(IdNum char(4),

1126

Program

Chapter 44

Gender char(1),
Jobcode char(3),
Salary num,
Birth num informat=date7.
format=date7.,
Hired num informat=date7.
format=date7.);

Insert values into the PROCLIB.PAYLIST table. The INSERT statement inserts data
values into PROCLIB.PAYLIST according to the position in the VALUES clause. Therefore, in
the rst VALUES clause, 1639 is inserted into the rst column, F into the second column, and
so forth. Dates in SAS are stored as integers with 0 equal to January 1, 1960. Sufxing the date
with a d is one way to use the internal value for dates.
insert into proclib.paylist
values(1639,F,TA1,42260,26JUN70d,28JAN91d)
values(1065,M,ME3,38090,26JAN54d,07JAN92d)
values(1400,M,ME1,29769.05NOV67d,16OCT90d)

Include missing values in the data. The value null represents a missing value for the
character column Jobcode. The period represents a missing value for the numeric column Salary.
values(1561,M,null,36514,30NOV63d,07OCT87d)
values(1221,F,FA3,.,22SEP63d,04OCT94d);

Specify the title.

title PROCLIB.PAYLIST Table;

Display the entire PROCLIB.PAYLIST table. The SELECT clause selects columns from
PROCLIB.PAYLIST. The asterisk (*) selects all columns. The FROM clause species
PROCLIB.PAYLIST as the table to select from.
select *
from proclib.paylist;

The SQL Procedure

Example 2: Creating a Table from a Querys Result

Output Table

PROCLIB.PAYLIST

PROCLIB.PAYLIST Table
Id
Num
Gender Jobcode
Salary
Birth
Hired
------------------------------------------------1639 F
TA1
42260 26JUN70 28JAN91
1065 M
ME3
38090 26JAN54 07JAN92
1400 M
ME1
29769 05NOV67 16OCT90
1561 M
36514 30NOV63 07OCT87
1221 F
FA3
. 22SEP63 04OCT94

Example 2: Creating a Table from a Querys Result

Procedure features:

CREATE TABLE statement

AS query-expression
SELECT clause
column alias
FORMAT= column-modier
object-item
Other features:

data set option

OBS=
Tables:

PROCLIB.PAYROLL, PROCLIB.BONUS

This example builds a column with an arithmetic expression and creates the
PROCLIB.BONUS table from the querys result.

1127

1128

Input Table

Chapter 44

Input Table

PROCLIB.PAYROLL (Partial Listing)

PROCLIB.PAYROLL
First 10 Rows Only
Id
Number Gender Jobcode
Salary
Birth
Hired
--------------------------------------------------1919
M
TA2
34376 12SEP60 04JUN87
1653
F
ME2
35108 15OCT64 09AUG90
1400
M
ME1
29769 05NOV67 16OCT90
1350
F
FA3
32886 31AUG65 29JUL90
1401
M
TA3
38822 13DEC50 17NOV85
1499
M
ME3
43025 26APR54 07JUN80
1101
M
SCP
18723 06JUN62 01OCT90
1333
M
PT2
88606 30MAR61 10FEB81
1402
M
TA2
32615 17JAN63 02DEC90
1479
F
TA3
38785 22DEC68 05OCT89

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Create the PROCLIB.BONUS table. The CREATE TABLE statement creates the table
PROCLIB.BONUS from the result of the subsequent query.
proc sql;
create table proclib.bonus as

Select the columns to include. The SELECT clause species that three columns will be in
the new table: IdNumber, Salary, and Bonus. FORMAT= assigns the DOLLAR8. format to
Salary. The Bonus column is built with the SQL expression salary*.025.
select IdNumber, Salary format=dollar8.,
salary*.025 as Bonus format=dollar8.
from proclib.payroll;

The SQL Procedure

Example 3: Updating Data in a PROC SQL Table

Specify the title.

title BONUS Information;

Display the rst 10 rows of the PROCLIB.BONUS table. The SELECT clause selects
columns from PROCLIB.BONUS. The asterisk (*) selects all columns. The FROM clause
species PROCLIB.BONUS as the table to select from. The OBS= data set option limits the
printing of the output to 10 rows.
select *
from proclib.bonus(obs=10);

Output

PROCLIB.BONUS

BONUS Information
Id
Number
Salary
Bonus
-------------------------1919
$34,376
$859
1653
$35,108
$878
1400
$29,769
$744
1350
$32,886
$822
1401
$38,822
$971
1499
$43,025
$1,076
1101
$18,723
$468
1333
$88,606
$2,215
1402
$32,615
$815
1479
$38,785
$970

Example 3: Updating Data in a PROC SQL Table

Procedure features:

ALTER TABLE statement

DROP clause
MODIFY clause
UPDATE statement
SET clause
CASE expression
Table:

EMPLOYEES

This example updates data values in the EMPLOYEES table and drops a column.

1129

1130

Input

Chapter 44

Input
data Employees;
input IdNum $4. +2 LName $11. FName $11. JobCode $3.
+1 Salary 5. +1 Phone $12.;
datalines;
1876 CHIN
JACK
TA1 42400 212/588-5634
1114 GREENWALD JANICE
ME3 38000 212/588-1092
1556 PENNINGTON MICHAEL
ME1 29860 718/383-5681
1354 PARKER
MARY
FA3 65800 914/455-2337
1130 WOOD
DEBORAH
PT2 36514 212/587-0013
;

Program

Display the entire EMPLOYEES table. The SELECT clause displays the table before the
updates. The asterisk (*) selects all columns for display. The FROM clause species
EMPLOYEES as the table to select from.
proc sql;
title Employees Table;
select * from Employees;

Update the values in the Salary column. The UPDATE statement updates the values in
EMPLOYEES. The SET clause species that the data in the Salary column be multiplied by
1.04 when the job code ends with a 1 and 1.025 for all other job codes. (The two underscores
represent any character.) The CASE expression returns a value for each row that completes the
SET clause.
update employees
set salary=salary*
case when jobcode like __1 then 1.04
else 1.025
end;

Modify the format of the Salary column and delete the Phone column. The ALTER
TABLE statement species EMPLOYEES as the table to alter. The MODIFY clause
permanently modies the format of the Salary column. The DROP clause permanently drops the
Phone column.
alter table employees
modify salary num format=dollar8.
drop phone;

The SQL Procedure

Example 4: Joining Two Tables

Specify the title.

title Updated Employees Table;

Display the entire updated EMPLOYEES table. The SELECT clause displays the
EMPLOYEES table after the updates. The asterisk (*) selects all columns.
select * from employees;

Output
Employees Table

Id
Job
Num
LName
FName
Code
Salary Phone
-----------------------------------------------------------1876 CHIN
JACK
TA1
42400 212/588-5634
1114 GREENWALD
JANICE
ME3
38000 212/588-1092
1556 PENNINGTON
MICHAEL
ME1
29860 718/383-5681
1354 PARKER
MARY
FA3
65800 914/455-2337
1130 WOOD
DEBORAH
PT2
36514 212/587-0013

Updated Employees Table

Id
Job
Num
LName
FName
Code
Salary
---------------------------------------------1876 CHIN
JACK
TA1
$44,096
1114 GREENWALD
JANICE
ME3
$38,950
1556 PENNINGTON
MICHAEL
ME1
$31,054
1354 PARKER
MARY
FA3
$67,445
1130 WOOD
DEBORAH
PT2
$37,427

Example 4: Joining Two Tables

Procedure features:

FROM clause
table alias
inner join
joined-table component
PROC SQL statement option
NUMBER
WHERE clause
IN condition

1131

1132

Input Tables

Tables:

Chapter 44

PROCLIB.STAFF, PROCLIB.PAYROLL

This example joins two tables in order to get more information about data that are
common to both tables.

Input Tables

PROCLIB.STAFF (Partial Listing)

PROCLIB.STAFF
First 10 Rows Only
Id
Num
Lname
Fname
City
State Hphone
---------------------------------------------------------------------------1919 ADAMS
GERALD
STAMFORD
CT
203/781-1255
1653 ALIBRANDI
MARIA
BRIDGEPORT
CT
203/675-7715
1400 ALHERTANI
ABDULLAH
NEW YORK
NY
212/586-0808
1350 ALVAREZ
MERCEDES
NEW YORK
NY
718/383-1549
1401 ALVAREZ
CARLOS
PATERSON
NJ
201/732-8787
1499 BAREFOOT
JOSEPH
PRINCETON
NJ
201/812-5665
1101 BAUCOM
WALTER
NEW YORK
NY
212/586-8060
1333 BANADYGA
JUSTIN
STAMFORD
CT
203/781-1777
1402 BLALOCK
RALPH
NEW YORK
NY
718/384-2849
1479 BALLETTI
MARIE
NEW YORK
NY
718/384-8816

PROCLIB.PAYROLL (Partial Listing)

The SQL Procedure

Program

1133

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Add row numbers to PROC SQL output. NUMBER adds a column that contains the row
number.
proc sql number;

Specify the title.

title Information for Certain Employees Only;

Select the columns to display. The SELECT clause selects the columns to show in the output.
select Lname, Fname, City, State,
IdNumber, Salary, Jobcode

Specify the tables from which to obtain the data. The FROM clause lists the tables to
select from.
from proclib.staff, proclib.payroll

Specify the join criterion and subset the query. The WHERE clause species that the
tables are joined on the ID number from each table. WHERE also further subsets the query
with the IN condition, which returns rows for only four employees.
where idnumber=idnum and idnum in
(1919, 1400, 1350, 1333);

1134

Output

Chapter 44

Output
Information for Certain Employees Only
Id
Lname
Fname
City
State Number
Salary Jobcode
-----------------------------------------------------------------------1 ADAMS
GERALD
STAMFORD
CT
1919
34376 TA2
Row

ALHERTANI
29769 ME1

ABDULLAH

NEW YORK

1400

ALVAREZ
32886

MERCEDES

NEW YORK

1350

FA3

BANADYGA
88606

JUSTIN

STAMFORD

1333

PT2

Example 5: Combining Two Tables

Procedure features:

DELETE statement
IS condition
RESET statement option
DOUBLE
UNION set operator
Tables:

PROCLIB.NEWPAY, PROCLIB.PAYLIST, PROCLIB.PAYLIST2

This example creates a new table, PROCLIB.NEWPAY, by concatenating two other

tables: PROCLIB.PAYLIST and PROCLIB.PAYLIST2.

Input Tables

PROCLIB.PAYLIST

Information for Certain Employees Only

Id
Num
Gender Jobcode
Salary
Birth
Hired
------------------------------------------------1639 F
TA1
42260 26JUN70 28JAN91
1065 M
ME3
38090 26JAN54 07JAN92
1400 M
ME1
29769 05NOV67 16OCT90
1561 M
36514 30NOV63 07OCT87
1221 F
FA3
. 22SEP63 04OCT94

The SQL Procedure

Program

1135

PROCLIB.PAYLIST2

PROCLIB.PAYLIST2 Table
Id
Num
Gender Jobcode
Salary
Birth
Hired
------------------------------------------------1919 M
TA2
34376 12SEP66 04JUN87
1653 F
ME2
31896 15OCT64 09AUG92
1350 F
FA3
36886 31AUG55 29JUL91
1401 M
TA3
38822 13DEC55 17NOV93
1499 M
ME1
23025 26APR74 07JUN92

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Create the PROCLIB.NEWPAY table. The SELECT clauses select all the columns from the
tables that are listed in the FROM clauses. The UNION set operator concatenates the query
results that are produced by the two SELECT clauses.
proc sql;
create table proclib.newpay as
select * from proclib.paylist
union
select * from proclib.paylist2;

Delete rows with missing Jobcode or Salary values. The DELETE statement deletes rows
from PROCLIB.NEWPAY that satisfy the WHERE expression. The IS condition species rows
that contain missing values in the Jobcode or Salary column.
delete
from proclib.newpay
where jobcode is missing or salary is missing;

1136

Output

Chapter 44

Reset the PROC SQL environment and double-space the output. RESET changes the
procedure environment without stopping and restarting PROC SQL. The DOUBLE option
double-spaces the output. (The DOUBLE option has no effect on ODS output.)
reset double;

Specify the title.

title Personnel Data;

Display the entire PROCLIB.NEWPAY table. The SELECT clause selects all columns from
the newly created table, PROCLIB.NEWPAY.
select *
from proclib.newpay;

Output
Personnel Data
Id
Num
Gender Jobcode
Salary
Birth
Hired
------------------------------------------------1065 M
ME3
38090 26JAN54 07JAN92
1350

FA3

36886

31AUG55

29JUL91

1400

ME1

29769

05NOV67

16OCT90

1401

TA3

38822

13DEC55

17NOV93

1499

ME1

23025

26APR74

07JUN92

1639

TA1

42260

26JUN70

28JAN91

1653

ME2

31896

15OCT64

09AUG92

1919

TA2

34376

12SEP66

04JUN87

Example 6: Reporting from DICTIONARY Tables

Procedure features:

DESCRIBE TABLE statement

DICTIONARY.table-name component
Table: DICTIONARY.MEMBERS

This example uses DICTIONARY tables to show a list of the SAS les in a SAS data
library. If you do not know the names of the columns in the DICTIONARY table that
you are querying, then use a DESCRIBE TABLE statement with the table.

The SQL Procedure

Program

1137

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. SOURCE writes
the programming statements to the SAS log.
options nodate pageno=1 source linesize=80 pagesize=60;

List the column names from the DICTIONARY.MEMBERS table. DESCRIBE TABLE
writes the column names from DICTIONARY.MEMBERS to the SAS log.
proc sql;
describe table dictionary.members;

Specify the title.

title SAS Files in the PROCLIB Library;

Display a list of les in the PROCLIB library. The SELECT clause selects the MEMNAME
and MEMTYPE columns. The FROM clause species DICTIONARY.MEMBERS as the table to
select from. The WHERE clause subsets the output to include only those rows that have a libref
of PROCLIB in the LIBNAME column.
select memname, memtype
from dictionary.members
where libname=PROCLIB;

1138

Log

Chapter 44

Log
277 options nodate pageno=1 source linesize=80 pagesize=60;
278
279 proc sql;
280
describe table dictionary.members;
NOTE: SQL table DICTIONARY.MEMBERS was created like:
create table DICTIONARY.MEMBERS
(
libname char(8) label=Library Name,
memname char(32) label=Member Name,
memtype char(8) label=Member Type,
engine char(8) label=Engine Name,
index char(32) label=Indexes,
path char(1024) label=Path Name
);
281
282
283
284
285

title SAS Files in the PROCLIB Library;

select memname, memtype
from dictionary.members
where libname=PROCLIB;

Output
SAS Files in the PROCLIB Library
Member
Member Name
Type
-----------------------------------------ALL
DATA
BONUS
DATA
BONUS95
DATA
DELAY
DATA
HOUSES
DATA
INTERNAT
DATA
MARCH
DATA
NEWPAY
DATA
PAYLIST
DATA
PAYLIST2
DATA
PAYROLL
DATA
PAYROLL2
DATA
SCHEDULE
DATA
SCHEDULE2
DATA
STAFF
DATA
STAFF2
DATA
SUPERV
DATA
SUPERV2
DATA

Example 7: Performing an Outer Join

Procedure features:

joined-table component
left outer join
SELECT clause

The SQL Procedure

Input Tables

COALESCE function
WHERE clause
CONTAINS condition
Tables:

PROCLIB.PAYROLL, PROCLIB.PAYROLL2

This example illustrates a left outer join of the PROCLIB.PAYROLL and

PROCLIB.PAYROLL2 tables.

Input Tables

PROCLIB.PAYROLL (Partial Listing)

PROCLIB.PAYROLL
First 10 Rows Only
Id
Number Gender Jobcode
Salary
Birth
Hired
--------------------------------------------------1009
M
TA1
28880 02MAR59 26MAR92
1017
M
TA3
40858 28DEC57 16OCT81
1036
F
TA3
39392 19MAY65 23OCT84
1037
F
TA1
28558 10APR64 13SEP92
1038
F
TA1
26533 09NOV69 23NOV91
1050
M
ME2
35167 14JUL63 24AUG86
1065
M
ME2
35090 26JAN44 07JAN87
1076
M
PT1
66558 14OCT55 03OCT91
1094
M
FA1
22268 02APR70 17APR91
1100
M
BCK
25004 01DEC60 07MAY88

PROCLIB.PAYROLL2

PROCLIB.PAYROLL2
Id
Num
Sex Jobcode
Salary
Birth
Hired
---------------------------------------------1036 F
TA3
42465 19MAY65 23OCT84
1065 M
ME3
38090 26JAN44 07JAN87
1076 M
PT1
69742 14OCT55 03OCT91
1106 M
PT3
94039 06NOV57 16AUG84
1129 F
ME3
36758 08DEC61 17AUG91
1221 F
FA3
29896 22SEP67 04OCT91
1350 F
FA3
36098 31AUG65 29JUL90
1369 M
TA3
36598 28DEC61 13MAR87
1447 F
FA1
22123 07AUG72 29OCT92
1561 M
TA3
36514 30NOV63 07OCT87
1639 F
TA3
42260 26JUN57 28JAN84
1998 M
SCP
23100 10SEP70 02NOV92

1139

1140

Program

Chapter 44

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Limit the number of output rows. OUTOBS= limits the output to 10 rows.
proc sql outobs=10;

Specify the title for the rst query.

title Most Current Jobcode and Salary Information;

Select the columns. The SELECT clause lists the columns to select. Some column names are
prexed with a table alias because they are in both tables. LABEL= and FORMAT= are column
modiers.
select p.IdNumber, p.Jobcode, p.Salary,
p2.jobcode label=New Jobcode,
p2.salary label=New Salary format=dollar8.

Specify the type of join. The FROM clause lists the tables to join and assigns table aliases.
The keywords LEFT JOIN specify the type of join. The order of the tables in the FROM clause
is important. PROCLIB.PAYROLL is listed rst and is considered the left table.
PROCLIB.PAYROLL2 is the right table.
from proclib.payroll as p left join proclib.payroll2 as p2

Specify the join criterion. The ON clause species that the join be performed based on the
values of the ID numbers from each table.
on p.IdNumber=p2.idnum;

The SQL Procedure

Output

1141

Output

As the output shows, all rows from the left table, PROCLIB.PAYROLL, are returned. PROC
SQL assigns missing values for rows in the left table, PAYROLL, that have no matching values
for IdNum in PAYROLL2.

Most Current Jobcode and Salary Information

Id
New
New
Number Jobcode
Salary Jobcode
Salary
-------------------------------------------1009
TA1
28880
.
1017
TA3
40858
.
1036
TA3
39392 TA3
$42,465
1037
TA1
28558
.
1038
TA1
26533
.
1050
ME2
35167
.
1065
ME2
35090 ME3
$38,090
1076
PT1
66558 PT1
$69,742
1094
FA1
22268
.
1100
BCK
25004
.

Specify the title for the second query.

title Most Current Jobcode and Salary Information;

Select the columns and coalesce the Jobcode columns.The SELECT clause lists the
columns to select. COALESCE overlays the like-named columns. For each row, COALESCE
returns the rst nonmissing value of either P2.JOBCODE or P.JOBCODE. Because
P2.JOBCODE is the rst argument, if there is a nonmissing value for P2.JOBCODE,
COALESCE returns that value. Thus, the output contains the most recent job code information
for every employee. LABEL= assigns a column label.
select p.idnumber, coalesce(p2.jobcode,p.jobcode)
label=Current Jobcode,

Coalesce the Salary columns. For each row, COALESCE returns the rst nonmissing value
of either P2.SALARY or P.SALARY. Because P2.SALARY is the rst argument, if there is a
nonmissing value for P2.SALARY, then COALESCE returns that value. Thus, the output
contains the most recent salary information for every employee.
coalesce(p2.salary,p.salary) label=Current Salary
format=dollar8.

1142

Output

Chapter 44

Specify the type of join and the join criterion. The FROM clause lists the tables to join and
assigns table aliases. The keywords LEFT JOIN specify the type of join. The ON clause species
that the join is based on the ID numbers from each table.
from proclib.payroll p left join proclib.payroll2 p2
on p.IdNumber=p2.idnum;

Output

Most Current Jobcode and Salary Information

Id
Current
Current
Number Jobcode
Salary
------------------------1009
TA1
$28,880
1017
TA3
$40,858
1036
TA3
$42,465
1037
TA1
$28,558
1038
TA1
$26,533
1050
ME2
$35,167
1065
ME3
$38,090
1076
PT1
$69,742
1094
FA1
$22,268
1100
BCK
$25,004

Subset the query. The WHERE clause subsets the left join to include only those rows
containing the value TA.
title Most Current Information for Ticket Agents;
select p.IdNumber,
coalesce(p2.jobcode,p.jobcode) label=Current Jobcode,
coalesce(p2.salary,p.salary) label=Current Salary
from proclib.payroll p left join proclib.payroll2 p2
on p.IdNumber=p2.idnum
where p2.jobcode contains TA;

Output

Most Current Information for Ticket Agents

Id
Current
Current
Number Jobcode
Salary
------------------------1036
TA3
42465
1369
TA3
36598
1561
TA3
36514
1639
TA3
42260

The SQL Procedure

Program

Example 8: Creating a View from a Querys Result

Procedure features:

CREATE VIEW statement

GROUP BY clause
SELECT clause
COUNT function
HAVING clause
Other features:

AVG summary function

data set option
PW=
Tables:

PROCLIB.PAYROLL, PROCLIB.JOBS

This example creates the PROC SQL view PROCLIB.JOBS from the result of a
query-expression.

Input Table

PROCLIB.PAYROLL (Partial Listing)

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

1143

1144

Program

Chapter 44

Create the PROCLIB.JOBS view. CREATE VIEW creates the PROC SQL view
PROCLIB.JOBS. The PW= data set option assigns password protection to the data that is
generated by this view.
proc sql;
create view proclib.jobs(pw=red) as

Select the columns. The SELECT clause species four columns for the view: Jobcode and
three columns, Number, AVGAGE, and AVGSAL, whose values are the products functions.
COUNT returns the number of nonmissing values for each job code because the data is grouped
by Jobcode. LABEL= assigns a label to the column.
select Jobcode,
count(jobcode) as number label=Number,

Calculate the Avgage and Avgsal columns. The AVG summary function calculates the
average age and average salary for each job code.
avg(int((today()-birth)/365.25)) as avgage
format=2. label=Average Age,
avg(salary) as avgsal
format=dollar8. label=Average Salary

Specify the table from which the data is obtained. The FROM clause species PAYROLL
as the table to select from. PROC SQL assumes the libref of PAYROLL to be PROCLIB because
PROCLIB is used in the CREATE VIEW statement.
from payroll

Organize the data into groups and specify the groups to include in the output. The
GROUP BY clause groups the data by the values of Jobcode. Thus, any summary statistics are
calculated for each grouping of rows by value of Jobcode. The HAVING clause subsets the
grouped data and returns rows for job codes that contain an average age of greater than or
equal to 30.
group by jobcode
having avgage ge 30;

Specify the titles.

title Current Summary Information for Each Job Category;
title2 Average Age Greater Than or Equal to 30;

The SQL Procedure

Example 9: Joining Three Tables

1145

Display the entire PROCLIB.JOBS view. The SELECT statement selects all columns from
PROCLIB.JOBS. PW=RED is necessary because the view is password protected.
select * from proclib.jobs(pw=red);

Output
Current Summary Information for Each Job Category
Average Age Greater Than Or Equal to 30
Average
Average
Jobcode
Number
Age
Salary
-----------------------------------BCK
9
36
$25,794
FA1
11
33
$23,039
FA2
16
37
$27,987
FA3
7
39
$32,934
ME1
8
34
$28,500
ME2
14
39
$35,577
ME3
7
42
$42,411
NA1
5
30
$42,032
NA2
3
42
$52,383
PT1
8
38
$67,908
PT2
10
43
$87,925
PT3
2
54
$10,505
SCP
7
37
$18,309
TA1
9
36
$27,721
TA2
20
36
$33,575
TA3
12
40
$39,680

Example 9: Joining Three Tables

Procedure features:

FROM clause
joined-table component
WHERE clause
Tables: PROCLIB.STAFF2, PROCLIB.SCHEDULE2, PROCLIB.SUPERV2

This example joins three tables and produces a report that contains columns from
each table.

1146

Input Tables

Chapter 44

Input Tables

PROCLIB.STAFF2

PROCLIB.STAFF2
Id
Num
Lname
Fname
City
State Hphone
---------------------------------------------------------------------------1106 MARSHBURN
JASPER
STAMFORD
CT
203/781-1457
1430 DABROWSKI
SANDRA
BRIDGEPORT
CT
203/675-1647
1118 DENNIS
ROGER
NEW YORK
NY
718/383-1122
1126 KIMANI
ANNE
NEW YORK
NY
212/586-1229
1402 BLALOCK
RALPH
NEW YORK
NY
718/384-2849
1882 TUCKER
ALAN
NEW YORK
NY
718/384-0216
1479 BALLETTI
MARIE
NEW YORK
NY
718/384-8816
1420 ROUSE
JEREMY
PATERSON
NJ
201/732-9834
1403 BOWDEN
EARL
BRIDGEPORT
CT
203/675-3434
1616 FUENTAS
CARLA
NEW YORK
NY
718/384-3329

PROCLIB.SCHEDULE2

PROCLIB.SCHEDULE2
Id
Flight
Date Dest Num
--------------------------132
01MAR94 BOS
1118
132
01MAR94 BOS
1402
219
02MAR94 PAR
1616
219
02MAR94 PAR
1478
622
03MAR94 LON
1430
622
03MAR94 LON
1882
271
04MAR94 NYC
1430
271
04MAR94 NYC
1118
579
05MAR94 RDU
1126
579
05MAR94 RDU
1106

The SQL Procedure

Program

1147

PROCLIB.SUPERV2

PROCLIB.SUPERV2
Supervisor
Job
Id
State Category
--------------------------1417
NJ
NA
1352
NY
NA
1106
CT
PT
1442
NJ
PT
1118
NY
PT
1405
NJ
SC
1564
NY
SC
1639
CT
TA
1126
NY
TA
1882
NY
ME

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Select the columns. The SELECT clause species the columns to select. IdNum is prexed
with a table alias because it appears in two tables.
proc sql;
title All Flights for Each Supervisor;
select s.IdNum, Lname, City Hometown, Jobcat,
Flight, Date

Specify the tables to include in the join. The FROM clause lists the three tables for the join
and assigns an alias to each table.
from proclib.schedule2 s, proclib.staff2 t, proclib.superv2 v

1148

Output

Chapter 44

Specify the join criteria. The WHERE clause species the columns that join the tables. The
STAFF2 and SCHEDULE2 tables have an IdNum column, which has related values in both
tables. The STAFF2 and SUPERV2 tables have the IdNum and SUPID columns, which have
related values in both tables.
where s.idnum=t.idnum and t.idnum=v.supid;

Output
All Flights for Each Supervisor
Id
Job
Num
Lname
Hometown
Category Flight
Date
----------------------------------------------------------------1106 MARSHBURN
STAMFORD
PT
579
05MAR94
1118 DENNIS
NEW YORK
PT
132
01MAR94
1118 DENNIS
NEW YORK
PT
271
04MAR94
1126 KIMANI
NEW YORK
TA
579
05MAR94
1882 TUCKER
NEW YORK
ME
622
03MAR94

Example 10: Querying an In-Line View

Procedure features:

FROM clause
in-line view
Tables: PROCLIB.STAFF2, PROCLIB.SCHEDULE2, PROCLIB.SUPERV2

This example shows an alternative way to construct the query that is explained in
Example 9 on page 1145 by joining one of the tables with the results of an in-line view.
The example also shows how to rename columns with an in-line view.

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

The SQL Procedure

Output

1149

Select the columns. The SELECT clause selects all columns that are returned by the in-line
view (which will have the alias Three assigned to it), plus one column from the third table
(which will have the alias V assigned to it).
proc sql;
title All Flights for Each Supervisor;
select three.*, v.jobcat

Specify the in-line query. Instead of including the name of a table or view, the FROM clause
includes a query that joins two of the three tables. In the in-line query, the SELECT clause lists
the columns to select. IdNum is prexed with a table alias because it appears in both tables.
The FROM clause lists the two tables for the join and assigns an alias to each table. The
WHERE clause species the columns that join the tables. The STAFF2 and SCHEDULE2 tables
have an IdNum column, which has related values in both tables.
from (select lname, s.idnum, city, flight, date
from proclib.schedule2 s, proclib.staff2 t
where s.idnum=t.idnum)

Specify an alias for the query and names for the columns. The alias Three refers to the
results of the in-line view. The names in parentheses become the names for the columns in the
view.
as three (Surname, Emp_ID, Hometown,
FlightNumber, FlightDate),

Join the results of the in-line view with the third table. The WHERE clause species the
columns that join the table with the in-line view. Note that the WHERE clause species the
renamed Emp_ID column from the in-line view.
proclib.superv2 v
where three.Emp_ID=v.supid;

Output
All Flights for Each Supervisor
Job
Surname
Emp_ID Hometown
FlightNumber FlightDate Category
---------------------------------------------------------------------------MARSHBURN
1106
STAMFORD
579
05MAR94 PT
DENNIS
1118
NEW YORK
132
01MAR94 PT
DENNIS
1118
NEW YORK
271
04MAR94 PT
KIMANI
1126
NEW YORK
579
05MAR94 TA
TUCKER
1882
NEW YORK
622
03MAR94 ME

1150

Example 11: Retrieving Values with the SOUNDS-LIKE Operator

Chapter 44

Example 11: Retrieving Values with the SOUNDS-LIKE Operator

Procedure features:

ORDER BY clause
SOUNDS-LIKE operator
Table:

PROCLIB.STAFF

This example returns rows based on the functionality of the SOUNDS-LIKE operator
in a WHERE clause.
Note: The SOUNDS-LIKE operator is based on the SOUNDEX algorithm for
identifying words that sound alike. The SOUNDEX algorithm is English-biased and is
less useful for languages other than English. For more information on the SOUNDEX
algorithm, see SAS Language Reference: Dictionary. 4

Input Table

PROCLIB.STAFF

Program

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

options nodate pageno=1 linesize=80 pagesize=60;

The SQL Procedure

Output

1151

Select the columns and the table from which the data is obtained. The SELECT clause
selects all columns from the table in the FROM clause, PROCLIB.STAFF.
proc sql;
title "Employees Whose Last Name Sounds Like Johnson";
select idnum, upcase(lname), fname
from proclib.staff

Subset the query and sort the output. The WHERE clause uses the SOUNDS-LIKE
operator to subset the table by those employees whose last name sounds like Johnson. The
ORDER BY clause orders the output by the second column.
where lname=*"Johnson"
order by 2;

Output
Employees Whose Last Name Sounds Like Johnson

Id
Num
Fname
-------------------------------------1411 JOHNSEN
JACK
1113 JOHNSON
LESLIE
1369 JONSON
ANTHONY

SOUNDS-LIKE is useful, but there might be instances where it does not return every row that
seems to satisfy the condition. PROCLIB.STAFF has an employee with the last name SANDERS
and an employee with the last name SANYERS. The algorithm does not nd SANYERS, but it does
nd SANDERS and SANDERSON.
title "Employees Whose Last Name Sounds Like Sanders";
select *
from proclib.staff
where lname=*"Sanders"
order by 2;

Employees Whose Last Name Sounds Like Sanders

Id
Num
Lname
Fname
City
State Hphone
---------------------------------------------------------------------------1561 SANDERS
RAYMOND
NEW YORK
NY
212/588-6615
1414 SANDERSON
NATHAN
BRIDGEPORT
CT
203/675-1715
1434 SANDERSON
EDITH
STAMFORD
CT
203/781-1333

1152

Example 12: Joining Two Tables and Calculating a New Value

Chapter 44

Example 12: Joining Two Tables and Calculating a New Value

Procedure features:

GROUP BY clause
HAVING clause
SELECT clause
ABS function
FORMAT= column-modier
LABEL= column-modier
MIN summary function
** operator, exponentiation
SQRT function
Tables: STORES, HOUSES

This example joins two tables in order to compare and analyze values that are unique
to each table yet have a relationship with a column that is common to both tables.
options ls=80 ps=60 nodate pageno=1 ;
data stores;
input Store $ x y;
datalines;
store1 5 1
store2 5 3
store3 3 5
store4 7 5
;
data houses;
input House $ x y;
datalines;
house1 1 1
house2 3 3
house3 2 3
house4 7 7
;

Input Tables

STORES and HOUSES

The tables contain X and Y coordinates that represent the location of the stores and houses.

STORES Table
Coordinates of Stores
Store
x
y
---------------------------store1
6
1
store2
5
2
store3
3
5
store4
7
5

The SQL Procedure

HOUSES Table
Coordinates of Houses

Program

1153

House
x
y
---------------------------house1
1
1
house2
3
3
house3
2
3
house4
7
7

Program

Specify the query. The SELECT clause species three columns: HOUSE, STORE, and DIST.
The arithmetic expression uses the square root function (SQRT) to create the values of DIST,
which contain the distance from HOUSE to STORE for each row. The double asterisk (**)
represents exponentiation. LABEL= assigns a label to STORE and to DIST.
proc sql;
title Each House and the Closest Store;
select house, store label=Closest Store,
sqrt((abs(s.x-h.x)**2)+(abs(h.y-s.y)**2)) as dist
label=Distance format=4.2
from stores s, houses h

Organize the data into groups and subset the query. The minimum distance from each
house to all the stores is calculated because the data are grouped by house. The HAVING clause
species that each row be evaluated to determine if its value of DIST is the same as the
minimum distance from that house to any store.
group by house
having dist=min(dist);

1154

Output

Chapter 44

Output

Note that two stores are tied for shortest distance from house2.

Each House and the Closest Store

Closest
House
Store
Distance
---------------------------house1
store1
4.00
house2
store2
2.00
house2
store3
2.00
house3
store3
2.24
house4
store4
2.00

Example 13: Producing All the Possible Combinations of the Values in a

Column
Procedure features:

CASE expression
joined-table component
Cross join
SELECT clause
DISTINCT keyword
Tables: PROCLIB.MARCH, FLIGHTS

This example joins a table with itself to get all the possible combinations of the
values in a column.

The SQL Procedure

Program to Create the Flights Table

1155

Input Table

PROCLIB.MARCH (Partial Listing)

PROCLIB.MARCH
First 10 Rows Only

Flight
Date Depart Orig Dest
Miles
Boarded Capacity
----------------------------------------------------------------114
01MAR94
7:10 LGA
LAX
2475
172
210
202
01MAR94
10:43 LGA
ORD
740
151
210
219
01MAR94
9:31 LGA
LON
3442
198
250
622
01MAR94
12:19 LGA
FRA
3857
207
250
132
01MAR94
15:35 LGA
YYZ
366
115
178
271
01MAR94
13:17 LGA
PAR
3635
138
250
302
01MAR94
20:22 LGA
WAS
229
105
180
114
02MAR94
7:10 LGA
LAX
2475
119
210
202
02MAR94
10:43 LGA
ORD
740
120
210
219
02MAR94
9:31 LGA
LON
3442
147
250

Program to Create the Flights Table

Declare the PROCLIB library. The PROCLIB library is used in these examples to store
created tables.
libname proclib SAS-data-library;

Create the FLIGHTS table. The CREATE TABLE statement creates the table FLIGHTS from
the output of the query. The SELECT clause selects the unique values of Dest. DISTINCT
species that only one row for each value of city be returned by the query and stored in the
table FLIGHTS. The FROM clause species PROCLIB.MARCH as the table to select from.
proc sql;
create table flights as
select distinct dest
from proclib.march;

1156

Output

Chapter 44

Specify the title.

title Cities Serviced by the Airline;

Display the entire FLIGHTS table.

select * from flights;

Output

FLIGHTS Table

Cities Serviced by the Airline

Dest
---FRA
LAX
LON
ORD
PAR
WAS
YYZ

Program Using Conventional Join

Specify the title.

title All Possible Connections;

Select the columns. The SELECT clause species three columns for the output. The prexes
on DEST are table aliases to specify which table to take the values of Dest from. The CASE
expression creates a column that contains the character string to and from.
select f1.Dest, case
when f1.dest ne then to and from
end,
f2.Dest

Specify the type of join. The FROM clause joins FLIGHTS with itself and creates a table that
contains every possible combination of rows (a Cartesian product). The table contains two rows
for each possible route, for example, PAR <-> WAS and WAS <-> PAR.
from flights as f1, flights as f2

The SQL Procedure

Program Using Cross Join

1157

Specify the join criterion. The WHERE clause subsets the internal table by choosing only
those rows where the name in F1.Dest sorts before the name in F2.Dest. Thus, there is only one
row for each possible route.
where f1.dest < f2.dest

Sort the output. ORDER BY sorts the result by the values of F1.Dest.
order by f1.dest;

Output

All Possible Connections

Dest
Dest
----------------------FRA
to and from LAX
FRA
to and from LON
FRA
to and from WAS
FRA
to and from ORD
FRA
to and from PAR
FRA
to and from YYZ
LAX
to and from LON
LAX
to and from PAR
LAX
to and from WAS
LAX
to and from ORD
LAX
to and from YYZ
LON
to and from ORD
LON
to and from WAS
LON
to and from PAR
LON
to and from YYZ
ORD
to and from WAS
ORD
to and from PAR
ORD
to and from YYZ
PAR
to and from WAS
PAR
to and from YYZ
WAS
to and from YYZ

Program Using Cross Join

Specify a cross join. Because a cross join is functionally the same as a Cartesian product join,
the cross join syntax can be substituted for the conventional join syntax.
proc sql;
title All Possible Connections;
select f1.Dest, case
when f1.dest ne then to and from
end,
f2.Dest
from flights as f1 cross join flights as f2
where f1.dest < f2.dest
order by f1.dest;

1158

Output

Chapter 44

Output
All Possible Connections

Example 14: Matching Case Rows and Control Rows

Procedure features:

joined-table component
Tables:

MATCH_11 on page 1388, MATCH

This example uses a table that contains data for a case-control study. Each row
contains information for a case or a control. To perform statistical analysis, you need a
table with one row for each case-control pair. PROC SQL joins the table with itself in
order to match the cases with their appropriate controls. After the rows are matched,
differencing can be performed on the appropriate columns.
The input table MATCH_11 contains one row for each case and one row for each
control. Pair contains a number that associates the case with its control. Low is 0 for
the controls and 1 for the cases. The remaining columns contain information about the
cases and controls.

The SQL Procedure

Program

1159

Input Table

MATCH_11 Table
First 10 Rows Only

Pair
Low
Age
Lwt
Race
Smoke
Ptd
Ht
UI
race1
race2
-----------------------------------------------------------------------------------------------------------1
1
2

0
1
0

14
14
15

135
101
98

1
3
2

0
1
0

0
0
0

0
0
1

0
1
0

2
3
3

1
0
1

15
16
16

115
95
130

3
3
3

0
0
0

1
0
0

0
0
0

1
1
1

4
4

0
1

17
17

103
130

3
3

0
1

0
0

0
1

0
0

1
1

5
5

0
1

17
17

122
110

1
1

0
0

Program

Create the MATCH table. The SELECT clause species the columns for the table MATCH.
SQL expressions in the SELECT clause calculate the differences for the appropriate columns
and create new columns.
proc sql;
create table match as
select
one.Low,
one.Pair,
(one.lwt - two.lwt) as Lwt_d,
(one.smoke - two.smoke) as Smoke_d,
(one.ptd - two.ptd) as Ptd_d,
(one.ht - two.ht) as Ht_d,
(one.ui - two.ui) as UI_d

Specify the type of join and the join criterion. The FROM clause lists the table MATCH_11
twice. Thus, the table is joined with itself. The WHERE clause returns only the rows for each
pair that show the difference when the values for control are subtracted from the values for case.
from match_11 one, match_11 two
where (one.pair=two.pair and one.low>two.low);

1160

Output

Chapter 44

Specify the title.

title Differences for Cases and Controls;

Display the rst ve rows of the MATCH table. The SELECT clause selects all the columns
from MATCH. The OBS= data set option limits the printing of the output to ve rows.
select *
from match(obs=5);

Output

MATCH Table

Differences for Cases and Controls

Low
Pair
Lwt_d
Smoke_d
Ptd_d
Ht_d
UI_d
-------------------------------------------------------------------1
1
-34
1
1
0
0
1
2
17
0
0
0
1
1
3
35
0
0
0
0
1
4
27
1
1
0
1
1
5
-12
0
0
0
0

Example 15: Counting Missing Values with a SAS Macro

Procedure feature:

COUNT function
Table:

SURVEY

This example uses a SAS macro to create columns. The SAS macro is not explained
here. See SAS Macro Language: Reference for information on SAS macros.

The SQL Procedure

Program

1161

Input Table

SURVEY contains data from a questionnaire about diet and exercise habits. SAS enables you to
use a special notation for missing values. In the EDUC column, the .x notation indicates that
the respondent gave an answer that is not valid, and .n indicates that the respondent did not
answer the question. A period as a missing value indicates a data entry error.
data survey;
input id $
datalines;
1001 yes yes
1002 no yes
1003 no no
1004 yes yes
1005 no yes
1006 yes yes
1007 no yes
1008 no no
;

diet $ exer $ hours xwk educ;

1 3 1
1 4 2
. . .n
2 3 .x
2 3 .x
2 4 .x
.5 3 .
. . .

Program

Count the nonmissing responses. The COUNTM macro uses the COUNT function to perform
various counts for a column. Each COUNT function uses a CASE expression to select the rows
to be counted. The rst COUNT function uses only the column as an argument to return the
number of nonmissing rows.
%macro countm(col);
count(&col) "Valid Responses for &col",

Count missing or invalid responses. The NMSS function returns the number of rows for
which the column has any type of missing value: .n, .x, or a period.
nmiss(&col) "Missing or NOT VALID Responses for &col",

1162

Output

Chapter 44

Count the occurrences of various sources of missing or invalid responses. The last
three COUNT functions use CASE expressions to count the occurrences of the three notations
for missing values. The count me character string gives the COUNT function a nonmissing
value to count.
count(case
when
end)
count(case
when
end)
count(case
when
end)
%mend;

&col=.n then "count me"

"Coded as NO ANSWER for &col",
&col=.x then "count me"
"Coded as NOT VALID answers for &col",
&col=. then "count me"
"Data Entry Errors for &col"

Use the COUNTM macro to create the columns. The SELECT clause species the columns
that are in the output. COUNT(*) returns the total number of rows in the table. The COUNTM
macro uses the values of the EDUC column to create the columns that are dened in the macro.
proc sql;
title Counts for Each Type of Missing Response;
select count(*) "Total No. of Rows",
%countm(educ)
from survey;

Output
Counts for Each Type of Missing Response
Missing
Coded as
or NOT Coded as
NOT
Data
Total
Valid
VALID
NO
VALID
Entry
No. of Responses Responses
ANSWER
answers
Errors
Rows
for educ
for educ for educ for educ for educ
-----------------------------------------------------------8
2
6
1
3
2

1163

CHAPTER

45
The STANDARD Procedure
Overview: STANDARD Procedure 1163
What Does the STANDARD Procedure Do? 1163
Standardizing Data 1163
Syntax: STANDARD Procedure 1165
PROC STANDARD Statement 1166
BY Statement 1168
FREQ Statement 1169
VAR Statement 1169
WEIGHT Statement 1169
Results: STANDARD Procedure 1170
Missing Values 1170
Output Data Set 1170
Statistical Computations: STANDARD Procedure 1171
Examples: STANDARD Procedure 1171
Example 1: Standardizing to a Given Mean and Standard Deviation
Example 2: Standardizing BY Groups and Replacing Missing Values

1171
1173

Overview: STANDARD Procedure

What Does the STANDARD Procedure Do?
The STANDARD procedure standardizes variables in a SAS data set to a given mean
and standard deviation, and it creates a new SAS data set containing the standardized
values.

Standardizing Data
Output 45.1 shows a simple standardization where the output data set contains
standardized student exam scores. The statements that produce the output follow:
proc standard data=score mean=75 std=5
out=stndtest;
run;
proc print data=stndtest;
run;

1164

Standardizing Data

Output 45.1

Chapter 45

Standardized Test Scores Using PROC STANDARD

The SAS System
Obs
1
2
3
4
5
6
7
8
9
10
11
12

Student

1
Test1

Capalleti
Dubose
Engles
Grant
Krupski
Lundsford
McBane
Mullen
Nguyen
Patel
Si
Tanaka

80.5388
64.3918
80.9143
68.8980
75.2816
79.7877
73.4041
78.6612
74.9061
71.9020
73.4041
77.9102

Output 45.2 shows a more complex example that uses BY-group processing. PROC
STANDARD computes Z scores separately for two BY groups by standardizing
life-expectancy data to a mean of 0 and a standard deviation of 1. The data are 1950
and 1993 life expectancies at birth for 16 countries. The birth rates for each country,
classied as stable or rapid, form the two BY groups. The statements that produce the
analysis also
3 print statistics for each variable to standardize
3 replace missing values with the given mean
3 calculate standardized values using a given mean and standard deviation
3 print the data set with the standardized values.
For an explanation of the program that produces this output, see Example 2 on page
1173.

Output 45.2

Z Scores for Each BY Group Using PROC STANDARD

Life Expectancies by Birth Rate

-------------------- PopulationRate=Stable --------------------The STANDARD Procedure

Name
Label

Mean

Life50
67.400000
1950 life expectancy
Life93
74.500000
1993 life expectancy

Standard
Deviation

1.854724

4.888763

--------------------- PopulationRate=Rapid ---------------------

Name
Label

Mean

Life50
42.000000
1950 life expectancy
Life93
59.100000
1993 life expectancy

Standard
Deviation

5.033223

8.225300

The STANDARD Procedure

Standardized Life Expectancies at Birth

by a Countrys Birth Rate
Population
Rate
Stable
Stable
Stable
Stable
Stable
Stable
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid

Country
France
Germany
Japan
Russia
United Kingdom
United States
Bangladesh
Brazil
China
Egypt
Ethiopia
India
Indonesia
Mozambique
Philippines
Turkey

Life50

1165

Life93

-0.21567
0.32350
-1.83316
0.00000
0.86266
0.86266
0.00000
1.78812
-0.19868
0.00000
-1.78812
-0.59604
-0.79472
0.00000
1.19208
0.39736

Syntax: STANDARD Procedure

0.51138
0.10228
0.92048
-1.94323
0.30683
0.10228
-0.74161
0.96045
1.32518
0.10942
-1.59265
-0.01216
-0.01216
-1.47107
0.59572
0.83888

Syntax: STANDARD Procedure

Supports the Output Delivery System. See Output Delivery System on page 32
for details.

Tip:

ODS Table Name:

OUT=

Computational options
Exclude observations with nonpositive weights

EXCLNPWGT

Specify the mean value

MEAN=

Replace missing values with a variable mean or

MEAN= value

REPLACE

Specify the standard deviation value

STD=

Specify the divisor for variance calculations

VARDEF=

Control printed output

Print statistics for each variable to standardize

Without Options
If you do not specify MEAN=, REPLACE, or STD=, the output data set is an
identical copy of the input data set.

Options
DATA=SAS-data-set

identies the input SAS data set.

Main discussion: Input Data Sets on page 19
Restriction: You cannot use PROC STANDARD with an engine that supports

concurrent access if another user is updating the data set at the same time.

The STANDARD Procedure

PROC STANDARD Statement

1167

EXCLNPWGT

excludes observations with nonpositive weight values (zero or negative). The

procedure does not use the observation to calculate the mean and standard deviation,
but the observation is still standardized. By default, the procedure treats
observations with negative weights like those with zero weights and counts them in
the total number of observations.
MEAN=mean-value

standardizes variables to a mean of mean-value.

Alias: M=
Default: mean of the input values
Featured in: Example 1 on page 1171
OUT=SAS-data-set

identies the output data set. If SAS-data-set does not exist, PROC STANDARD
creates it. If you omit OUT=, the data set is named DATAn, where n is the smallest
integer that makes the name unique.
Default: DATAn
Featured in: Example 1 on page 1171
PRINT

prints the original frequency, mean, and standard deviation for each variable to
standardize.
Featured in: Example 2 on page 1173
REPLACE

replaces missing values with the variable mean.

Interaction: If you use MEAN=, PROC STANDARD replaces missing values with
the given mean.
Featured in: Example 2 on page 1173
STD=std-value

standardizes variables to a standard deviation of

std-value.
Alias: S=
Default: standard deviation of the input values
Featured in: Example 1 on page 1171
VARDEF=divisor

species the divisor to use in the calculation of variances and standard deviation.
Table 45.1 on page 1167 shows the possible values for divisor and the associated
divisors.
Table 45.1

Possible Values for VARDEF=

Value

Divisor

Formula for Divisor

degrees of freedom

number of observations

WDF

sum of weights minus one

WEIGHT
|WGT

sum of weights

P (x 0

n1
n
(6i wi) 1

6i wi

CSS=divisor

CSS

The procedure computes the variance as

, where
is the corrected
x)2 . When you weight the analysis variables,
sums of squares and equals
i
CSS equals wi (xi xw )2 where xw is the weighted mean.

1168

BY Statement

Chapter 45

Default: DF

When you use the WEIGHT statement and VARDEF=DF, the variance is an
estimate of 2 , where the variance of the ith observation is var (xi ) = 2 =wi and
wi is the weight for the ith observation. This yields an estimate of the variance of
an observation with unit weight.

Tip:

When you use the WEIGHT statement and VARDEF=WGT, the computed
variance is asymptotically (for large n) an estimate of 2 =w, where w is the
average weight. This yields an asymptotic estimate of the variance of an
observation with average weight.

Tip:

Example 2 on page 1173

BY < DESCENDING> variable-1 <<DESCENDING> variable-n>< NOTSORTED>;

Required Arguments
variable

Options
DESCENDING

species that the data set is sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data are grouped in another way, such as chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the
NOTSORTED option. In fact, the procedure does not use an index if you specify
NOTSORTED. The procedure denes a BY group as a set of contiguous observations
that have the same values for all BY variables. If observations with the same values
for the BY variables are not contiguous, the procedure treats each contiguous set as a
separate BY group.

The STANDARD Procedure

WEIGHT Statement

1169

FREQ Statement
Species a numeric variable whose values represent the frequency of the observation.
The effects of the FREQ and WEIGHT statements are similar except when
calculating degrees of freedom.
See also: For an example that uses the FREQ statement, see FREQ on page 61
Tip:

FREQ variable;

Required Arguments
variable

species a numeric variable whose value represents the frequency of the observation.
If you use the FREQ statement, the procedure assumes that each observation
represents n observations, where n is the value of variable. If n is not an integer, the
SAS System truncates it. If n is less than 1 or is missing, the procedure does not use
that observation to calculate statistics but the observation is still standardized.
The sum of the frequency variable represents the total number of observations.

VAR Statement
Species the variables to standardize and their order in the printed output.
If you omit the VAR statement, PROC STANDARD standardizes all numeric
variables not listed in the other statements.
Featured in: Example 1 on page 1171
Default:

VAR variable(s);

Required Arguments
variable(s)

identies one or more variables to standardize.

WEIGHT Statement
Species weights for analysis variables in the statistical calculations.
See also: For information about calculating weighted statistics and for an example that
uses the WEIGHT statement, see WEIGHT on page 63

1170

Results: STANDARD Procedure

Chapter 45

WEIGHT variable;

Required Arguments
variable

species a numeric variable whose values weight the values of the analysis variables.
The values of the variable do not have to be integers. If the value of the weight
variable is
Weight value
0
less than 0
missing

PROC STANDARD
counts the observation in the total number of observations
converts the weight value to zero and counts the observation in
the total number of observations
excludes the observation from the calculation of mean and
standard deviation

To exclude observations that contain negative and zero weights from the calculation
of mean and standard deviation, use EXCLNPWGT. Note that most SAS/STAT
procedures, such as PROC GLM, exclude negative and zero weights by default.
Tip: When you use the WEIGHT statement, consider which value of the VARDEF=
option is appropriate. See VARDEF= on page 1167 and the calculation of weighted
statistics in Keywords and Formulas on page 1340 for more information.
Note: Prior to Version 7 of the SAS System, the procedure did not exclude the
observations with missing weights from the count of observations. 4

Results: STANDARD Procedure

Missing Values
By default, PROC STANDARD excludes missing values for the analysis variables
from the standardization process, and the values remain missing in the output data set.
When you specify the REPLACE option, the procedure replaces missing values with the
variables mean or the MEAN= value.
If the value of the WEIGHT variable or the FREQ variable is missing then the
procedure does not use the observation to calculate the mean and the standard
deviation. However, the observation is standardized.

Output Data Set

PROC STANDARD always creates an output data set that stores the standardized
values in the VAR statement variables, regardless of whether you specify the OUT=
option. The output data set contains all the input data set variables, including those
not standardized. PROC STANDARD does not print the output data set. Use PROC
PRINT, PROC REPORT, or another SAS reporting tool to print the output data set.

The STANDARD Procedure

Example 1: Standardizing to a Given Mean and Standard Deviation

Statistical Computations: STANDARD Procedure

Standardizing values removes the location and scale attributes from a set of data.
The formula to compute standardized values is

xi =
0

S 3 (xi 0 x)
+M
sx

where

xi
S
M
xi
x
sx
0

is a new standardized value

is the value of STD=
is the value of MEAN=
is an observations value
is a variables mean
is a variables standard deviation.

PROC STANDARD calculates the mean (x) and standard deviation (sx ) from the
input data set. The resulting standardized variable has a mean of M and a standard
deviation of S.
If the data are normally distributed, standardizing is also studentizing since the
resulting data have a Students t distribution.

Examples: STANDARD Procedure

Example 1: Standardizing to a Given Mean and Standard Deviation

Procedure features:

PROC STANDARD statement options:

MEAN=
OUT=
STD=
VAR statement
Other features:

PRINT procedure

This example
3 standardizes two variables to a mean of 75 and a standard deviation of 5
3 species the output data set
3 combines standardized variables with original variables
3 prints the output data set.

1171

1172

Program

Chapter 45

Program

Create the SCORE data set. This data set contains test scores for students who took two tests
and a nal exam. The FORMAT statement assigns the Zw.d format to StudentNumber. This
format pads right-justied output with 0s instead of blanks. The LENGTH statement species
the number of bytes to use to store values of Student.
data score;
length Student $ 9;
input Student $ StudentNumber Section $
Test1 Test2 Final @@;
format studentnumber z4.;
datalines;
Capalleti 0545 1 94 91 87 Dubose
1252 2
Engles
1167 1 95 97 97 Grant
1230 2
Krupski
2527 2 80 69 71 Lundsford 4860 1
McBane
0674 1 75 78 72 Mullen
6445 2
Nguyen
0886 1 79 76 80 Patel
9164 2
Si
4915 1 75 71 73 Tanaka
8534 2
;

51
63
92
89
71
87

65
75
40
82
77
73

91
80
86
93
83
76

Generate the standardized data and create the output data set STNDTEST. PROC
STANDARD uses a mean of 75 and a standard deviation of 5 to standardize the values. OUT=
identies STNDTEST as the data set to contain the standardized values.
proc standard data=score mean=75 std=5 out=stndtest;

Specify the variables to standardize. The VAR statement species the variables to
standardize and their order in the output.
var test1 test2;
run;

Create a data set that combines the original values with the standardized values.
PROC SQL joins SCORE and STNDTEST to create the COMBINED data set (table) that
contains standardized and original test scores for each student. Using AS to rename the
standardized variables NEW.TEST1 to StdTest1 and NEW.TEST2 to StdTest2 makes the
variable names unique.
proc sql;
create table combined as

The STANDARD Procedure

Example 2: Standardizing BY Groups and Replacing Missing Values

select old.student, old.studentnumber,

old.section,
old.test1, new.test1 as StdTest1,
old.test2, new.test2 as StdTest2,
old.final
from score as old, stndtest as new
where old.student=new.student;

Print the data set. PROC PRINT prints the COMBINED data set. ROUND rounds the
standardized values to two decimal places. The TITLE statement species a title.
proc print data=combined noobs round;
title Standardized Test Scores for a College Course;
run;

Output

Chapter 45

Output

PROC STANDARD prints the variable name, mean, standard deviation, input frequency, and
label of each variable to standardize for each BY group.
Life expectancies for Bangladesh, Mozambique, and Russia are no longer missing. The missing
values are replaced with the given mean (0).

Life Expectancies by Birth Rate

---------------------------- PopulationRate=Stable -----------------------------

Name
Life50
Life93

Mean

Standard
Deviation

Label

67.400000
74.500000

1.854724
4.888763

5
6

1950 life expectancy

1993 life expectancy

----------------------------- PopulationRate=Rapid -----------------------------

Name
Life50
Life93

Mean

Standard
Deviation

Label

42.000000
5.033223
8
1950 life expectancy
59.100000
8.225300
10
1993 life expectancy
Standardized Life Expectancies at Birth
2
by a Countrys Birth Rate
Population
Rate
Stable
Stable
Stable
Stable
Stable
Stable
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid
Rapid

Country
France
Germany
Japan
Russia
United Kingdom
United States
Bangladesh
Brazil
China
Egypt
Ethiopia
India
Indonesia
Mozambique
Philippines
Turkey

Life50

Life93

-0.21567
0.32350
-1.83316
0.00000
0.86266
0.86266
0.00000
1.78812
-0.19868
0.00000
-1.78812
-0.59604
-0.79472
0.00000
1.19208
0.39736

0.51138
0.10228
0.92048
-1.94323
0.30683
0.10228
-0.74161
0.96045
1.32518
0.10942
-1.59265
-0.01216
-0.01216
-1.47107
0.59572
0.83888

1177

CHAPTER

46
The SUMMARY Procedure
Overview: SUMMARY Procedure 1177
Syntax: SUMMARY Procedure 1177
PROC SUMMARY Statement 1178
VAR Statement 1178

Overview: SUMMARY Procedure

The SUMMARY procedure provides data summarization tools that compute
descriptive statistics for variables across all observations or within groups of
observations. The SUMMARY procedure is very similar to the MEANS procedure; for
full syntax details, see Chapter 27, The MEANS Procedure, on page 523. Except for
the differences that are discussed here, all the PROC MEANS information also applies
to PROC SUMMARY.

Syntax: SUMMARY Procedure

PROC SUMMARY <option(s)> <statistic-keyword(s)>;

BY <DESCENDING> variable-1< < DESCENDING> variable-n>
<NOTSORTED>;
CLASS variable(s) </ option(s)>;
FREQ variable;
ID variable(s);
OUTPUT <OUT=SAS-data-set>< output-statistic-specication(s)>
<id-group-specication(s)> <maximum-id-specication(s)>
<minimum-id-specication(s)></ option(s)> ;
TYPES request(s);

1178

PROC SUMMARY Statement

Chapter 46

VAR variable(s)</ WEIGHT=weight-variable>;

WAYS list;
WEIGHT variable;

PROC SUMMARY Statement

PRINT | NOPRINT

species whether PROC SUMMARY displays the descriptive statistics. By default,

PROC SUMMARY produces no display output, but PROC MEANS does produce
display output.
Default: NOPRINT

VAR Statement
Identies the analysis variables and their order in the results.
If you omit the VAR statement, then PROC SUMMARY produces a simple
count of observations, whereas PROC MEANS tries to analyze all the numeric variables
that are not listed in the other statements.
Interaction: If you specify statistics on the PROC SUMMARY statement and the VAR
statement is omitted, then PROC SUMMARY stops processing and an error message is
written to the SAS log.
Default:

1179

CHAPTER

47
The TABULATE Procedure
Overview: TABULATE Procedure 1180
What Does the TABULATE Procedure Do? 1180
Simple Tables 1180
Complex Tables 1181
PROC TABULATE and the Output Delivery System 1182
Terminology: TABULATE Procedure 1183
Syntax: TABULATE Procedure 1186
PROC TABULATE Statement 1187
BY Statement 1196
CLASS Statement 1197
CLASSLEV Statement 1200
FREQ Statement 1201
KEYLABEL Statement 1202
KEYWORD Statement 1202
TABLE Statement 1203
VAR Statement 1211
WEIGHT Statement 1212
Concepts: TABULATE Procedure 1213
Statistics That Are Available in PROC TABULATE 1213
Formatting Class Variables 1214
Formatting Values in Tables 1215
How Using BY-Group Processing Differs from Using the Page Dimension 1215
Calculating Percentages 1216
Calculating the Percentage of the Value of in a Single Table Cell 1216
Using PCTN and PCTSUM 1217
Specifying a Denominator for the PCTN Statistic 1217
Specifying a Denominator for the PCTSUM Statistic 1218
Using Style Elements in PROC TABULATE 1220
What Are Style Elements? 1220
Using the STYLE= Option 1220
Applying Style Attributes to Table Cells 1221
Using a Format to Assign a Style Attribute 1221
Results: TABULATE Procedure 1222
Missing Values 1222
How PROC TABULATE Treats Missing Values 1222
No Missing Values 1223
A Missing Class Variable 1224
Including Observations with Missing Class Variables 1225
Formatting Headings for Observations with Missing Class Variables 1226
Providing Headings for All Categories 1227
Providing Text for Cells That Contain Missing Values 1228

1180

Overview: TABULATE Procedure

Chapter 47

Providing Headings for All Values of a Format 1229

Understanding the Order of Headings with ORDER=DATA 1230
Portability of ODS Output with PROC TABULATE 1231
Examples: TABULATE Procedure 1232
Example 1: Creating a Basic Two-Dimensional Table 1232
Example 2: Specifying Class Variable Combinations to Appear in a Table 1235
Example 3: Using Preloaded Formats with Class Variables 1237
Example 4: Using Multilabel Formats 1242
Example 5: Customizing Row and Column Headings 1244
Example 6: Summarizing Information with the Universal Class Variable ALL 1246
Example 7: Eliminating Row Headings 1249
Example 8: Indenting Row Headings and Eliminating Horizontal Separators 1251
Example 9: Creating Multipage Tables 1253
Example 10: Reporting on Multiple-Response Survey Data 1255
Example 11: Reporting on Multiple-Choice Survey Data 1260
Example 12: Calculating Various Percentage Statistics 1266
Example 13: Using Denominator Denitions to Display Basic Frequency Counts and
Percentages 1269
Example 14: Specifying Style Elements for ODS Output 1279
References 1283

Overview: TABULATE Procedure

What Does the TABULATE Procedure Do?
The TABULATE procedure displays descriptive statistics in tabular format, using
some or all of the variables in a data set. You can create a variety of tables ranging
from simple to highly customized.
PROC TABULATE computes many of the same statistics that are computed by other
descriptive statistical procedures such as MEANS, FREQ, and REPORT. PROC
TABULATE provides
3 simple but powerful methods to create tabular reports
3 exibility in classifying the values of variables and establishing hierarchical
relationships between the variables
3 mechanisms for labeling and formatting variables and procedure-generated
statistics.

Simple Tables
Output 47.1 shows a simple table that was produced by PROC TABULATE. The data
setENERGY on page 1387 contains data on expenditures of energy by two types of
customers, residential and business, in individual states in the Northeast (1) and West
(4) regions of the United States. The table sums expenditures for states within a
geographic division. (The RTS option provides enough space to display the column
headers without hyphenating them.)
options nodate pageno=1 linesize=64
pagesize=40;
proc tabulate data=energy;
class region division type;

The TABULATE Procedure

Complex Tables

1181

var expenditures;
table region*division, type*expenditures /
rts=20;
run;

Output 47.1

Simple Table Produced by PROC TABULATE

The SAS System

---------------------------------------------|
|
Type
|
|
|-------------------------|
|
|
1
|
2
|
|
|------------+------------|
|
|Expenditures|Expenditures|
|
|------------+------------|
|
|
Sum
|
Sum
|
|------------------+------------+------------|
|Region |Division |
|
|
|--------+---------|
|
|
|1
|1
|
7477.00|
5129.00|
|
|---------+------------+------------|
|
|2
|
19379.00|
15078.00|
|--------+---------+------------+------------|
|4
|3
|
5476.00|
4729.00|
|
|---------+------------+------------|
|
|4
|
13959.00|
12619.00|
----------------------------------------------

Complex Tables
Output 47.2 is a more complicated table using the same data set that was used to
create Output 47.1. The statements that create this report

3
3
3
3
3

customize column and row headers

apply a format to all table cells
sum expenditures for residential and business customers
compute subtotals for each division
compute totals for all regions.

For an explanation of the program that produces this report, see Example 6 on page
1246.

1182

PROC TABULATE and the Output Delivery System

Output 47.2

Chapter 47

Complex Table Produced by PROC TABULATE

Energy Expenditures for Each Region
(millions of dollars)

---------------------------------------------------------------|
|
Customer Base
|
|
|
|-------------------------|
|
|
|Residential | Business |
All
|
|
| Customers | Customers | Customers |
|-----------------------+------------+------------+------------|
|Region
|Division
|
|
|
|
|-----------+-----------|
|
|
|
|Northeast |New England|
7,477|
5,129|
12,606|
|
|-----------+------------+------------+------------|
|
|Middle
|
|
|
|
|
|Atlantic
|
19,379|
15,078|
34,457|
|
|-----------+------------+------------+------------|
|
|Subtotal
|
26,856|
20,207|
47,063|
|-----------+-----------+------------+------------+------------|
|West
|Division
|
|
|
|
|
|-----------|
|
|
|
|
|Mountain
|
5,476|
4,729|
10,205|
|
|-----------+------------+------------+------------|
|
|Pacific
|
13,959|
12,619|
26,578|
|
|-----------+------------+------------+------------|
|
|Subtotal
|
19,435|
17,348|
36,783|
|-----------------------+------------+------------+------------|
|Total for All Regions |
$46,291|
$37,555|
$83,846|
----------------------------------------------------------------

PROC TABULATE and the Output Delivery System

Display 47.1 on page 1183 shows a table that is created in Hypertext Markup
Language (HTML). You can use the Output Delivery System with PROC TABULATE to
create customized output in HTML, Rich Text Format (RTF), Portable Document
Format (PDF), and other output formats. For an explanation of the program that
produces this table, see Example 14 on page 1279.

The TABULATE Procedure

Display 47.1

Terminology: TABULATE Procedure

HTML Table Produced by PROC TABULATE

Terminology: TABULATE Procedure

The following gures illustrate some of the terms that are commonly used in
discussions of PROC TABULATE.

1183

1184

Terminology: TABULATE Procedure

Figure 47.1

Chapter 47

Parts of a PROC TABULATE Table

Column headings

Column

The SAS System

-----------------------------------------------|

Type

|-----------------------|

|Residential|

| Customers | Customers |

Business |

|----------------------+-----------+-----------|
|Region

|Division

|----------+-----------|

|Northeast |New England|

$7,477|

$5,129 |

|-----------+-----------+-----------|

|Middle

|Atlantic

$19,379|

$15,078 |

|----------+-----------+-----------+-----------|
|West

|Mountain

|-----------+-----------+-----------|

|Pacific

$5,476|

$13,959|

$4,729 |

$12,619 |

-----------------------------------------------Row

Row headings

Cell

The TABULATE Procedure

Figure 47.2

Terminology: TABULATE Procedure

1185

PROC TABULATE Table Dimensions

The SAS System

Year: 2000

Year: 2001
Year: 2002

page
dimension

The SAS System

column dimension

row dimension

In addition, the following terms frequently appear in discussions of PROC

TABULATE:
category
the combination of unique values of class variables. The TABULATE procedure
creates a separate category for each unique combination of values that exists in
the observations of the data set. Each category that is created by PROC
TABULATE is represented by one or more cells in the table where the pages, rows,
and columns that describe the category intersect.
The table in Figure 47.1 on page 1184 contains three class variables: Region,
Division, and Type. These class variables form the eight categories listed in Table
47.1 on page 1185. (For convenience, the categories are described in terms of their
formatted values.)
Table 47.1 Categories Created from Three Class Variables
Region

Division

Type

Northeast

New England

Residential Customers

Northeast

New England

Business Customers

Northeast

Middle Atlantic

Residential Customers

Northeast

Middle Atlantic

Business Customers

West

Mountain

Residential Customers

West

Mountain

Business Customers

West

Pacic

Residential Customers

West

Pacic

Business Customers

continuation message
the text that appears below the table if it spans multiple physical pages.

1186

Syntax: TABULATE Procedure

Chapter 47

nested variable
a variable whose values appear in the table with each value of another variable.
In Figure 47.1 on page 1184, Division is nested under Region.
page dimension text
the text that appears above the table if the table has a page dimension. However,
if you specify BOX=_PAGE_ in the TABLE statement, then the text that would
appear above the table appears in the box. In Figure 47.2 on page 1185, the word
Year:, followed by the value, is the page dimension text.
Page dimension text has a style. The default style is Beforecaption. For more
information about using styles, see STYLE= on page 1194 in the PROC
TABULATE statement and Output Delivery System on page 32.
subtable
the group of cells that is produced by crossing a single element from each
dimension of the TABLE statement when one or more dimensions contain
concatenated elements.
Figure 47.1 on page 1184 contains no subtables. For an illustration of a table
that is composed of multiple subtables, see Figure 47.18 on page 1274.

Syntax: TABULATE Procedure

At least one TABLE statement is required.
Requirements: Depending on the variables that appear in the TABLE statement, a
CLASS statement, a VAR statement, or both are required.
Tip: Supports the Output Delivery System. See Output Delivery System on page 32
for details.
ODS Table Name: Table
Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See
Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.
Requirements:

PROC TABULATE <option(s)>;

BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
CLASS variable(s) </ options>;
CLASSLEV variable(s) / STYLE=<style-element-name | PARENT>
<[style-attribute-specication(s)] >;
FREQ variable;
KEYLABEL keyword-1=description-1
<keyword-n=description-n>;
KEYWORD keyword(s) / STYLE=<style-element-name | PARENT>
<[style-attribute-specication(s)] >;
TABLE <<page-expression,> row-expression,> column-expression< / table-option(s)>;
VAR analysis-variable(s)< / options>;
WEIGHT variable;

The TABULATE Procedure

PROC TABULATE Statement

To do this

Use this statement

Create a separate table for each BY group

Identify variables in the input data set as class

variables

CLASS

Specify a style for class variable level value headings

CLASSLEV

Identify a variable in the input data set whose values

represent the frequency of each observation

FREQ

Specify a label for a keyword

KEYLABEL

Specify a style for keyword headings

KEYWORD

Describe the table to create

TABLE

Identify variables in the input data set as analysis

variables

VAR

Identify a variable in the input data set whose values

weight each observation in the statistical calculations

WEIGHT

PROC TABULATE Statement

PROC TABULATE <option(s)>;

To do this

Use this option

Customize the HTML contents link to the output

CONTENTS=

Specify the input data set

DATA=

Specify the output data set

OUT=

Override the SAS system option THREADS |

NOTHREADS

THREADS | NOTHREADS

Enable oating point exception recovery

TRAP

Identify categories of data that are of interest

Specify a secondary data set that contains the
combinations of values of class variables to
include in tables and output data sets

CLASSDATA=

Exclude from tables and output data sets all

combinations of class variable values that are
not in the CLASSDATA= data set

EXCLUSIVE

Consider missing values as valid values for class

variables

MISSING

Control the statistical analysis

Specify the condence level for the condence
limits

ALPHA=

1187

1188

PROC TABULATE Statement

Chapter 47

To do this

Use this option

Exclude observations with nonpositive weights

EXCLNPWGTS

Specify the sample size to use for the P quantile

estimation method

QMARKERS=

Specify the quantile estimation method

QMETHOD=

Specify the mathematical denition to calculate

quantiles

QNTLDEF=

Specify the variance divisor

VARDEF=

Customize the appearance of the table

Specify a default format for each cell in the table

FORMAT=

Dene the characters to use to construct the

table outlines and dividers

FORMCHAR=

Eliminate horizontal separator lines from the

row titles and the body of the table

NOSEPS

Order the values of a class variable according to

the specied order

ORDER=

Specify the default style element or style

elements (for the Output Delivery System) to
use for each cell of the table

STYLE=

Options
ALPHA=value

between 0 and 1

Interaction: To compute condence limits specify the statistic-keyword LCLM or

UCLM.
CLASSDATA=SAS-data-set

species a data set that contains the combinations of values of the class variables
that must be present in the output. Any combinations of values of the class variables
that occur in the CLASSDATA= data set but not in the input data set appear in each
table or output data set and have a frequency of zero.
Restriction: The CLASSDATA= data set must contain all class variables. Their

data type and format must match the corresponding class variables in the input
data set.
Interaction: If you use the EXCLUSIVE option, then PROC TABULATE excludes

any observations in the input data set whose combinations of values of class
variables are not in the CLASSDATA= data set.
Tip:

Use the CLASSDATA= data set to lter or supplement the input data set.

Featured in:

Example 2 on page 1235

The TABULATE Procedure

PROC TABULATE Statement

1189

CONTENTS=link-name

enables you to name the link in the HTML table of contents that points to the ODS
output of the rst table that was produced by using the TABULATE procedure.
Note: CONTENTS= affects only the contents le of ODS HTML output. It has no
effect on the actual TABULATE procedure reports. 4
DATA=SAS-data-set

species the input data set.

Main Discussion:

Input Data Sets on page 19

EXCLNPWGTS

excludes observations with nonpositive weight values (zero or negative) from the
analysis. By default, PROC TABULATE treats observations with negative weights
like those with zero weights and counts them in the total number of observations.
Alias:

EXCLNPWGT

See also: WEIGHT= on page 1212 and WEIGHT Statement on page 1212
EXCLUSIVE

excludes from the tables and the output data sets all combinations of the class
variable that are not found in the CLASSDATA= data set.
Requirement:

If a CLASSDATA= data set is not specied, then this option is

ignored.
Featured in:

Example 2 on page 1235

FORMAT=format-name

species a default format for the value in each table cell. You can use any SAS or
user-dened format.
Alias:

Default: If you omit FORMAT=, then PROC TABULATE uses BEST12.2 as the

default format.
Interaction: Formats that are specied in a TABLE statement override the format

that is specied with FORMAT=.

This option is especially useful for controlling the number of print positions
that are used to print a table.

Tip:

Featured in:

Example 1 on page 1232 and Example 6 on page 1246

FORMCHAR <(position(s))>=formatting-character(s)

denes the characters to use for constructing the table outlines and dividers.
position(s)
identies the position of one or more characters in the SAS formatting-character
string. A space or a comma separates the positions.
Default: Omitting position(s) is the same as specifying all 20 possible SAS

formatting characters, in order.

Range: PROC TABULATE uses 11 of the 20 formatting characters that SAS

provides. Table 47.2 on page 1190 shows the formatting characters that PROC
TABULATE uses. Figure 47.3 on page 1191 illustrates the use of each
formatting character in the output from PROC TABULATE.
formatting-character(s)
lists the characters to use for the specied positions. PROC TABULATE assigns
characters in formatting-character(s) to position(s), in the order that they are
listed. For example, the following option assigns the asterisk (*) to the third

1190

PROC TABULATE Statement

Chapter 47

formatting character, the pound sign (#) to the seventh character, and does not
alter the remaining characters:
formchar(3,7)=*#

Interaction: The SAS system option FORMCHAR= species the default formatting

characters. The system option denes the entire string of formatting characters.
The FORMCHAR= option in a procedure can redene selected characters.
Restriction: The FORMCHAR= option affects only the traditional SAS monospace

output destination.
You can use any character in formatting-characters, including hexadecimal
characters. If you use hexadecimal characters, then you must put an x after the
closing quotation mark. For instance, the following option assigns the hexadecimal
character 2D to the third formatting character, assigns the hexadecimal character
7C to the seventh character, and does not alter the remaining characters:

Tip:

formchar(3,7)=2D7Cx

Specifying all blanks for formatting-character(s) produces tables with no

outlines or dividers.

Tip:

formchar(1,2,3,4,5,6,7,8,9,10,11)
=
(11 blanks)

See also: For more information about formatting output, see Chapter 5,

Controlling the Tables Appearance, in the SAS Guide to TABULATE Processing.

For information about which hexadecimal codes to use for which characters,
consult the documentation for your hardware.

Table 47.2

Formatting Characters Used by PROC TABULATE

Position

Default

Used to draw

the right and left borders and the vertical separators

between columns

the top and bottom borders and the horizontal separators

between rows

the top character in the left border

the top character in a line of characters that separate

columns

the top character in the right border

the leftmost character in a row of horizontal separators

the intersection of a column of vertical characters and a

row of horizontal characters

the rightmost character in a row of horizontal separators

the bottom character in the left border

the bottom character in a line of characters that separate

columns

the bottom character in the right border

The TABULATE Procedure

Figure 47.3

PROC TABULATE Statement

1191

Formatting Characters in PROC TABULATE Output

-----------------------------------|
| Expend |
|
|----------|
4
|
|
Sum
|
|-----------------------+----------|
|Region
|Division
|
|
|-----------+-----------|
|
|Northeast |New England|
$12,606|
|
|-----------+----------|
|
|Middle
|
|
|
|Atlantic
|
$34,457|
|-----------+-----------+----------|
|West
|Mountain
|
$10,205|
|
|-----------+----------|
|
|Pacific
|
$26,578|
------------------------------------

MISSING

considers missing values as valid values to create the combinations of class variables.
Special missing values that are used to represent numeric values (the letters A
through Z and the underscore (_) character) are each considered as a separate value.
A heading for each missing value appears in the table.
Default: If you omit MISSING, then PROC TABULATE does not include

observations with a missing value for any class variable in the report.
Main Discussion:

Including Observations with Missing Class Variables on page

1225
See also: SAS Language Reference: Concepts for a discussion of missing values that

have special meaning.

NOSEPS

eliminates horizontal separator lines from the row titles and the body of the table.
Horizontal separator lines remain between nested column headers.
Restriction: The NOSEPS option affects only the traditional SAS monospace

output destination.
If you want to replace the separator lines with blanks rather than remove
them, then use the FORMCHAR= option on page 1189.

Tip:

Featured in:

Example 8 on page 1251

NOTHREADS

See THREADS | NOTHREADS on page 1195.

NOTRAP

See TRAP | NOTRAP on page 1195.

ORDER=DATA | FORMATTED | FREQ | UNFORMATTED

species the sort order to create the unique combinations of the values of the class
variables, which form the headings of the table, according to the specied order.
DATA
orders values according to their order in the input data set.

1192

PROC TABULATE Statement

Chapter 47

Interaction: If you use PRELOADFMT in the CLASS statement, then the order

for the values of each class variable matches the order that PROC FORMAT
uses to store the values of the associated user-dened format. If you use the
CLASSDATA= option, then PROC TABULATE uses the order of the unique
values of each class variable in the CLASSDATA= data set to order the output
levels. If you use both options, then PROC TABULATE rst uses the
user-dened formats to order the output. If you omit EXCLUSIVE, then PROC
TABULATE appends after the user-dened format and the CLASSDATA=
values the unique values of the class variables in the input data set in the same
order in which they are encountered.
Tip: By default, PROC FORMAT stores a format denition in sorted order. Use
the NOTSORTED option to store the values or ranges of a user dened format
in the order that you dene them.
FORMATTED
orders values by their ascending formatted values. If no format has been assigned
to a numeric class variable, then the default format, BEST12., is used. This order
depends on your operating environment.
Alias: FMT | EXTERNAL
FREQ
orders values by descending frequency count.
Interaction: Use the ASCENDING option in the CLASS statement to order
values by ascending frequency count.
UNFORMATTED
orders values by their unformatted values, which yields the same order as PROC
SORT. This order depends on your operating environment. This sort sequence is
particularly useful for displaying dates chronologically.
Alias: UNFMT | INTERNAL
Default: UNFORMATTED
Interaction: If you use the PRELOADFMT option in the CLASS statement, then
PROC TABULATE orders the levels by the order of the values in the user-dened
format.
Featured in: Understanding the Order of Headings with ORDER=DATA on page
1230
OUT=SAS-data-set

names the output data set. If SAS-data-set does not exist, then PROC TABULATE
creates it.
The number of observations in the output data set depends on the number of
categories of data that are used in the tables and the number of subtables that are
generated. The output data set contains these variables (in this order):
by variables
variables that are listed in the BY statement.
class variables
variables that are listed in the CLASS statement.
_TYPE_
a character variable that shows which combination of class variables produced the
summary statistics in that observation. Each position in _TYPE_ represents one
variable in the CLASS statement. If that variable is in the category that produced
the statistic, then the position contains a 1; if it is not, then the position contains a
0. In simple PROC TABULATE steps that do not use the universal class variable
ALL, all values of _TYPE_ contain only 1s because the only categories that are

The TABULATE Procedure

PROC TABULATE Statement

1193

being considered involve all class variables. If you use the variable ALL, then your
tables will contain data for categories that do not include all the class variables,
and positions of _TYPE_ will, therefore, include both 1s and 0s.
_PAGE_
The logical page that contains the observation.
_TABLE_
The number of the table that contains the observation.
statistics
statistics that are calculated for each observation in the data set.
Featured in:

Example 3 on page 1237

PCTLDEF=

See QNTLDEF= on page 1193.

QMARKERS=number
2

species the default number of markers to use for the P quantile estimation method.
The number of markers controls the size of xed memory space.
Default: The default value depends on which quantiles you request. For the median

(P50), number is 7. For the quartiles (P25 and P75), number is 25. For the
quantiles P1, P5, P10, P90, P95, or P99, number is 105. If you request several
quantiles, then PROC TABULATE uses the largest default value of number.
Range: an odd integer greater than 3

Increase the number of markers above the default settings to improve the
accuracy of the estimates; reduce the number of markers to conserve memory and
computing time.

Tip:

Main Discussion:

Quantiles on page 555

QMETHOD=OS|P2|HIST

species the method PROC TABULATE uses to process the input data when it
computes quantiles. If the number of observations is less than or equal to the
QMARKERS= value and QNTLDEF=5, then both methods produce the same results.
OS
uses order statistics. This is the technique that PROC UNIVARIATE uses.
Note: This technique can be very memory-intensive.

P2|HIST
2
uses the P method to approximate the quantile.
Default: OS
Restriction: When QMETHOD=P2, PROC TABULATE does not compute weighted

quantiles.
When QMETHOD=P2, reliable estimates of some quantiles (P1, P5, P95, P99)
may not be possible for some types of data.

Tip:

Main Discussion:

Quantiles on page 555

QNTLDEF=1|2|3|4|5

species the mathematical denition that the procedure uses to calculate quantiles
when QMETHOD=OS is specied. When QMETHOD=P2, you must use
QNTLDEF=5.
Default: 5
Alias:

PCTLDEF=

Main discussion:

Quantile and Related Statistics on page 1345

1194

PROC TABULATE Statement

Chapter 47

STYLE=<style-element-name|PARENT>[style-attribute-name=style-attributevalue< style-attribute-name=style-attribute-value>]

species the style element to use for the data cells of a table when it is used in the
PROC TABULATE statement. For example, the following statement species that
the background color for data cells be red:
proc tabulate data=one style=[background=red];

Note: This option can be used in other statements, or in dimension expressions,

to specify style elements for other parts of a table. 4
Note:

You can use braces ({ and }) instead of square brackets ([ and ]).

style-element-name
is the name of a style element that is part of a style denition that is registered
with the Output Delivery System. SAS provides some style denitions. You can
create your own style denitions with PROC TEMPLATE.
Default: If you do not specify a style element, then PROC TABULATE uses Data.
See also: See SAS Output Delivery System: Users Guide for information about
PROC TEMPLATE and the default style denitions.
PARENT
species that the data cell use the style element of its parent heading. The parent
style element of a data cell is one of the following:
3 the style element of the leaf heading above the column that contains the data
cell, if the table species no row dimension, or if the table species the style
element in the column dimension expression.
3 the style element of the leaf heading above the row that contains the cell, if
the table species the style element in the row dimension expression.
3 the Beforecaption style element, if the table species the style element in the
page dimension expression.
3 undened, otherwise.
Note: The parent of a heading (not applicable to STYLE= in the PROC
TABULATE statement) is the heading under which the current heading is
nested. 4
style-attribute-name
species the attribute to change. The following table shows attributes that you can
set or change with the STYLE= option in the PROC TABULATE statement (or in
any other statement that uses STYLE=, except for the TABLE statement). Note
that not all attributes are valid in all destinations.
ASIS=

FONT_WIDTH=

BACKGROUND=

HREFTARGET=

BACKGROUNDIMAGE=

HTMLCLASS=

BORDERCOLOR=

JUST=

BORDERCOLORDARK=

NOBREAKSPACE=

BORDERCOLORLIGHT=

POSTHTML=

BORDERWIDTH=

POSTIMAGE=

CELLHEIGHT=

POSTTEXT=

CELLWIDTH=

PREHTML=

FLYOVER=

PREIMAGE=

The TABULATE Procedure

PROC TABULATE Statement

FONT=

PRETEXT=

FONT_FACE=

PROTECTSPECIALCHARS=

FONT_SIZE=

TAGATTR=

FONT_STYLE=

URL=

FONT_WEIGHT=

1195

VJUST=

style-attribute-value
species a value for the attribute. Each attribute has a different set of valid
values. See SAS Output Delivery System: Users Guide for more information about
these style attributes, their valid values, and their applicable destinations.
Alias:

Restriction: This option affects only the HTML, RTF, and Printer destinations.

To specify a style element for data cells with missing values, use STYLE= in
the TABLE statement MISSTEXT= option.

Tip:

Example 14 on page 1279

THREADS | NOTHREADS

except when a BY statement is specied or the value of the SAS system option
CPUCOUNT is equal to 1. In those cases, you can use THREADS in the PROC
TABULATE statement to force PROC TABULATE to use parallel processing.
TRAP | NOTRAP

enables or disables oating point exception (FPE) recovery during data processing
beyond that provided by normal SAS FPE handling, which terminates PROC
TABULATE in the case of math exceptions. Note that with NOTRAP, normal SAS
FPE handling is still in effect so that PROC TABULATE terminates in the case of
math exceptions.
Default: NOTRAP
VARDEF=divisor

species the divisor to use in the calculation of the variance and standard deviation.
Table 47.3 on page 1195 shows the possible values for divisor and the associated
divisors.
Table 47.3

Possible Values for VARDEF=

Value

Divisor

degrees of freedom

number of observations

WDF

sum of weights minus one

WEIGHT | WGT

sum of weights

Formula for Divisor

n1
n
(6i wi) 1

6i wi

1196

BY Statement

Chapter 47

P x 0x

CSS=divisor

CSS

The procedure computes the variance as

, where
is the corrected
2
) . When you weight the analysis variables,
sums of squares and equals
( i
2
) where w is the weighted mean.
( i
equals
i
w

CSS

x 0x

Default: DF
Requirement:

To compute standard error of the mean, use the default value of

VARDEF=.
When you use the WEIGHT statement and VARDEF=DF, the variance is an
( i) = 2
estimate of 2 , where the variance of the ith observation is
i , and
i is the weight for the ith observation. This yields an estimate of the variance of
an observation with unit weight.

Tip:

var x

When you use the WEIGHT statement and VARDEF=WGT, the computed
, where
is the
variance is asymptotically (for large n) an estimate of 2
average weight. This yields an asymptotic estimate of the variance of an
observation with average weight.

Tip:

BY < DESCENDING> variable-1

<< DESCENDING> variable-n>
<NOTSORTED>;

Required Arguments
variable

Options
DESCENDING

species that the observations are sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

The TABULATE Procedure

CLASS Statement

1197

option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations with the same values for the BY variables
are not contiguous, then the procedure treats each contiguous set as a separate BY
group.

CLASS Statement
Identies class variables for the table. Class variables determine the categories that PROC
TABULATE uses to calculate statistics.
You can use multiple CLASS statements.
Tip: Some CLASS statement options are also available in the PROC TABULATE
statement. They affect all CLASS variables rather than just the one(s) that you specify
in a CLASS statement.
Tip:

CLASS variable(s) </option(s)>;

Required Arguments
variable(s)

species one or more variables that the procedure uses to group the data. Variables
in a CLASS statement are referred to as class variables. Class variables can be
numeric or character. Class variables can have continuous values, but they typically
have a few discrete values that dene the classications of the variable. You do not
have to sort the data by class variables.

Options
ASCENDING

species to sort the class variable values in ascending order.

Alias: ASCEND
Interaction: PROC TABULATE issues a warning message if you specify both
ASCENDING and DESCENDING and ignores both options.
DESCENDING

species to sort the class variable values in descending order.

Alias: DESCEND
Default: ASCENDING
Interaction: PROC TABULATE issues a warning message if you specify both
ASCENDING and DESCENDING and ignores both options.
EXCLUSIVE

excludes from tables and output data sets all combinations of class variables that are
not found in the preloaded range of user-dened formats.
Requirement: You must specify the PRELOADFMT option in the CLASS statement
to preload the class variable formats.

1198

CLASS Statement

Chapter 47

Featured in:

Example 3 on page 1237

GROUPINTERNAL

species not to apply formats to the class variables when PROC TABULATE groups
the values to create combinations of class variables.
Interaction: If you specify the PRELOADFMT option in the CLASS statement,

then PROC TABULATE ignores the GROUPINTERNAL option and uses the
formatted values.
Interaction: If you specify the ORDER=FORMATTED option, then PROC
TABULATE ignores the GROUPINTERNAL option and uses the formatted values.
This option saves computer resources when the class variables contain discrete
numeric values.

Tip:

MISSING

considers missing values as valid class variable levels. Special missing values that
represent numeric values (the letters A through Z and the underscore (_) character)
are each considered as a separate value.
Default: If you omit MISSING, then PROC TABULATE excludes the observations

with any missing CLASS variable values from tables and output data sets.
See also: SAS Language Reference: Concepts for a discussion of missing values with

special meanings.
MLF

enables PROC TABULATE to use the format label or labels for a given range or
overlapping ranges to create subgroup combinations when a multilabel format is
assigned to a class variable.
You must use PROC FORMAT and the MULTILABEL option in the
VALUE statement to create a multilabel format.
Interaction: Using MLF with ORDER=FREQ may not produce the order that you
expect for the formatted values.
Requirement:

Interaction: When you specify MLF, the formatted values of the class variable

become internal values. Therefore, specifying ORDER=FORMATTED produces the

same results as specifying ORDER=UNFORMATTED.
Tip: If you omit MLF, then PROC TABULATE uses the primary format labels,
which correspond to the rst external format value, to determine the subgroup
combinations.
See also: The MULTILABEL option on page 449 in the VALUE statement of the

FORMAT procedure.
Featured in:

Example 4 on page 1242

species the order to group the levels of the class variables in the output, where
DATA
orders values according to their order in the input data set.
Interaction: If you use PRELOADFMT, then the order for the values of each class

variable matches the order that PROC FORMAT uses to store the values of the
associated user-dened format. If you use the CLASSDATA= option in the
PROC statement, then PROC TABULATE uses the order of the unique values of
each class variable in the CLASSDATA= data set to order the output levels. If

The TABULATE Procedure

CLASS Statement

1199

you use both options, then PROC TABULATE rst uses the user-dened
formats to order the output. If you omit EXCLUSIVE in the PROC statement,
then PROC TABULATE places, in the order in which they are encountered, the
unique values of the class variables that are in the input data set after the
user-dened format and the CLASSDATA= values.
Tip: By default, PROC FORMAT stores a format denition in sorted order. Use

the NOTSORTED option to store the values or ranges of a user-dened format

in the order that you dene them.
FORMATTED
orders values by their ascending formatted values. This order depends on your
operating environment.
Alias: FMT | EXTERNAL

FREQ
orders values by descending frequency count.
Interaction: Use the ASCENDING option to order values by ascending frequency

count.
UNFORMATTED
orders values by their unformatted values, which yields the same order as PROC
SORT. This order depends on your operating environment. This sort sequence is
particularly useful for displaying dates chronologically.
Alias: UNFMT | INTERNAL
Default: UNFORMATTED
Interaction: If you use the PRELOADFMT option in the CLASS statement, then

PROC TABULATE orders the levels by the order of the values in the user-dened
format.
By default, all orders except FREQ are ascending. For descending orders, use
the DESCENDING option.

Tip:

Featured in:

Understanding the Order of Headings with ORDER=DATA on page

1230
PRELOADFMT

species that all formats are preloaded for the class variables.
PRELOADFMT has no effect unless you specify EXCLUSIVE,
ORDER=DATA, or PRINTMISS and you assign formats to the class variables.

Requirement:

Note: If you specify PRELOADFMT without also specifying EXCLUSIVE,

ORDER=DATA, or PRINTMISS, then SAS writes a warning message to the SAS
log. 4
Interaction: To limit PROC TABULATE output to the combinations of formatted

class variable values present in the input data set, use the EXCLUSIVE option in
the CLASS statement.
Interaction: To include all ranges and values of the user-dened formats in the

output, use the PRINTMISS option in the TABLE statement.

Note: Use care when you use PRELOADFMT with PRINTMISS. This feature
creates all possible combinations of formatted class variables. Some of these
combinations may not make sense. 4
Featured in:

Example 3 on page 1237

1200

CLASSLEV Statement

Chapter 47

STYLE=<style-element-name|PARENT>[style-attribute-name=style-attributevalue< style-attribute-name=style-attribute-value>]

species the style element to use for page dimension text and class variable name
headings. For information about the arguments of this option, and how it is used, see
STYLE= on page 1194 in the PROC TABULATE statement.
Note: When you use STYLE= in the CLASS statement, it differs slightly from its
use in the PROC TABULATE statement. In the CLASS statement, the parent of the
heading is the page dimension text or heading under which the current heading is
nested. 4
Note: If a page dimension expression contains multiple nested elements, then the
Beforecaption style element is the style element of the rst element in the nesting. 4
Alias:

Restriction: This option affects only the HTML, RTF, and Printer destinations.

To override a style element that is specied for page dimension text in the
CLASS statement, you can specify a style element in the TABLE statement page
dimension expression.

Tip:

To override a style element that is specied for a class variable name heading
in the CLASS statement, you can specify a style element in the related TABLE
statement dimension expression.

Tip:

Featured in:

Example 14 on page 1279

How PROC TABULATE Handles Missing Values for Class Variables

By default, if an observation contains a missing value for any class variable, then
PROC TABULATE excludes that observation from all tables that it creates. CLASS
statements apply to all TABLE statements in the PROC TABULATE step. Therefore, if
you dene a variable as a class variable, then PROC TABULATE omits observations
that have missing values for that variable from every table even if the variable does not
appear in the TABLE statement for one or more tables.
If you specify the MISSING option in the PROC TABULATE statement, then the
procedure considers missing values as valid levels for all class variables. If you specify
the MISSING option in a CLASS statement, then PROC TABULATE considers missing
values as valid levels for the class variable(s) that are specied in that CLASS
statement.

CLASSLEV Statement
Species a style element for class variable level value headings.
Restriction:

This statement affects only the HTML, RTF, and Printer destinations.

CLASSLEV variable(s) / STYLE=<style-element-name | PARENT>

[style-attribute-name=style-attribute-value<
style-attribute-name=style-attribute-value>] ;

The TABULATE Procedure

FREQ Statement

1201

Required Arguments
variable(s)

species one or more class variables from the CLASS statement for which you want
to specify a style element.

Options
STYLE=<style-element-name|PARENT>[style-attribute-name=style-attributevalue< style-attribute-name=style-attribute-value>]

species a style element for class variable level value headings. For information on
the arguments of this option and how it is used, see STYLE= on page 1194 in the
PROC TABULATE statement.
Note: When you use STYLE= in the CLASSLEV statement, it differs slightly
from its use in the PROC TABULATE statement. In the CLASSLEV statement, the
parent of the heading is the heading under which the current heading is nested. 4
Alias:

Restriction: This option affects only the HTML, RTF, and Printer destinations.

To override a style element that is specied in the CLASSLEV statement, you

can specify a style element in the related TABLE statement dimension expression.

Tip:

Featured in:

Example 14 on page 1279

FREQ Statement
Species a numeric variable that contains the frequency of each observation.
The effects of the FREQ and WEIGHT statements are similar except when
calculating degrees of freedom.

Tip:

See also: For an example that uses the FREQ statement, see FREQ on page 61.

FREQ variable;

Required Arguments
variable

1202

KEYLABEL Statement

Chapter 47

KEYLABEL Statement
Labels a keyword for the duration of the PROC TABULATE step. PROC TABULATE uses the label
anywhere that the specied keyword would otherwise appear.
KEYLABEL keyword-1=description-1
<keyword-n=description-n>;

Required Arguments
keyword

is one of the keywords for statistics that is discussed in Statistics That Are Available
in PROC TABULATE on page 1213 or is the universal class variable ALL (see
Elements That You Can Use in a Dimension Expression on page 1208).
description

is up to 256 characters to use as a label. As the syntax shows, you must enclose
description in quotation marks.
Restriction: Each keyword can have only one label in a particular PROC

TABULATE step; if you request multiple labels for the same keyword, then PROC
TABULATE uses the last one that is specied in the step.

KEYWORD Statement
Species a style element for keyword headings.
Restriction:

This statement affects only the HTML, RTF, and Printer output.

KEYWORD keyword(s) / STYLE=<style-element-name | PARENT>

[style-attribute-name=style-attribute-value<
style-attribute-name=style-attribute-value>] ;

Required Arguments
keyword

Options
STYLE=<style-element-name|PARENT>[style-attribute-name=style-attributevalue< style-attribute-name=style-attribute-value>]

The TABULATE Procedure

TABLE Statement

1203

species a style element for the keyword headings. For information on the
arguments of this option and how it is used, see STYLE= on page 1194 in the PROC
TABULATE statement.
Note: When you use STYLE= in the KEYWORD statement, it differs slightly
from its use in the PROC TABULATE statement. In the KEYWORD statement, the
parent of the heading is the heading under which the current heading is nested. 4
Alias: S=
Restriction: This option affects only the HTML, RTF, and Printer destinations.
Tip: To override a style element that is specied in the KEYWORD statement, you
can specify a style element in the related TABLE statement dimension expression.
Featured in: Example 14 on page 1279

TABLE Statement
Describes a table to print.
Requirement: All variables in the TABLE statement must appear in either the VAR
statement or the CLASS statement.
Tip: Use multiple TABLE statements to create several tables.

TABLE <<page-expression,> row-expression,>

column-expression </ table-option(s)>;

Required Arguments
column-expression

denes the columns in the table. For information on constructing dimension

expressions, see Constructing Dimension Expressions on page 1208.
Restriction: A column dimension is the last dimension in a TABLE statement. A
row dimension or a row dimension and a page dimension may precede a column
dimension.

Options
To do this

Use this option

Add dimensions
Dene the pages in a table

page-expression

Dene the rows in a table

row-expression

Customize the HTML contents entry link to the output

CONTENTS=

Modify the appearance of the table

Change the order of precedence for specied format modiers

FORMAT_PRECEDENCE=

Specify a style element for various parts of the table

STYLE=

1204

TABLE Statement

Chapter 47

To do this

Use this option

Change the order of precedence for specied style attribute

values

STYLE_PRECEDENCE=

Customize text in the table

Specify the text to place in the empty box above row titles

BOX=

Supply up to 256 characters to print in table cells that contain

missing values

MISSTEXT=

Suppress the continuation message for tables that span

multiple physical pages

NOCONTINUED

Modify the layout of the table

Print as many complete logical pages as possible on a single
printed page or, if possible, print multiple pages of tables that
are too wide to t on a page one below the other on a single
page, instead of on separate pages.

CONDENSE

Create the same row and column headings for all logical pages
of the table

PRINTMISS

Customize row headings

Specify the number of spaces to indent nested row headings

INDENT=

Control allocation of space for row titles within the available

space

ROW=

Specify the number of print positions available for row titles

RTSPACE=

BOX=value
BOX={<label=value>
<STYLE=<style-element-name>[style-attribute-name=style-attribute-value<
style-attribute-name=style-attribute-value>]>}

species text and a style element for the empty box above the row titles.
Value can be one of the following:
_PAGE_
writes the page-dimension text in the box. If the page-dimension text does not t,
then it is placed in its default position above the box, and the box remains empty.
string
writes the quoted string in the box. Any string that does not t in the box is
truncated.
variable
writes the name (or label, if the variable has one) of a variable in the box. Any
name or label that does not t in the box is truncated.
For details about the arguments of the STYLE= option and how it is used, see
STYLE= on page 1194 in the PROC TABULATE statement.
Featured in:

Example 9 on page 1253 and Example 14 on page 1279

CONDENSE

prints as many complete logical pages as possible on a single printed page or, if
possible, prints multiple pages of tables that are too wide to t on a page one below

The TABULATE Procedure

TABLE Statement

1205

the other on a single page, instead of on separate pages. A logical page is all the
rows and columns that fall within one of the following:
3 a page-dimension category (with no BY-group processing)
3 a BY group with no page dimension

3 a page-dimension category within a single BY group.

CONDENSE has no effect on the pages that are generated by the BY
statement. The rst table for a BY group always begins on a new page.
Featured in: Example 9 on page 1253
Restrictions:

CONTENTS=link-name

enables you to name the link in the HTML table of contents that points to the ODS
output of the table that is produced by using the TABLE statement.
Note: CONTENTS= affects only the contents le of ODS HTML output. It has no
effect on the actual TABULATE procedure reports. 4
FORMAT_PRECEDENCE=PAGE|ROW|COLUMN|COL

species whether the format that is specied for the page dimension (PAGE), row
dimension (ROW), or column dimension (COLUMN or COL) is applied to the
contents of the table cells.
Default: COLUMN
FUZZ=number

supplies a numeric value against which analysis variable values and table cell values
other than frequency counts are compared to eliminate trivial values (absolute values
less than the FUZZ= value) from computation and printing. A number whose
absolute value is less than the FUZZ= value is treated as zero in computations and
printing. The default value is the smallest representable oating-point number on
the computer that you are using.
INDENT=number-of-spaces

species the number of spaces to indent nested row headings, and suppresses the
row headings for class variables.
Tip: When there are no crossings in the row dimension, there is nothing to indent,
so the value of number-of-spaces has no effect. However, in such cases INDENT=
still suppresses the row headings for class variables.
Restriction: In the HTML, RTF, and Printer destinations, the INDENT= option
suppresses the row headings for class variables but does not indent nested row
headings.
Featured in: Example 8 on page 1251 (with crossings) and Example 9 on page 1253
(without crossings)

MISSTEXT=text
MISSTEXT={<label= text> <STYLE=<style-element-name>
[style-attribute-name=style-attribute-value<
style-attribute-name=style-attribute-value>]>}
supplies up to 256 characters of text to print and species a style element for table
cells that contain missing values. For details on the arguments of the STYLE= option
and how it is used, see STYLE= on page 1194 in the PROC TABULATE statement.
Interaction: A style element that is specied in a dimension expression overrides a
style element that is specied in the MISSTEXT= option for any given cell(s).
Providing Text for Cells That Contain Missing Values on page 1228
and Example 14 on page 1279

Featured in:

1206

TABLE Statement

Chapter 47

NOCONTINUED

suppresses the continuation message, continued, that is displayed at the bottom of

tables that span multiple pages. The text is rendered with the Aftercaption style
element.
Note: Because HTML browsers do not break pages, NOCONTINUED has no
effect on the HTML destination. 4
page-expression

denes the pages in a table. For information on constructing dimension expressions,

see Constructing Dimension Expressions on page 1208.
Restriction: A page dimension is the rst dimension in a table statement. Both a

row dimension and a column dimension must follow a page dimension.

Example 9 on page 1253

Featured in:
PRINTMISS

prints all values that occur for a class variable each time headings for that variable
are printed, even if there are no data for some of the cells that these headings create.
Consequently, PRINTMISS creates row and column headings that are the same for
all logical pages of the table, within a single BY group.
Default: If you omit PRINTMISS, then PROC TABULATE suppresses a row or
column for which there are no data, unless you use the CLASSDATA= option in
the PROC TABULATE statement.
Restrictions: If an entire logical page contains only missing values, then that page
does not print regardless of the PRINTMISS option.
See also: CLASSDATA= option on page 1188
Featured in: Providing Headings for All Categories on page 1227
ROW=spacing

species whether all title elements in a row crossing are allotted space even when
they are blank. The possible values for spacing are as follows:
CONSTANT
allots space to all row titles even if the title has been blanked out (for example,
N= ).
Alias: CONST
FLOAT
divides the row title space equally among the nonblank row titles in the crossing.
Default: CONSTANT
Featured in: Example 7 on page 1249
row-expression

denes the rows in the table. For information on constructing dimension expressions,
see Constructing Dimension Expressions on page 1208.
Restriction: A row dimension is the next to last dimension in a table statement. A
column dimension must follow a row dimension. A page dimension may precede a
row dimension.
RTSPACE=number

species the number of print positions to allot to all of the headings in the row
dimension, including spaces that are used to print outlining characters for the row
headings. PROC TABULATE divides this space equally among all levels of row
headings.
Alias: RTS=
Default: one-fourth of the value of the SAS system option LINESIZE=

The TABULATE Procedure

TABLE Statement

1207

Restriction: The RTSPACE= option affects only the traditional SAS monospace

output destination.
Interaction: By default, PROC TABULATE allots space to row titles that are blank.

Use ROW=FLOAT in the TABLE statement to divide the space among only
nonblank titles.
See also: For more information about controlling the space for row titles, see

Chapter 5, Controlling the Tables Appearance, in SAS Guide to TABULATE

Processing.
Featured in:

Example 1 on page 1232

STYLE=<style-element-name> [style-attribute-name=style-attribute-value<
style-attribute-name=style-attribute-value>]

species a style element to use for parts of the table other than table cells. For
information about the arguments of this option and how it is used, see STYLE= on
page 1194 in the PROC TABULATE statement.
Note: The list of attributes that you can set or change with the STYLE= option in
the TABLE statement differs from that of the PROC TABULATE statement. 4
The following table shows the attributes that you can set or change with the
STYLE= option in the TABLE statement. Most of these attributes apply to parts of
the table other than cells (for example, table borders and the lines between columns
and rows). Attributes that you apply in the PROC TABULATE statement and in
other locations in the PROC TABULATE step apply to cells within the table. Note
that not all attributes are valid in all destinations. See SAS Output Delivery System:
Users Guide for more information about these style attributes, their valid values,
and their applicable destinations.
BACKGROUND=

FONT_WIDTH=*

BACKGROUNDIMAGE=

FOREGROUND=*

BORDERCOLOR=

FRAME=

BORDERCOLORDARK=

HTMLCLASS=

BORDERCOLORLIGHT=

JUST=

BORDERWIDTH=

OUTPUTWIDTH=

CELLPADDING=

POSTHTML=

CELLSPACING=

POSTIMAGE=

FONT=*

POSTTEXT=
*

FONT_FACE=
*

PREHTML=

FONT_SIZE=

PREIMAGE=

FONT_STYLE=*

PRETEXT=

FONT_WEIGHT=*

RULES=

Note: You can use braces ({ and }) instead of square brackets ([ and ]).
Alias:

1208

TABLE Statement

Chapter 47

Restriction: This option affects only the HTML, RTF, and Printer destinations.

To override a style element specication that is made as an option in the

TABLE statement, specify STYLE= in a dimension expression of the TABLE
statement.

Tip:

Featured in:

Example 14 on page 1279

STYLE_PRECEDENCE=PAGE|ROW|COLUMN|COL

species whether the style that is specied for the page dimension (PAGE), row
dimension (ROW), or column dimension (COLUMN or COL) is applied to the
contents of the table cells.
Default: COLUMN

Constructing Dimension Expressions

What Are Dimension Expressions?
A dimension expression denes the content and appearance of a dimension (the
columns, rows, or pages in the table) by specifying the combination of variables,
variable values, and statistics that make up that dimension. A TABLE statement
consists of from one to three dimension expressions separated by commas. Options can
follow the dimension expressions.
If all three dimensions are specied, then the leftmost dimension expression denes
pages, the middle dimension expression denes rows, and the rightmost dimension
expression denes columns. If two dimensions are specied, then the left dimension
expression denes rows, and the right dimension expression denes columns. If a single
dimension is specied, then the dimension expression denes columns.
A dimension expression is composed of one or more elements and operators.

Elements That You Can Use in a Dimension Expression

analysis variables
(see VAR Statement on page 1211).
class variables
(see CLASS Statement on page 1197).
the universal class variable ALL
summarizes all of the categories for class variables in the same parenthetical
group or dimension (if the variable ALL is not contained in a parenthetical group).
Featured in: Example 6 on page 1246, Example 9 on page 1253, and Example 13

on page 1269
Note: If the input data set contains a variable named ALL, then enclose the
name of the universal class variable in quotation marks. 4
keywords for statistics
See Statistics That Are Available in PROC TABULATE on page 1213 for a list of
available statistics. Use the asterisk (*) operator to associate a statistic keyword
with a variable. The N statistic (number of nonmissing values) can be specied in
a dimension expression without associating it with a variable.
Restriction: Statistic keywords other than N must be associated with an

analysis variable.
Default: For analysis variables, the default statistic is SUM. Otherwise, the

default statistic is N.

The TABULATE Procedure

TABLE Statement

1209

Examples:
n
Region*n
Sales*max

Featured in: Example 10 on page 1255 and Example 13 on page 1269

format modiers
dene how to format values in cells. Use the asterisk (*) operator to associate a
format modier with the element (an analysis variable or a statistic) that produces
the cells that you want to format. Format modiers have the form
f=format

Example:
Sales*f=dollar8.2

Tip:

Format modiers have no effect on CLASS variables.

For more information on specifying formats in tables, see Formatting

Values in Tables on page 1215.

See also:

Featured in: Example 6 on page 1246

labels
temporarily replace the names of variables and statistics. Labels affect only the
variable or statistic that immediately precedes the label. Labels have the form
statistic-keyword-or-variable-name=label-text

PROC TABULATE eliminates the space for blank column headings from a
table but by default does not eliminate the space for blank row headings unless
all row headings are blank. Use ROW=FLOAT in the TABLE statement to
remove the space for blank row headings.

Tip:

Examples:
Region=Geographical Region
Sales*max=Largest Sale

Featured in: Example 5 on page 1244 and Example 7 on page 1249

style-element specications
specify style elements for page dimension text, headings, or data cells. For details,
see Specifying Style Elements in Dimension Expressions on page 1210.

Operators That You Can Use in a Dimension Expression

asterisk *
creates categories from the combination of values of the class variables and
constructs the appropriate headers for the dimension. If one of the elements is an
analysis variable, then the statistics for the analysis variable are calculated for the
categories that are created by the class variables. This process is called crossing.
Examples:
Region*Division
Quarter*Sales*f=dollar8.2

Featured in: Example 1 on page 1232

(blank)
places the output for each element immediately after the output for the preceding
element. This process is called concatenation.

1210

TABLE Statement

Chapter 47

Example:
n Region*Sales ALL

Featured in:

Example 6 on page 1246

parentheses ()
group elements and associate an operator with each concatenated element in the
group.
Examples:
Division*(Sales*max Sales*min)
(Region ALL)*Sales

Featured in:

Example 6 on page 1246

angle brackets <>

specify denominator denitions, which determine the value of the denominator in
the calculation of a percentage. For a discussion of how to construct denominator
denitions, see Calculating Percentages on page 1216.
Featured in:

Example 10 on page 1255 and Example 13 on page 1269

Specifying Style Elements in Dimension Expressions

You can specify a style element in a dimension expression to control the appearance in
HTML, RTF, and Printer output of the following table elements:
analysis variable name headings
class variable name headings
class variable level value headings
data cells
keyword headings
page dimension text
Specifying a style element in a dimension expression is useful when you want to
override a style element that you have specied in another statement, such as the
PROC TABULATE, CLASS, CLASSLEV, KEYWORD, TABLE, or VAR statements.
The syntax for specifying a style element in a dimension expression is
[STYLE< (CLASSLEV)>=< style-element-name |
PARENT>[style-attribute-name=style-attribute-value<
style-attribute-name=style-attribute-value>]]
Some examples of style elements in dimension expressions are
dept={label=Department
style=[foreground=red]}, N
dept*[style=MyDataStyle], N
dept*[format=12.2 style=MyDataStyle], N

Note: When used in a dimension expression, the STYLE= option must be enclosed
within square brackets ([ and ]) or braces ({ and }). 4
With the exception of (CLASSLEV), all arguments are described in STYLE= on page
1194 in the PROC TABULATE statement.

The TABULATE Procedure

VAR Statement

1211

(CLASSLEV)
assigns a style element to a class variable level value heading. For example, the
following TABLE statement species that the level value heading for the class
variable, DEPT, has a foreground color of yellow:
table dept=[style(classlev)=
[foreground=yellow]]*sales;

Note: This option is used only in dimension expressions.

For an example that shows how to specify style elements within dimension
expressions, see Example 14 on page 1279.

VAR Statement
Identies numeric variables to use as analysis variables.
Alias:
Tip:

VARIABLES
You can use multiple VAR statements.

VAR analysis-variable(s) </ option(s)>;

Required Arguments
analysis-variable(s);

identies the analysis variables in the table. Analysis variables are numeric
variables for which PROC TABULATE calculates statistics. The values of an analysis
variable can be continuous or discrete.
If an observation contains a missing value for an analysis variable, then PROC
TABULATE omits that value from calculations of all statistics except N (the number
of observations with nonmissing variable values) and NMISS (the number of
observations with missing variable values). For example, the missing value does not
increase the SUM, and it is not counted when you are calculating statistics such as
the MEAN.

Options
STYLE=<style-element-name|PARENT>[style-attribute-name=style-attributevalue< style-attribute-name=style-attribute-value>]

species a style element for analysis variable name headings. For information on the
arguments of this option and how it is used, see STYLE= on page 1194 in the PROC
TABULATE statement.
Note: When you use STYLE= in the VAR statement, it differs slightly from its
use in the PROC TABULATE statement. In the VAR statement, the parent of the
heading is the heading under which the current heading is nested. 4

1212

WEIGHT Statement

Alias:

Chapter 47

Restriction: This option affects only the HTML, RTF, and Printer destinations.

To override a style element that is specied in the VAR statement, you can
specify a style element in the related TABLE statement dimension expression.
Featured in: Example 14 on page 1279
Tip:

WEIGHT=weight-variable

PROC TABULATE

counts the observation in the total number of observations

less than 0

converts the value to zero and counts the observation in the total
number of observations

missing

excludes the observation

To exclude observations that contain negative and zero weights from the analysis,
use EXCLNPWGT. Note that most SAS/STAT procedures, such as PROC GLM,
exclude negative and zero weights by default.
Restriction: To compute weighted quantiles, use QMETHOD=OS in the PROC
statement.
Tip: When you use the WEIGHT= option, consider which value of the VARDEF=
option is appropriate (see the discussion of VARDEF= on page 1195).
Tip: Use the WEIGHT option in multiple VAR statements to specify different
weights for the analysis variables.
Note: Prior to Version 7 of SAS, the procedure did not exclude the observations
with missing weights from the count of observations. 4

WEIGHT Statement
Species weights for analysis variables in the statistical calculations.
See also: For information on calculating weighted statistics and for an example that
uses the WEIGHT statement, see Calculating Weighted Statistics on page 64

WEIGHT variable;

Required Arguments
variable

species a numeric variable whose values weight the values of the analysis variables.
The values of the variable do not have to be integers. PROC TABULATE responds to
weight values in accordance with the following table.

The TABULATE Procedure

Statistics That Are Available in PROC TABULATE

Weight value

PROC TABULATE response

counts the observation in the total number of observations

less than 0

converts the value to zero and counts the observation in the

total number of observations

missing

1213

excludes the observation

statement.
Interaction: If you use the WEIGHT= option in a VAR statement to specify a

weight variable, then PROC TABULATE uses this variable instead to weight those
VAR statement variables.
When you use the WEIGHT statement, consider which value of the VARDEF=
option is appropriate. See the discussion of VARDEF= on page 1195 and the
calculation of weighted statistics in Keywords and Formulas on page 1340 for
more information.

Tip:

Note: Prior to Version 7 of SAS, the procedure did not exclude the observations
with missing weights from the count of observations. 4

Concepts: TABULATE Procedure

Statistics That Are Available in PROC TABULATE

Use the following keywords to request statistics in the TABLE statement or to
specify statistic keywords in the KEYWORD or KEYLABEL statement. If a variable
name (class or analysis) and a statistic name are the same, then enclose the statistic
name in single quotation marks for example, MAX.
Descriptive statistic keywords
COLPCTN

PCTSUM

COLPCTSUM

RANGE

CSS

REPPCTN

REPPCTSUM

KURTOSIS | KURT

ROWPCTN

LCLM

ROWPCTSUM

MAX

SKEWNESS | SKEW

MEAN

STDDEV|STD

MIN

STDERR

1214

Formatting Class Variables

Chapter 47

SUM

NMISS

SUMWGT

PAGEPCTN

UCLM

PAGEPCTSUM

USS

PCTN

VAR

Quantile statistic keywords

MEDIAN|P50

Q3|P75

P90

P95

P10

P99

Q1|P25

QRANGE

Hypothesis testing keywords

PROBT

These statistics, the formulas that are used to calculate them, and their data
requirements are discussed in Keywords and Formulas on page 1340.
To compute standard error of the mean (STDERR) or Students t-test, you must use
the default value of the VARDEF= option, which is DF. The VARDEF= option is
specied in the PROC TABULATE statement.
To compute weighted quantiles, you must use QMETHOD=OS in the PROC
TABULATE statement.
Use both LCLM and UCLM to compute a two-sided condence limit for the mean.
Use only LCLM or UCLM to compute a one-sided condence limit. Use the ALPHA=
option in the PROC TABULATE statement to specify a condence level.

Formatting Class Variables

Use the FORMAT statement to assign a format to a class variable for the duration of
a PROC TABULATE step. When you assign a format to a class variable, PROC
TABULATE uses the formatted values to create categories, and it uses the formatted
values in headings. If you do not specify a format for a class variable, and the variable
does not have any other format assigned to it, then the default format, BEST12., is
used, unless the GROUPINTERNAL option is specied.
User-dened formats are particularly useful for grouping values into fewer
categories. For example, if you have a class variable, Age, with values ranging from 1
to 99, then you could create a user-dened format that groups the ages so that your
tables contain a manageable number of categories. The following PROC FORMAT step
creates a format that condenses all possible values of age into six groups of values.
proc format;
value agefmt

run;

0-29=Under 30
30-39=30-39
40-49=40-49
50-59=50-59
60-69=60-69
other=70 or over;

The TABULATE Procedure

How Using BY-Group Processing Differs from Using the Page Dimension

1215

For information on creating user-dened formats, see Chapter 22, The FORMAT
Procedure, on page 429.
By default, PROC TABULATE includes in a table only those formats for which the
frequency count is not zero and for which values are not missing. To include missing
values for all class variables in the output, use the MISSING option in the PROC
TABULATE statement, and to include missing values for selected class variables, use
the MISSING option in a CLASS statement. To include formats for which the frequency
count is zero, use the PRELOADFMT option in a CLASS statement and the
PRINTMISS option in the TABLE statement, or use the CLASSDATA= option in the
PROC TABULATE statement.

Formatting Values in Tables

The formats for data in table cells serve two purposes. They determine how PROC
TABULATE displays the values, and they determine the width of the columns. The
default format for values in table cells is 12.2. You can modify the format for printing
values in table cells by

3 changing the default format with the FORMAT= option in the PROC TABULATE
statement

3 crossing elements in the TABLE statement with the F= format modier.

PROC TABULATE determines the format to use for a particular cell from the
following default order of precedence for formats:
1 If no other formats are specied, then PROC TABULATE uses the default format

(12.2).
2 The FORMAT= option in the PROC TABULATE statement changes the default

format. If no format modiers affect a cell, then PROC TABULATE uses this
format for the value in that cell.
3 A format modier in the page dimension applies to the values in all the table cells

on the logical page unless you specify another format modier for a cell in the row
or column dimension.
4 A format modier in the row dimension applies to the values in all the table cells

in the row unless you specify another format modier for a cell in the column
dimension.
5 A format modier in the column dimension applies to the values in all the table

cells in the column.

You can change this order of precedence by using the FORMAT_PRECEDENCE=
option in the TABLE statement. For example, if you specify
FORMAT_PRECEDENCE=ROW and specify a format modier in the row dimension,
then that format overrides all other specied formats for the table cells.

How Using BY-Group Processing Differs from Using the Page

Dimension
Using the page-dimension expression in a TABLE statement can have an effect
similar to using a BY statement.
Table 47.4 on page 1216 contrasts the two methods.

1216

Calculating Percentages

Table 47.4

Chapter 47

Contrasting the BY Statement and the Page Dimension

PROC TABULATE with a page dimension
in the TABLE statement

Issue

PROC TABULATE with a BY statement

Order of observations
in the input data set

The observations in the input data set must

be sorted by the BY variables. 1

Sorting is unnecessary.

One report
summarizing all BY
groups

You cannot create one report for all the BY

groups.

Use ALL in the page dimension to create a

report for all classes. (See Example 6 on
page 1246.)

Percentages

The percentages in the tables are

percentages of the total for that BY group.
You cannot calculate percentages for a BY
group compared to the totals for all BY
groups because PROC TABULATE prepares
the individual reports separately. Data for
the report for one BY group are not
available to the report for another BY
group.

You can use denominator denitions to

control the meaning of PCTN (see
Calculating Percentages on page 1216.)

Titles

You can use the #BYVAL, #BYVAR, and

#BYLINE specications in TITLE
statements to customize the titles for each
BY group (see Creating Titles That Contain
BY-Group Information on page 20).

The BOX= option in the TABLE statement

customizes the page headers, but you must
use the same title on each page.

Ordering class
variables

ORDER=DATA and ORDER=FREQ order

each BY group independently.

The order of class variables is the same on

every page.

Obtaining uniform
headings

You may need to insert dummy

observations into BY groups that do not
have all classes represented.

The PRINTMISS option ensures that each

page of the table has uniform headings.

Multiple ranges with

the same format

PROC TABULATE produces a table for

each range.

PROC TABULATE combines observations

from the two ranges.

1 You can use the BY statement without sorting the data set if the data set has an index for the BY variable.

Calculating Percentages
Calculating the Percentage of the Value of in a Single Table Cell
The following statistics print the percentage of the value in a single table cell in
relation to the total of the values in a group of cells. No denominator denitions are
required; however, an analysis variable may be used as a denominator denition for
percentage sum statistics.
REPPCTN and REPPCTSUM statisticsprint the percentage of the value in a single
table cell in relation to the total of the values in the report.
COLPCTN and COLPCTSUM statisticsprint the percentage of the value in a single
table cell in relation to the total of the values in the column.
ROWPCTN and ROWPCTSUM statisticsprint the percentage of the value in a
single table cell in relation to the total of the values in the row.
PAGEPCTN and PAGEPCTSUM statisticsprint the percentage of the value in a
single table cell in relation to the total of the values in the page.

The TABULATE Procedure

Calculating Percentages

1217

These statistics calculate the most commonly used percentages. See Example 12 on
page 1266 for an example.

Using PCTN and PCTSUM

PCTN and PCTSUM statistics can be used to calculate these same percentages. They
allow you to manually dene denominators. PCTN and PCTSUM statistics print the
percentage of the value in a single table cell in relation to the value (used in the
denominator of the calculation of the percentage) in another table cell or to the total of
the values in a group of cells. By default, PROC TABULATE summarizes the values in
all N cells (for PCTN) or all SUM cells (for PCTSUM) and uses the summarized value
for the denominator. You can control the value that PROC TABULATE uses for the
denominator with a denominator denition.
You place a denominator denition in angle brackets (< and >) next to the PCTN or
PCTSUM statistic. The denominator denition species which categories to sum for the
denominator.
This section illustrates how to specify denominator denitions in a simple table.
Example 13 on page 1269 illustrates how to specify denominator denitions in a table
that is composed of multiple subtables. For more examples of denominator denitions,
see How Percentages Are Calculated in Chapter 3, Details of TABULATE
Processing, in SAS Guide to TABULATE Processing.

Specifying a Denominator for the PCTN Statistic

The following PROC TABULATE step calculates the N statistic and three different
versions of PCTN using the data set ENERGYENERGY on page 1387.
proc tabulate data=energy;
class division type;
table division*
(n=Number of customers
pctn<type>=% of row u
pctn<division>=% of column v
pctn=% of all customers), w
type/rts=50;
title Number of Users in Each Division;
run;

The TABLE statement creates a row for each value of Division and a column for
each value of Type. Within each row, the TABLE statement nests four statistics: N and
three different calculations of PCTN (see Figure 47.4 on page 1218). Each occurrence of
PCTN uses a different denominator denition.

1218

Calculating Percentages

Figure 47.4
Highlighted

Chapter 47

Three Different Uses of the PCTN Statistic with Frequency Counts

Number of Users in Each Division

1
-----------------------------------------------------------|
|
Type
|
|
|-------------------------|
|
|
1
|
2
|
|--------------------------------+------------+------------|
|Division |
|
|
|
|
|---------+----------------------|
|
|1
|Number of customers
|
6.00|
6.00|
|
|----------------------+------------+------------|
|
|% of row
|
50.00|
50.00|
|
|----------------------+------------+------------|
|
27.27|
27.27|
|
|% of column
|
|----------------------+------------+------------|
13.64|
13.64|
|
|% of all customers |
|---------+----------------------+------------+------------|
|2
|Number of customers
|
3.00|
3.00|
|
|----------------------+------------+------------|
|
|% of row
|
50.00|
50.00|
|
|----------------------+------------+------------|
|
|% of column
|
13.64|
13.64|
|
|----------------------+------------+------------|
|
|% of all customers
|
6.82|
6.82|
|---------+----------------------+------------+------------|
|3
|Number of customers
|
8.00|
8.00|
|
|----------------------+------------+------------|
|
|% of row
|
50.00|
50.00|
|
|----------------------+------------+------------|
|
|% of column
|
36.36|
36.36|
|
|----------------------+------------+------------|
|
|% of all customers
|
18.18|
18.18|
|---------+----------------------+------------+------------|
|4
|Number of customers
|
5.00|
5.00|
|
|----------------------+------------+------------|
|
|% of row
|
50.00|
50.00|
|
|----------------------+------------+------------|
|
|% of column
|
22.73|
22.73|
|
|----------------------+------------+------------|
|
|% of all customers
|
11.36|
11.36|

u <type> sums the frequency counts for all occurrences of Type within the same
value of Division. Thus, for Division=1, the denominator is 6 + 6, or 12.
v <division> sums the frequency counts for all occurrences of Division within the
same value of Type. Thus, for Type=1, the denominator is 6 + 3 + 8 + 5, or 22.
w The third use of PCTN has no denominator denition. Omitting a denominator
denition is the same as including all class variables in the denominator denition.
Thus, for all cells, the denominator is 6 + 3 + 8 + 5 + 6 + 3 + 8 + 5, or 44.

Specifying a Denominator for the PCTSUM Statistic

The following PROC TABULATE step sums expenditures for each combination of
Type and Division and calculates three different versions of PCTSUM.
proc tabulate data=energy format=8.2;
class division type;
var expenditures;
table division*
(sum=Expenditures*f=dollar10.2
pctsum<type>=% of row u

The TABULATE Procedure

Calculating Percentages

1219

pctsum<division>=% of column v
pctsum=% of all customers), w
type*expenditures/rts=40;
title Expenditures in Each Division;
run;

The TABLE statement creates a row for each value of Division and a column for each
value of Type. Because Type is crossed with Expenditures, the value in each cell is the
sum of the values of Expenditures for all observations that contribute to the cell.
Within each row, the TABLE statement nests four statistics: SUM and three different
calculations of PCTSUM (see Figure 47.5 on page 1219). Each occurrence of PCTSUM
uses a different denominator denition.
Figure 47.5

Three Different Uses of the PCTSUM Statistic with Sums Highlighted

Expenditures in Each Division

-------------------------------------------------------|
|
Type
|
|
|---------------------|
|
|
1
|
2
|
|
|----------+----------|
|
| Expend | Expend |
|--------------------------------+----------+----------|
|Division
|
|
|
|
|-----------+--------------------|
|
|
|1
|Expenditures
| $7,477.00| $5,129.00|
|
|--------------------+----------+----------|
|
|% of row
|
59.31|
40.69|
|
|--------------------+----------+----------|
|
|% of column
|
16.15|
13.66|
|
|--------------------+----------+----------|
|
|% of all customers |
8.92|
6.12|
|-----------+--------------------+----------+----------|
|2
|Expenditures
|$19,379.00|$15,078.00|
|
|--------------------+----------+----------|
|
|% of row
|
56.24|
43.76|
|
|--------------------+----------+----------|
|
|% of column
|
41.86|
40.15|
|
|--------------------+----------+----------|
|
|% of all customers |
23.11|
17.98|
|-----------+--------------------+----------+----------|
|3
|Expenditures
| $5,476.00| $4,729.00|
|
|--------------------+----------+----------|
|
|% of row
|
53.66|
46.34|
|
|--------------------+----------+----------|
|
|% of column
|
11.83|
12.59|
|
|--------------------+----------+----------|
|
|% of all customers |
6.53|
5.64|
|-----------+--------------------+----------+----------|
|4
|Expenditures
|$13,959.00|$12,619.00|
|
|--------------------+----------+----------|
|
|% of row
|
52.52|
47.48|
|
|--------------------+----------+----------|
|
|% of column
|
30.15|
33.60|
|
|--------------------+----------+----------|
|
|% of all customers |
16.65|
15.05|
--------------------------------------------------------

u <type> sums the values of Expenditures for all occurrences of Type within the
same value of Division. Thus, for Division=1, the denominator is $7,477 + $5,129.
v <division> sums the frequency counts for all occurrences of Division within the
same value of Type. Thus, for Type=1, the denominator is $7,477 + $19,379 +
$5,476 + $13,959.
w The third use of PCTN has no denominator denition. Omitting a denominator
denition is the same as including all class variables in the denominator

1220

Using Style Elements in PROC TABULATE

Chapter 47

denition. Thus, for all cells, the denominator is $7,477 + $19,379 + $5,476 +
$13,959 + $5,129 + $15,078 + $4,729 + $12,619.

Using Style Elements in PROC TABULATE

What Are Style Elements?

If you use the Output Delivery System to create HTML, RTF, or Printer output from
PROC TABULATE, then you can set the style element that the procedure uses for
various parts of the table. Style elements determine presentation attributes, such as
font face, font weight, color, and so forth. See Output Delivery System on page 32 for
more information. The following table lists the default styles for various regions of a
table.
Table 47.5

Default Styles for Table Regions

Region

Style

column headings

Header

box

Header

page dimension text

Beforecaption

row headings

Rowheader

data cells

Data

table

Table

Using the STYLE= Option

You specify style elements for PROC TABULATE with the STYLE= option. The
following table shows where you can use this option. Specications in the TABLE
statement override the same specication in the PROC TABULATE statement.
However, any style attributes that you specify in the PROC TABULATE statement and
that you do not override in the TABLE statement are inherited. For instance, if you
specify a blue background and a white foreground for all data cells in the PROC
TABULATE statement, and you specify a gray background for the data cells of a
particular crossing in the TABLE statement, then the background for those data cells is
gray, and the foreground is white (as specied in the PROC TABULATE statement).
Detailed information on STYLE= is provided in the documentation for individual
statements.

Table 47.6

Using the STYLE= Option in PROC TABULATE

To set the style element for

Use STYLE in this statement

data cells

PROC TABULATE or dimension expression(s)

page dimension text and class variable name headings

CLASS

class level value headings

CLASSLEV

keyword headings

KEYWORD

The TABULATE Procedure

Using Style Elements in PROC TABULATE

To set the style element for

Use STYLE in this statement

table borders, rules, and other parts that are not

specied elsewhere

TABLE

box text

TABLE statement, BOX= option

missing values

TABLE statement, MISSTEXT= option

analysis variable name headings

1221

VAR

Applying Style Attributes to Table Cells

PROC TABULATE determines the style attributes to use for a particular cell from
the following default order of precedence for styles:
1 If no other style attributes are specied, then PROC TABULATE uses the default

style attributes from the default style (Data).

2 The STYLE= option in the PROC TABULATE statement changes the default style

attributes. If no other STYLE= option specications affect a cell, then PROC

TABULATE uses these style attributes for that cell.
3 A STYLE= option that is specied in the page dimension applies to all the table

cells on the logical page unless you specify another STYLE= option for a cell in the
row or column dimension.
4 A STYLE= option that is specied in the row dimension applies to all the table

cells in the row unless you specify another STYLE= option for a cell in the column
dimension.
5 A STYLE= option that is specied in the column dimension applies to all the table

cells in the column.

You can change this order of precedence by using the STYLE_PRECEDENCE=
option in the TABLE statement. For example, if you specify
STYLE_PRECEDENCE=ROW and specify a STYLE= option in the row dimension, then
those style attribute values override all others that are specied for the table cells.

Using a Format to Assign a Style Attribute

You can use a format to assign a style attribute value to any cell whose content is
determined by value(s) of a class or analysis variable. For example, the following code
assigns a red background to cells whose values are less than 10,000, a yellow
background to cells whose values are at least 10,000 but less than 20,000, and a green
background to cells whose values are at least 20,000:
proc format;
value expfmt low-<10000=red
10000-<20000=yellow
20000-high=green;
run;
ods html body=external-HTML-file;
proc tabulate data=energy style=[background=expfmt.];
class region division type;
var expenditures;
table (region all)*(division all),
type*expenditures;
run;
ods html close;

1222

Results: TABULATE Procedure

Chapter 47

Results: TABULATE Procedure

Missing Values
How PROC TABULATE Treats Missing Values
How a missing value for a variable in the input data set affects your output depends
on how you use the variable in the PROC TABULATE step. Table 47.7 on page 1222
summarizes how the procedure treats missing values.
Table 47.7

Summary of How PROC TABULATE Treats Missing Values

PROC TABULATE, by default,

To override the default

an observation contains a missing

value for an analysis variable

excludes that observation from the

calculation of statistics (except N and
NMISS) for that particular variable

no alternative

an observation contains a missing

value for a class variable

excludes that observation from the

table1

use MISSING in the PROC

TABULATE statement, or MISSING
in the CLASS statement

there are no data for a category

does not show the category in the

table

use PRINTMISS in the TABLE

statement, or use CLASSDATA= in
the PROC TABULATE statement

every observation that contributes to

a table cell contains a missing value
for an analysis variable

displays a missing value for any

statistics (except N and NMISS) in
that cell

use MISSTEXT= in the TABLE

statement

there are no data for a formatted

value

does not display that formatted

value in the table

use PRELOADFMT in the CLASS

statement with PRINTMISS in the
TABLE statement, or use
CLASSDATA= in the PROC
TABULATE statement, or add
dummy observations to the input
data set so that it contains data for
each formatted value

a FREQ variable value is missing or

is less than 1

does not use that observation to

calculate statistics

no alternative

a WEIGHT variable value is missing

or 0

uses a value of 0

no alternative

1 The CLASS statement applies to all TABLE statements in a PROC TABULATE step. Therefore, if you dene a variable as
a class variable, PROC TABULATE omits observations that have missing values for that variable even if you do not use the
variable in a TABLE statement.

This section presents a series of PROC TABULATE steps that illustrate how PROC
TABULATE treats missing values. The following program creates the data set and
formats that are used in this section and prints the data set. The data set COMPREV
contains no missing values (see Figure 47.6 on page 1223).

The TABULATE Procedure

Missing Values

proc format;
value cntryfmt 1=United States
2=Japan;
value compfmt 1=Supercomputer
2=Mainframe
3=Midrange
4=Workstation
5=Personal Computer
6=Laptop;
run;

data comprev;
input Country Computer Rev90 Rev91 Rev92;
datalines;
1 1 788.8 877.6 944.9
1 2 12538.1 9855.6 8527.9
1 3 9815.8 6340.3 8680.3
1 4 3147.2 3474.1 3722.4
1 5 18660.9 18428.0 23531.1
2 1 469.9 495.6 448.4
2 2 5697.6 6242.4 5382.3
2 3 5392.1 5668.3 4845.9
2 4 1511.6 1875.5 1924.5
2 5 4746.0 4600.8 4363.7
;

proc print data=comprev noobs;

format country cntryfmt. computer compfmt.;
title The Data Set COMPREV;
run;

Figure 47.6

The Data Set COMPREV

Country

Computer

United
United
United
United
United
Japan
Japan
Japan
Japan
Japan

Supercomputer
Mainframe
Midrange
Workstation
Personal Computer
Supercomputer
Mainframe
Midrange
Workstation
Personal Computer

States
States
States
States
States

Rev90

Rev91

Rev92

788.8
12538.1
9815.8
3147.2
18660.9
469.9
5697.6
5392.1
1511.6
4746.0

877.6
9855.6
6340.3
3474.1
18428.0
495.6
6242.4
5668.3
1875.5
4600.8

944.9
8527.9
8680.3
3722.4
23531.1
448.4
5382.3
4845.9
1924.5
4363.7

No Missing Values
The following PROC TABULATE step produces Figure 47.7 on page 1224:
proc tabulate data=comprev;
class country computer;
var rev90 rev91 rev92;

1223

1224

Missing Values

Chapter 47

table computer*country,rev90 rev91 rev92 /

rts=32;
format country cntryfmt. computer compfmt.;
title Revenues from Computer Sales;
title2 for 1990 to 1992;
run;

Figure 47.7

Computer Sales Data: No Missing Values

Because the data set contains no missing values, the table includes all observations. All headers
and cells contain nonmissing values.
Revenues from Computer Sales
for 1990 to 1992

-------------------------------------------------------------|
| Rev90
| Rev91
| Rev92 |
|
|----------+----------+--------|
|
| Sum
| Sum
| Sum
|
|-----------------------------+----------+----------+--------|
|Computer
|Country
|
|
|
|
|--------------+--------------|
|
|
|
|Supercomputer |United States |
788.80|
877.60| 944.90|
|
|--------------+----------+----------+--------|
|
|Japan
|
469.90|
495.60| 448.40|
|--------------+--------------+----------+----------+--------|
|Mainframe
|United States | 12538.10|
9855.60| 8527.90|
|
|--------------+----------+----------+--------|
|
|Japan
|
5697.60|
6242.40| 5382.30|
|--------------+--------------+----------+----------+--------|
|Midrange
|United States |
9815.80|
6340.30| 8680.30|
|
|--------------+----------+----------+--------|
|
|Japan
|
5392.10|
5668.30| 4845.90|
|--------------+--------------+----------+----------+--------|
|Workstation
|United States |
3147.20|
3474.10| 3722.40|
|
|--------------+----------+----------+--------|
|
|Japan
|
1511.60|
1875.50| 1924.50|
|--------------+--------------+----------+----------+--------|
|Personal
|United States | 18660.90| 18428.00|23531.10|
|Computer
|--------------+----------+----------+--------|
|
|Japan
|
4746.00|
4600.80| 4363.70|
--------------------------------------------------------------

A Missing Class Variable

The next program copies COMPREV and alters the data so that the eighth
observation has a missing value for Computer. Except for specifying this new data set,
the program that produces Figure 47.8 on page 1225 is the same as the program that
produces Figure 47.7 on page 1224. By default, PROC TABULATE ignores observations
with missing values for a class variable.
data compmiss;
set comprev;
if _n_=8 then computer=.;
run;

proc tabulate data=compmiss;

class country computer;

The TABULATE Procedure

Missing Values

1225

var rev90 rev91 rev92;

table computer*country,rev90 rev91 rev92 /
rts=32;
format country cntryfmt. computer compfmt.;
title Revenues from Computer Sales;
title2 for 1990 to 1992;
run;

Figure 47.8

Computer Sales Data: Midrange, Japan, Deleted

The observation with a missing value for Computer was the category Midrange, Japan. This
category no longer exists. By default, PROC TABULATE ignores observations with missing
values for a class variable, so this table contains one fewer row than Figure 47.7 on page 1224.

Revenues from Computer Sales

for 1990 to 1992

-------------------------------------------------------------|
| Rev90
| Rev91
| Rev92
|
|
|---------+----------+---------|
|
|
| Sum
| Sum
| Sum
|-----------------------------+---------+----------+---------|
|Computer
|Country
|
|
|
|
|--------------+--------------|
|
|
|
|Supercomputer |United States |
788.80|
877.60|
944.90|
|
|--------------+---------+----------+---------|
|
|Japan
|
469.90|
495.60|
448.40|
|--------------+--------------+---------+----------+---------|
|Mainframe
|United States | 12538.10|
9855.60| 8527.90|
|
|--------------+---------+----------+---------|
|
|Japan
| 5697.60|
6242.40| 5382.30|
|--------------+--------------+---------+----------+---------|
|Midrange
|United States | 9815.80|
6340.30| 8680.30|
|--------------+--------------+---------+----------+---------|
|Workstation
|United States | 3147.20|
3474.10| 3722.40|
|
|--------------+---------+----------+---------|
|
|Japan
| 1511.60|
1875.50| 1924.50|
|--------------+--------------+---------+----------+---------|
|Personal
|United States | 18660.90| 18428.00| 23531.10|
|Computer
|--------------+---------+----------+---------|
|
|Japan
| 4746.00|
4600.80| 4363.70|
--------------------------------------------------------------

Including Observations with Missing Class Variables

This program adds the MISSING option to the previous program. MISSING is
available either in the PROC TABULATE statement or in the CLASS statement. If you
want MISSING to apply only to selected class variables, but not to others, then specify
MISSING in a separate CLASS statement with the selected variable(s). The MISSING
option includes observations with missing values of a class variable in the report (see
Figure 47.9 on page 1226).
proc tabulate data=compmiss missing;
class country computer;
var rev90 rev91 rev92;
table computer*country,rev90 rev91 rev92 /
rts=32;
format country cntryfmt. computer compfmt.;

1226

Missing Values

Chapter 47

title Revenues from Computer Sales;

title2 for 1990 to 1992;
run;

Figure 47.9

Computer Sales Data: Missing Values for Computer

This table includes a category with missing values of Computer. This category makes up the
rst row of data in the table.
1
------------------------------------------------------------|
|
Animal
|-----------------------------------------------------------|
|
|
cat
|
dog
|-----------------------------+-----------------------------|
|
|
Food
|
Food
|-----------------------------+-----------------------------|
| fish
| meat
| milk
| fish
| meat
| bones |
|---------+---------+---------+---------+---------+---------|
|
N
|
N
|
N
|
N
|
N
|
N
|
|---------+---------+---------+---------+---------+---------|
|
1|
1|
1|
1|
1|
1|
-------------------------------------------------------------

Formatting Headings for Observations with Missing Class Variables

By default, as shown in Figure 47.9 on page 1226, PROC TABULATE displays
missing values of a class variable as one of the standard SAS characters for missing
values (a period, a blank, an underscore, or one of the letters A through Z). If you want
to display something else instead, then you must assign a format to the class variable
that has missing values, as shown in the following program (see Figure 47.10 on page
1227):
proc format;
value misscomp 1=Supercomputer
2=Mainframe
3=Midrange
4=Workstation
5=Personal Computer
6=Laptop
.=No type given;
run;

proc tabulate data=compmiss missing;

class country computer;
var rev90 rev91 rev92;
table computer*country,rev90 rev91 rev92 /
rts=32;
format country cntryfmt. computer misscomp.;

The TABULATE Procedure

Missing Values

1227

title Revenues for Computer Sales;

title2 for 1990 to 1992;
run;

Figure 47.10

Computer Sales Data: Text Supplied for Missing Computer Value

In this table, the missing value appears as the text that the MISSCOMP. format species.

Revenues for Computer Sales

for 1990 to 1992

---------------------------------------------------------|
| Rev90 | Rev91 | Rev92 |
|
|--------+--------+--------|
|
| Sum
| Sum
| Sum
|
|-----------------------------+--------+--------+--------|
|Computer
|Country
|
|
|
|
|--------------+--------------|
|
|
|
|No type given |Japan
| 5392.10| 5668.30| 4845.90|
|--------------+--------------+--------+--------+--------|
|Supercomputer |United States | 788.80| 877.60| 944.90|
|
|--------------+--------+--------+--------|
|
|Japan
| 469.90| 495.60| 448.40|
|--------------+--------------+--------+--------+--------|
|Mainframe
|United States |12538.10| 9855.60| 8527.90|
|
|--------------+--------+--------+--------|
|
|Japan
| 5697.60| 6242.40| 5382.30|
|--------------+--------------+--------+--------+--------|
|Midrange
|United States | 9815.80| 6340.30| 8680.30|
|--------------+--------------+--------+--------+--------|
|Workstation
|United States | 3147.20| 3474.10| 3722.40|
|
|--------------+--------+--------+--------|
|
|Japan
| 1511.60| 1875.50| 1924.50|
|--------------+--------------+--------+--------+--------|
|Personal
|United States |18660.90|18428.00|23531.10|
|Computer
|--------------+--------+--------+--------|
|
|Japan
| 4746.00| 4600.80| 4363.70|
----------------------------------------------------------

Providing Headings for All Categories

By default, PROC TABULATE evaluates each page that it prints and omits columns
and rows for categories that do not exist. For example, Figure 47.10 on page 1227 does
not include a row for No type given and for United States or for Midrange and for
Japan because there are no data in these categories. If you want the table to represent
all possible categories, then use the PRINTMISS option in the TABLE statement, as
shown in the following program (see Figure 47.11 on page 1228):
proc tabulate data=compmiss missing;
class country computer;
var rev90 rev91 rev92;
table computer*country,rev90 rev91 rev92 /
rts=32 printmiss;
format country cntryfmt. computer misscomp.;
title Revenues for Computer Sales;
title2 for 1990 to 1992;
run;

1228

Missing Values

Chapter 47

Figure 47.11

Computer Sales Data: Missing Statistics Values

This table contains a row for the categories No type given, United States and Midrange,
Japan. Because there are no data in these categories, the values for the statistics are all
missing.

Revenues for Computer Sales

for 1990 to 1992

-----------------------------------------------------------|
| Rev90
| Rev91 | Rev92 |
|
|---------+---------+---------|
|
| Sum
|
Sum
|
Sum
|
|----------------------------+---------+---------+---------|
|Computer
|Country
|
|
|
|
|--------------+-------------|
|
|
|
|No type given |United States|
.|
.|
.|
|
|-------------+---------+---------+---------|
|
|Japan
| 5392.10| 5668.30| 4845.90|
|--------------+-------------+---------+---------+---------|
|Supercomputer |United States|
788.80|
877.60|
944.90|
|
|-------------+---------+---------+---------|
|
|Japan
|
469.90|
495.60|
448.40|
|--------------+-------------+---------+---------+---------|
|Mainframe
|United States| 12538.10| 9855.60| 8527.90|
|
|-------------+---------+---------+---------|
|
|Japan
| 5697.60| 6242.40| 5382.30|
|--------------+-------------+---------+---------+---------|
|Midrange
|United States| 9815.80| 6340.30| 8680.30|
|
|-------------+---------+---------+---------|
|
|Japan
|
.|
.|
.|
|--------------+-------------+---------+---------+---------|
|Workstation
|United States| 3147.20| 3474.10| 3722.40|
|
|-------------+---------+---------+---------|
|
|Japan
| 1511.60| 1875.50| 1924.50|
|--------------+-------------+---------+---------+---------|
|Personal
|United States| 18660.90| 18428.00| 23531.10|
|Computer
|-------------+---------+---------+---------|
|
|Japan
| 4746.00| 4600.80| 4363.70|
------------------------------------------------------------

Providing Text for Cells That Contain Missing Values

If some observations in a category contain missing values for analysis variables, then
PROC TABULATE does not use those observations to calculate statistics (except N and
NMISS). However, if each observation in a category contains a missing value, then
PROC TABULATE displays a missing value for the value of the statistic. To replace
missing values for analysis variables with text, use the MISSTEXT= option in the
TABLE statement to specify the text to use, as shown in the following program (see
Figure 47.12 on page 1229).
proc tabulate data=compmiss missing;
class country computer;
var rev90 rev91 rev92;
table computer*country,rev90 rev91 rev92 /
rts=32 printmiss misstext=NO DATA!;
format country cntryfmt. computer misscomp.;
title Revenues for Computer Sales;
title2 for 1990 to 1992;
run;

The TABULATE Procedure

Figure 47.12

Missing Values

1229

Computer Sales Data: Text Supplied for Missing Statistics Values

This table replaces the period normally used to display missing values with the text of the
MISSTEXT= option.
Revenues for Computer Sales
for 1990 to 1992

---------------------------------------------------------|
| Rev90 | Rev91 | Rev92 |
|
|--------+--------+--------|
|
|
Sum | Sum
| Sum
|
|-----------------------------+--------+------------+----|
|Computer
|Country
|
|
|
|
|
|--------------+--------------|
|
|
|No type given |United States |NO DATA!|NO DATA!|NO DATA!|
|
|--------------+--------+--------+--------|
|
|Japan
| 5392.10| 5668.30| 4845.90|
|--------------+--------------+--------+--------+--------|
|Supercomputer |United States | 788.80| 877.60| 944.90|
|
|--------------+--------+--------+--------|
|
|Japan
| 469.90| 495.60| 448.40|
|--------------+--------------+--------+--------+--------|
|Mainframe
|United States |12538.10| 9855.60| 8527.90|
|
|--------------+--------+--------+--------|
|
|Japan
| 5697.60| 6242.40| 5382.30|
|--------------+--------------+--------+--------+--------|
|Midrange
|United States | 9815.80| 6340.30| 8680.30|
|
|--------------+--------+--------+--------|
|
|Japan
|NO DATA!|NO DATA!|NO DATA!|
|--------------+--------------+--------+--------+--------|
|Workstation
|United States | 3147.20| 3474.10| 3722.40|
|
|--------------+--------+--------+--------|
|
|Japan
| 1511.60| 1875.50| 1924.50|
|--------------+--------------+--------+--------+--------|
|Personal
|United States |18660.90|18428.00|23531.10|
|Computer
|--------------+--------+--------+--------|
|
|Japan
| 4746.00| 4600.80| 4363.70|
----------------------------------------------------------

Providing Headings for All Values of a Format

PROC TABULATE prints headings only for values that appear in the input data set.
For example, the format COMPFMT. provides for six possible values of Computer. Only
ve of these values occur in the data set COMPREV. The data set contains no data for
laptop computers.
If you want to include headings for all possible values of Computer (perhaps to make
it easier to compare the output with tables that are created later when you do have
data for laptops), then you have three different ways to create such a table:
3 Use the PRELOADFMT option in the CLASS statement with the PRINTMISS
option in the TABLE statement. See Example 3 on page 1237 for another example
that uses PRELOADFMT.
3 Use the CLASSDATA= option in the PROC TABULATE statement. See Example 2
on page 1235 for an example that uses the CLASSDATA= option.
3 Add dummy values to the input data set so that each value that the format
handles appears at least once in the data set.
The following program adds the PRELOADFMT option to a CLASS statement that
contains the relevant variable.
The results are shown in Figure 47.13 on page 1230.
proc tabulate data=compmiss missing;
class country;

1230

Understanding the Order of Headings with ORDER=DATA

Chapter 47

class computer / preloadfmt;

var rev90 rev91 rev92;
table computer*country,rev90 rev91 rev92 /
rts=32 printmiss misstext=NO DATA!;
format country cntryfmt. computer compfmt.;
title Revenues for Computer Sales;
title2 for 1990 to 1992;
run;

Figure 47.13

Computer Sales Data: All Possible Computer Values Included

This table contains a heading for each possible value of Computer.

Revenues for Computer Sales
for 1990 to 1992

--------------------------------------------------------|
| Rev90 | Rev91 | Rev92 |
|
|--------+--------+--------|
|
| Sum
|
Sum |
Sum |
|----------------------------+--------+--------+--------|
|Computer
|Country
|
|
|
|
|--------------+-------------|
|
|
|
|.
|United States|NO DATA!|NO DATA!|NO DATA!|
|
|-------------+--------+--------+--------|
|
|Japan
| 5392.10| 5668.30| 4845.90|
|--------------+-------------+--------+--------+--------|
|Supercomputer |United States| 788.80| 877.60| 944.90|
|
|-------------+--------+--------+--------|
|
|Japan
| 469.90| 495.60| 448.40|
|--------------+-------------+--------+--------+--------|
|Mainframe
|United States|12538.10| 9855.60| 8527.90|
|
|-------------+--------+--------+--------|
|
|Japan
| 5697.60| 6242.40| 5382.30|
|--------------+-------------+--------+--------+--------|
|Midrange
|United States| 9815.80| 6340.30| 8680.30|
|
|-------------+--------+------------+----|
|
|Japan
|NO DATA!|NO DATA!|NO DATA!|
|--------------+-------------+--------+------------+----|
|Workstation
|United States| 3147.20| 3474.10| 3722.40|
|
|-------------+--------+------------+----|
|
|Japan
| 1511.60| 1875.50| 1924.50|
|--------------+-------------+--------+--------+--------|
|Personal
|United States|18660.90|18428.00|23531.10|
|Computer
|-------------+--------+--------+--------|
|
|Japan
| 4746.00| 4600.80| 4363.70|
|--------------+-------------+--------+--------+--------|
|Laptop
|United States|NO DATA!|NO DATA!|NO DATA!|
|
|-------------+--------+--------+--------|
|
|Japan
|NO DATA!|NO DATA!|NO DATA!|
---------------------------------------------------------

Understanding the Order of Headings with ORDER=DATA

The ORDER= option applies to all class variables. Occasionally, you want to order
the headings for different variables differently. One method for doing this is to group
the data as you want them to appear and to specify ORDER=DATA.
For this technique to work, the rst value of the rst class variable must occur in the
data with all possible values of all the other class variables. If this criterion is not met,
then the order of the headings might surprise you.

The TABULATE Procedure

Portability of ODS Output with PROC TABULATE

1231

The following program creates a simple data set in which the observations are
ordered rst by the values of Animal, then by the values of Food. The ORDER= option
in the PROC TABULATE statement orders the heading for the class variables by the
order of their appearance in the data set (see Figure 47.14 on page 1231). Although
bones is the rst value for Food in the group of observations where Animal=dog, all
other values for Food appear before bones in the data set because bones never appears
when Animal=cat. Therefore, the header for bones in the table in Figure 47.14 on page
1231 is not in alphabetical order.
In other words, PROC TABULATE maintains for subsequent categories the order
that was established by earlier categories. If you want to re-establish the order of Food
for each value of Animal, then use BY-group processing. PROC TABULATE creates a
separate table for each BY group, so that the ordering can differ from one BY group to
the next.
data foodpref;
input Animal $ Food $;
datalines;
cat fish
cat meat
cat milk
dog bones
dog fish
dog meat
;
proc tabulate data=foodpref format=9.
order=data;
class animal food;
table animal*food;
run;

Figure 47.14

Ordering the Headings of Class Variables

------------------------------------------------------------|
|
Animal
|-----------------------------------------------------------|
|
|
cat
|
dog
|-----------------------------+-----------------------------|
|
|
Food
|
Food
|-----------------------------+-----------------------------|
| fish
| meat
| milk
| fish
| meat
| bones |
|---------+---------+---------+---------+---------+---------|
|
N
|
N
|
N
|
N
|
N
|
N
|
|---------+---------+---------+---------+---------+---------|
|
1|
1|
1|
1|
1|
1|
-------------------------------------------------------------

Portability of ODS Output with PROC TABULATE

Under certain circumstances, using PROC TABULATE with the Output Delivery
System produces les that are not portable. If the SAS system option FORMCHAR= in
your SAS session uses nonstandard line-drawing characters, then the output might
include strange characters instead of lines in operating environments in which the SAS

1232

Examples: TABULATE Procedure

Chapter 47

Monospace font is not installed. To avoid this problem, specify the following OPTIONS
statement before executing PROC TABULATE:
options formchar="|----|+|---+=|-/\<>*";

Examples: TABULATE Procedure

Example 1: Creating a Basic Two-Dimensional Table

Procedure features:

PROC TABULATE statement options:

FORMAT=
TABLE statement
crossing (*) operator
TABLE statement options:
RTS=
Other features: FORMAT statement

This example
3 creates a category for each type of user (residential or business) in each division of
each region
3 applies the same format to all cells in the table
3 applies a format to each class variable
3 extends the space for row headings.

Program

Create the ENERGY data set. ENERGY contains data on expenditures of energy for business
and residential customers in individual states in the Northeast and West regions of the United
States. A DATA step on page 1387 creates the data set.
data energy;
length State $2;
input Region Division state $ Type Expenditures;
datalines;
1 1 ME 1 708
1 1 ME 2 379
. . . more data lines . . .
4 4 HI 1 273

The TABULATE Procedure

Program

1233

4 4 HI 2 298
;

Create the REGFMT., DIVFMT., and USETYPE. formats. PROC FORMAT creates formats
for Region, Division, and Type.
proc format;
value regfmt 1=Northeast
2=South
3=Midwest
4=West;
value divfmt 1=New England
2=Middle Atlantic
3=Mountain
4=Pacific;
value usetype 1=Residential Customers
2=Business Customers;
run;

Specify the table options. The FORMAT= option species DOLLAR12. as the default format
for the value in each table cell.
proc tabulate data=energy format=dollar12.;

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of Region, Division, and Type.
class region division type;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

1234

Output

Chapter 47

Specify the row title space. RTS= provides 25 characters per line for row headings.
/ rts=25;

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

Output

Energy Expenditures for Each Region

The TABULATE Procedure

Program

1235

Example 2: Specifying Class Variable Combinations to Appear in a Table

Procedure features:

PROC TABULATE Statement options:

CLASSDATA=
EXCLUSIVE
Data set:

ENERGY ENERGY on page 1387

Formats: REGFMT., DIVFMT., and USETYPE. on page 1233

This example

3 uses the CLASSDATA= option to specify combinations of class variables to appear

in a table

3 uses the EXCLUSIVE option to restrict the output to only the combinations
specied in the CLASSDATA= data set. Without the EXCLUSIVE option, the
output would be the same as in Example 1 on page 1232.

Program

Create the CLASSES data set. CLASSES contains the combinations of class variable values
that PROC TABULATE uses to create the table.
data classes;
input region division type;
datalines;
1 1 1
1 1 2
4 4 1
4 4 2
;

Specify the table options. CLASSDATA= and EXCLUSIVE restrict the class level
combinations to those that are specied in the CLASSES data set.
proc tabulate data=energy format=dollar12.
classdata=classes exclusive;

1236

Program

Chapter 47

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of Region, Division, and Type.
class region division type;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

Specify the row title space. RTS= provides 25 characters per line for row headings.
/ rts=25;

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

The TABULATE Procedure

Example 3: Using Preloaded Formats with Class Variables

1237

Output
Energy Expenditures for Each Region
(millions of dollars)

--------------------------------------------------|
|
Type
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|
|------------+------------|
|
|Expenditures|Expenditures|
|
|------------+------------|
|
|
Sum
|
Sum
|
|-----------------------+------------+------------|
|Region
|Division
|
|
|
|-----------+-----------|
|
|
|Northeast |New England|
$7,477|
$5,129|
|-----------+-----------+------------+------------|
|West
|Pacific
|
$13,959|
$12,619|
---------------------------------------------------

Example 3: Using Preloaded Formats with Class Variables

Procedure features:

PROC TABULATE statement option:

OUT=
CLASS statement options:
EXCLUSIVE
PRELOADFMT
TABLE statement option:
PRINTMISS
Other features: PRINT procedure
Data set:

ENERGY ENERGY on page 1387

Formats: REGFMT., DIVFMT., and USETYPE. on page 1233

This example

3 creates a table that includes all possible combinations of formatted class variable
values (PRELOADFMT with PRINTMISS), even if those combinations have a zero
frequency and even if they do not make sense

3 uses only the preloaded range of user-dened formats as the levels of class
variables (PRELOADFMT with EXCLUSIVE).

3 writes the output to an output data set, and prints that data set.

1238

Program

Chapter 47

Program

Specify the table options. The FORMAT= option species DOLLAR12. as the default format
for the value in each table cell.
proc tabulate data=energy format=dollar12.;

Specify subgroups for the analysis. The CLASS statement separates the analysis by values
of Region, Division, and Type. PRELOADFMT species that PROC TABULATE use the
preloaded values of the user-dened formats for the class variables.
class region division type / preloadfmt;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

Dene the table rows and columns, and specify row and column options. PRINTMISS
species that all possible combinations of user-dened formats be used as the levels of the class
variables.
table region*division,
type*expenditures / rts=25 printmiss;

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

The TABULATE Procedure

Program

1239

Specify the table options and the output data set. The OUT= option species the name of
the output data set to which PROC TABULATE writes the data.
proc tabulate data=energy format=dollar12. out=tabdata;

Specify subgroups for the analysis. The EXCLUSIVE option, when used with
PRELOADFMT, uses only the preloaded range of user-dened formats as the levels of class
variables.
class region division type / preloadfmt exclusive;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

Dene the table rows and columns, and specify row and column options. The
PRINTMISS option is not specied in this case. If it were, then it would override the
EXCLUSIVE option in the CLASS statement.
table region*division,
type*expenditures / rts=25;

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

Print the output data set WORK.TABDATA.

proc print data=tabdata;
run;

1240

Output

Chapter 47

Output

This output, created with the PRELOADFMT and PRINTMISS options, contains all possible
combinations of preloaded user-dened formats for the class variable values. It includes
combinations with zero frequencies, and combinations that make no sense, such as Northeast
and Pacific.

Energy Expenditures for Each Region

(millions of dollars)
--------------------------------------------------|
|
Type
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|
|------------+------------|
|
|Expenditures|Expenditures|
|
|------------+------------|
|
|
Sum
|
Sum
|
|-----------------------+------------+------------|
|Region
|Division
|
|
|
|-----------+-----------|
|
|
|Northeast |New England|
$7,477|
$5,129|
|
|-----------+------------+------------|
|
|Middle
|
|
|
|
|Atlantic
|
$19,379|
$15,078|
|
|-----------+------------+------------|
|
|Mountain
|
.|
.|
|
|-----------+------------+------------|
|
|Pacific
|
.|
.|
|-----------+-----------+------------+------------|
|South
|New England|
.|
.|
|
|-----------+------------+------------|
|
|Middle
|
|
|
|
|Atlantic
|
.|
.|
|
|-----------+------------+------------|
|
|Mountain
|
.|
.|
|
|-----------+------------+------------|
|
|Pacific
|
.|
.|
|-----------+-----------+------------+------------|
|Midwest
|New England|
.|
.|
|
|-----------+------------+------------|
|
|Middle
|
|
|
|
|Atlantic
|
.|
.|
|
|-----------+------------+------------|
|
|Mountain
|
.|
.|
|
|-----------+------------+------------|
|
|Pacific
|
.|
.|
|-----------+-----------+------------+------------|
|West
|New England|
.|
.|
|
|-----------+------------+------------|
|
|Middle
|
|
|
|
|Atlantic
|
.|
.|
|
|-----------+------------+------------|
|
|Mountain
|
$5,476|
$4,729|
|
|-----------+------------+------------|
|
|Pacific
|
$13,959|
$12,619|
---------------------------------------------------

The TABULATE Procedure

Output

1241

This output, created with the PRELOADFMT and EXCLUSIVE options, contains only those
combinations of preloaded user-dened formats for the class variable values that appear in the
input data set. This output is identical to the output from Example 1 on page 1232.

Energy Expenditures for Each Region

(millions of dollars)

--------------------------------------------------|
|
Type
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|
|------------+------------|
|
|Expenditures|Expenditures|
|
|------------+------------|
|
|
Sum
|
Sum
|
|-----------------------+------------+------------|
|Region
|Division
|
|
|
|-----------+-----------|
|
|
|Northeast |New England|
$7,477|
$5,129|
|
|-----------+------------+------------|
|
|Middle
|
|
|
|
|Atlantic
|
$19,379|
$15,078|
|-----------+-----------+------------+------------|
|West
|Mountain
|
$5,476|
$4,729|
|
|-----------+------------+------------|
|
|Pacific
|
$13,959|
$12,619|
---------------------------------------------------

This output is a listing of the output data set TABDATA, which was created by the OUT= option
in the PROC TABULATE statement. TABDATA contains the data that is created by having the
PRELOADFMT and EXCLUSIVE options specied.

Energy Expenditures for Each Region

(millions of dollars)

O
b
s

R
e
g
i
o
n

D
i
v
i
s
i
o
n

1
2
3
4
5
6
7
8

Northeast
Northeast
Northeast
Northeast
West
West
West
West

New England
New England
Middle Atlantic
Middle Atlantic
Mountain
Mountain
Pacific
Pacific

T
y
p
e

_
T
Y
P
E
_

_
P
A
G
E
_

_
T
A
B
L
E
_

E
x
p
e
n
d
i
t
u
r
e
s
_
S
u
m

Residential Customers
Business Customers
Residential Customers
Business Customers
Residential Customers
Business Customers
Residential Customers
Business Customers

111
111
111
111
111
111
111
111

1
1
1
1
1
1
1
1

7477
5129
19379
15078
5476
4729
13959
12619

1242

Example 4: Using Multilabel Formats

Chapter 47

Example 4: Using Multilabel Formats

Procedure features:

CLASS statement options:

MLF
PROC TABULATE statement options:
FORMAT=
TABLE statement
ALL class variable
concatenation (blank) operator
crossing (*) operator
grouping elements (parentheses) operator
label
variable list
Other features:

FORMAT procedure
FORMAT statement
VALUE statement options:
MULTILABEL

This example

3 shows how to specify a multilabel format in the VALUE statement of PROC

FORMAT

3 shows how to activate multilabel format processing using the MLF option with the
CLASS statement

3 demonstrates the behavior of the N statistic when multilabel format processing is

activated.

Create the CARSURVEY data set. CARSURVEY contains data from a survey that was
distributed by a car manufacturer to a focus group of potential customers who were brought
together to evaluate new car names. Each observation in the data set contains an identication
number, the participants age, and the participants ratings of four car names. A DATA step
creates the data set.
data carsurvey;
input Rater Age Progressa Remark Jupiter Dynamo;
datalines;
1
38 94 98 84 80
2
49 96 84 80 77
3
16 64 78 76 73

The TABULATE Procedure

Program

1243

. . . more data lines . . .

77
78
79
80
;

61
24
18
62

92
87
54
90

88
88
50
91

77
88
62
90

85
91
74
86

Create the AGEFMT. format. The FORMAT procedure creates a multilabel format for ages by
using the MULTILABEL option on page 449. A multilabel format is one in which multiple labels
can be assigned to the same value, in this case because of overlapping ranges. Each value is
represented in the table for each range in which it occurs. The NOTSORTED option stores the
ranges in the order in which they are dened.
proc format;
value agefmt (multilabel notsorted)
15 - 29 = Below 30 years
30 - 50 = Between 30 and 50
51 - high = Over 50 years
15 - 19 = 15 to 19
20 - 25 = 20 to 25
25 - 39 = 25 to 39
40 - 55 = 40 to 55
56 - high = 56 and above;
run;

Specify the table options. The FORMAT= option species up to 10 digits as the default
format for the value in each table cell.
proc tabulate data=carsurvey format=10.;

Specify subgroups for the analysis. The CLASS statement identies Age as the class
variable and uses the MLF option to activate multilabel format processing.
class age / mlf;

Specify the analysis variables. The VAR statement species that PROC TABULATE
calculate statistics on the Progressa, Remark, Jupiter, and Dynamo variables.
var progressa remark jupiter dynamo;

Dene the table rows and columns. The row dimension of the TABLE statement creates a
row for each formatted value of Age. Multilabel formatting allows an observation to be included
in multiple rows or age categories. The row dimension uses the ALL class variable to
summarize information for all rows. The column dimension uses the N statistic to calculate the
number of observations for each age group. Notice that the result of the N statistic crossed with
the ALL class variable in the row dimension is the total number of observations instead of the
sum of the N statistics for the rows. The column dimension uses the ALL class variable at the
beginning of a crossing to assign a label, Potential Car Names. The four nested columns
calculate the mean ratings of the car names for each age group.
table age all, n all=Potential Car Names*(progressa remark
jupiter dynamo)*mean;

1244

Output

Chapter 47

Specify the titles.

title1 "Rating Four Potential Car Names";
title2 "Rating Scale 0-100 (100 is the highest rating)";

Format the output. The FORMAT statement assigns the user-dened format AGEFMT. to Age
for this analysis.
format age agefmt.;
run;

Output

Output 47.3
Rating Four Potential Car Names
Rating Scale 0-100 (100 is the highest rating)
--------------------------------------------------------------------------|
|
|
Potential Car Names
|
|
|
|-------------------------------------------|
|
|
|Progressa | Remark | Jupiter | Dynamo |
|
|
|----------+----------+----------+----------|
|
|
N
|
Mean
|
Mean
|
Mean
|
Mean
|
|------------------+----------+----------+----------+----------+----------|
|Age
|
|
|
|
|
|
|------------------|
|
|
|
|
|
|15 to 19
|
14|
75|
78|
81|
73|
|------------------+----------+----------+----------+----------+----------|
|20 to 25
|
11|
89|
88|
84|
89|
|------------------+----------+----------+----------+----------+----------|
|25 to 39
|
26|
84|
90|
82|
72|
|------------------+----------+----------+----------+----------+----------|
|40 to 55
|
14|
85|
87|
80|
68|
|------------------+----------+----------+----------+----------+----------|
|56 and above
|
15|
84|
82|
81|
75|
|------------------+----------+----------+----------+----------+----------|
|Below 30 years
|
36|
82|
84|
82|
75|
|------------------+----------+----------+----------+----------+----------|
|Between 30 and 50 |
25|
86|
89|
81|
73|
|------------------+----------+----------+----------+----------+----------|
|Over 50 years
|
19|
82|
84|
80|
76|
|------------------+----------+----------+----------+----------+----------|
|All
|
80|
83|
86|
81|
74|
---------------------------------------------------------------------------

Example 5: Customizing Row and Column Headings

Procedure features:

TABLE statement
labels
Data set:

ENERGYENERGY on page 1387

Formats:

REGFMT., DIVFMT., and USETYPE. on page 1233

The TABULATE Procedure

Program

1245

This example shows how to customize row and column headings. A label species
text for a heading. A blank label creates a blank heading. PROC TABULATE removes
the space for blank column headings from the table.

Program

Specify the table options. The FORMAT= option species DOLLAR12. as the default format
for the value in each table cell.
proc tabulate data=energy format=dollar12.;

Specify subgroups for the analysis. The CLASS statement identies Region, Division, and
Type as class variables.
class region division type;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

Dene the table rows and columns. The TABLE statement creates a row for each formatted
value of Region. Nested within each row are rows for each formatted value of Division. The
TABLE statement also creates a column for each formatted value of Type. Each cell that is
created by these rows and columns contains the sum of the analysis variable Expenditures for
all observations that contribute to that cell. Text in quotation marks species headings for the
corresponding variable or statistic. Although Sum is the default statistic, it is specied here so
that you can specify a blank for its heading.
table region*division,
type=Customer Base*expenditures= *sum=

Specify the row title space. RTS= provides 25 characters per line for row headings.
/ rts=25;

Format the output. The FORMAT statement assigns formats to Region, Division, and Type.
format region regfmt. division divfmt. type usetype.;

1246

Output

Chapter 47

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

Output

The heading for Type contains text that is specied in the TABLE statement. The TABLE
statement eliminated the headings for Expenditures and Sum.

Energy Expenditures for Each Region

(millions of dollars)

--------------------------------------------------|
|
Customer Base
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|-----------------------+------------+------------|
|Region
|Division
|
|
|
|-----------+-----------|
|
|
|Northeast |New England|
$7,477|
$5,129|
|
|-----------+------------+------------|
|
|Middle
|
|
|
|
|Atlantic
|
$19,379|
$15,078|
|-----------+-----------+------------+------------|
|West
|Mountain
|
$5,476|
$4,729|
|
|-----------+------------+------------|
|
|Pacific
|
$13,959|
$12,619|
---------------------------------------------------

Example 6: Summarizing Information with the Universal Class Variable ALL

Procedure features:

PROC TABULATE statement options:

FORMAT=
TABLE statement:
ALL class variable
concatenation (blank operator)
format modiers
grouping elements (parentheses operator)
Data set: ENERGYENERGY on page 1387
Formats: REGFMT., DIVFMT., and USETYPE. on page 1233

This example shows how to use the universal class variable ALL to summarize
information from multiple categories.

The TABULATE Procedure

Program

1247

Program

Specify the table options. The FORMAT= option species COMMA12. as the default format
for the value in each table cell.
proc tabulate data=energy format=comma12.;

Specify subgroups for the analysis. The CLASS statement identies Region, Division, and
Type as class variables.
class region division type;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

Dene the table rows. The row dimension of the TABLE statement creates a row for each
formatted value of Region. Nested within each row are rows for each formatted value of Division
and a row (labeled Subtotal) that summarizes all divisions in the region. The last row of the
report (labeled Total for All Regions) summarizes all regions. The format modier
f=DOLLAR12. assigns the DOLLAR12. format to the cells in this row.
table region*(division all=Subtotal)
all=Total for All Regions*f=dollar12.,

Dene the table columns. The column dimension of the TABLE statement creates a column
for each formatted value of Type and a column that is labeled All customers that shows
expenditures for all customers in a row of the table. Each cell that is created by these rows and
columns contains the sum of the analysis variable Expenditures for all observations that
contribute to that cell. Text in quotation marks species headings for the corresponding variable
or statistic. Although Sum is the default statistic, it is specied here so that you can specify a
blank for its heading.
type=Customer Base*expenditures= *sum=
all=All Customers*expenditures= *sum=

Specify the row title space. RTS= provides 25 characters per line for row headings.
/ rts=25;

1248

Output

Chapter 47

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

Output

The universal class variable ALL provides subtotals and totals in this
table.

Energy Expenditures for Each Region

(millions of dollars)

The TABULATE Procedure

Program

1249

Example 7: Eliminating Row Headings

Procedure features:

TABLE statement:
labels
ROW=FLOAT
Data set: ENERGYENERGY on page 1387
Formats: REGFMT., DIVFMT., and USETYPE. on page 1233

This example shows how to eliminate blank row headings from a table. To do so, you
must both provide blank labels for the row headings and specify ROW=FLOAT in the
TABLE statement.

Program

Specify the table options. The FORMAT= option species DOLLAR12. as the default format
for the value in each table cell.
proc tabulate data=energy format=dollar12.;

Specify subgroups for the analysis. The CLASS statement identies Region, Division, and
Type as class variables.
class region division type;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

Dene the table rows. The row dimension of the TABLE statement creates a row for each
formatted value of Region. Nested within these rows is a row for each formatted value of
Division. The analysis variable Expenditures and the Sum statistic are also included in the row
dimension, so PROC TABULATE creates row headings for them as well. The text in quotation
marks species the headings for the corresponding variable or statistic. Although Sum is the
default statistic, it is specied here so that you can specify a blank for its heading.
table region*division*expenditures= *sum= ,

1250

Output

Chapter 47

Dene the table columns. The column dimension of the TABLE statement creates a column
for each formatted value of Type.
type=Customer Base

Specify the row title space and eliminate blank row headings. RTS= provides 25
characters per line for row headings. ROW=FLOAT eliminates blank row headings.
/ rts=25 row=float;

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

Output

Compare this table with the output in Example 5 on page 1244. The two tables are identical,
but the program that creates this table uses Expenditures and Sum in the row dimension.
PROC TABULATE automatically eliminates blank headings from the column dimension,
whereas you must specify ROW=FLOAT to eliminate blank headings from the row dimension.

Energy Expenditures for Each Region

(millions of dollars)
--------------------------------------------------|
|
Customer Base
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|-----------------------+------------+------------|
|Region
|Division
|
|
|
|-----------+-----------|
|
|
|Northeast |New England|
$7,477|
$5,129|
|
|-----------+------------+------------|
|
|Middle
|
|
|
|
|Atlantic
|
$19,379|
$15,078|
|-----------+-----------+------------+------------|
|West
|Mountain
|
$5,476|
$4,729|
|
|-----------+------------+------------|
|
|Pacific
|
$13,959|
$12,619|
---------------------------------------------------

The TABULATE Procedure

Program

1251

Example 8: Indenting Row Headings and Eliminating Horizontal Separators

Procedure features:

PROC TABULATE statement options:

NOSEPS
TABLE statement options:
INDENT=
Data set:

ENERGYENERGY on page 1387

Formats: REGFMT., DIVFMT., and USETYPE. on page 1233

This example shows how to condense the structure of a table by

3 removing row headings for class variables

3 indenting nested rows underneath parent rows instead of placing them next to
each other

3 eliminating horizontal separator lines from the row titles and the body of the table.

Program

Specify the table options. The FORMAT= option species DOLLAR12. as the default format
for the value in each table cell. NOSEPS eliminates horizontal separator lines from row titles
and from the body of the table.
proc tabulate data=energy format=dollar12. noseps;

Specify subgroups for the analysis. The CLASS statement identies Region, Division, and
Type as class variables.
class region division type;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

1252

Output

Chapter 47

Dene the table rows and columns. The TABLE statement creates a row for each formatted
value of Region. Nested within each row are rows for each formatted value of Division. The
TABLE statement also creates a column for each formatted value of Type. Each cell that is
created by these rows and columns contains the sum of the analysis variable Expenditures for
all observations that contribute to that cell. Text in quotation marks in all dimensions species
headings for the corresponding variable or statistic. Although Sum is the default statistic, it is
specied here so that you can specify a blank for its heading.
table region*division,
type=Customer Base*expenditures= *sum=

Specify the row title space and indention value. RTS= provides 25 characters per line for
row headings. INDENT= removes row headings for class variables, places values for Division
beneath values for Region rather than beside them, and indents values for Division four spaces.
/ rts=25 indent=4;

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region;
title2 (millions of dollars);
run;

Output

NOSEPS removes the separator lines from the row titles and the body of the table. INDENT=
eliminates the row headings for Region and Division and indents values for Division underneath
values for Region.

Energy Expenditures for Each Region

(millions of dollars)
--------------------------------------------------|
|
Customer Base
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|-----------------------+------------+------------|
|Northeast
|
|
|
|
New England
|
$7,477|
$5,129|
|
Middle Atlantic
|
$19,379|
$15,078|
|West
|
|
|
|
Mountain
|
$5,476|
$4,729|
|
Pacific
|
$13,959|
$12,619|
---------------------------------------------------

The TABULATE Procedure

Program

1253

Example 9: Creating Multipage Tables

Procedure features:

TABLE statement
ALL class variable
BOX=
CONDENSE
INDENT=
page expression
Data set: ENERGY ENERGY on page 1387
Formats: REGFMT., DIVFMT., and USETYPE. on page 1233

This example creates a separate table for each region and one table for all regions.
By default, PROC TABULATE creates each table on a separate page, but the
CONDENSE option places them all on the same page.

Program

Specify the table options. The FORMAT= option species DOLLAR12. as the default format
for the value in each table cell.
proc tabulate data=energy format=dollar12.;

Specify subgroups for the analysis. The CLASS statement identies Region, Division, and
Type as class variables.
class region division type;

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Expenditures variable.
var expenditures;

Dene the table pages. The page dimension of the TABLE statement creates one table for
each formatted value of Region and one table for all regions. Text in quotation marks provides
the heading for each page.
table region=Region: all=All Regions,

1254

Program

Chapter 47

Dene the table rows. The row dimension creates a row for each formatted value of Division
and a row for all divisions. Text in quotation marks provides the row headings.
division all=All Divisions,

Dene the table columns. The column dimension of the TABLE statement creates a column
for each formatted value of Type. Each cell that is created by these pages, rows, and columns
contains the sum of the analysis variable Expenditures for all observations that contribute to
that cell. Text in quotation marks species headings for the corresponding variable or statistic.
Although Sum is the default statistic, it is specied here so that you can specify a blank for its
heading.
type=Customer Base*expenditures= *sum=

Specify additional table options. RTS= provides 25 characters per line for row headings.
BOX= places the page heading inside the box above the row headings. CONDENSE places as
many tables as possible on one physical page. INDENT= eliminates the row heading for
Division. (Because there is no nesting in the row dimension, there is nothing to indent.)
/ rts=25 box=_page_ condense indent=1;

Format the output. The FORMAT statement assigns formats to the variables Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

Specify the titles.

title Energy Expenditures for Each Region and All Regions;
title2 (millions of dollars);
run;

The TABULATE Procedure

Example 10: Reporting on Multiple-Response Survey Data

Output
Energy Expenditures for Each Region and All Regions
(millions of dollars)
--------------------------------------------------|Region: Northeast
|
Customer Base
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|-----------------------+------------+------------|
|New England
|
$7,477|
$5,129|
|-----------------------+------------+------------|
|Middle Atlantic
|
$19,379|
$15,078|
|-----------------------+------------+------------|
|All Divisions
|
$26,856|
$20,207|
---------------------------------------------------

--------------------------------------------------|Region: West
|
Customer Base
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|-----------------------+------------+------------|
|Mountain
|
$5,476|
$4,729|
|-----------------------+------------+------------|
|Pacific
|
$13,959|
$12,619|
|-----------------------+------------+------------|
|All Divisions
|
$19,435|
$17,348|
---------------------------------------------------

--------------------------------------------------|All Regions
|
Customer Base
|
|
|-------------------------|
|
|Residential | Business |
|
| Customers | Customers |
|-----------------------+------------+------------|
|New England
|
$7,477|
$5,129|
|-----------------------+------------+------------|
|Middle Atlantic
|
$19,379|
$15,078|
|-----------------------+------------+------------|
|Mountain
|
$5,476|
$4,729|
|-----------------------+------------+------------|
|Pacific
|
$13,959|
$12,619|
|-----------------------+------------+------------|
|All Divisions
|
$46,291|
$37,555|
---------------------------------------------------

Example 10: Reporting on Multiple-Response Survey Data

Procedure features:

TABLE statement:
denominator denition (angle bracket operators)
N statistic
PCTN statistic
variable list
Other features:

FORMAT procedure

1255

1256

Collecting the Data

Chapter 47

SAS system options:

FORMDLIM=
NONUMBER
SYMPUT routine

The two tables in this example show

3 which factors most inuenced customers decisions to buy products
3 where customers heard of the company.
The reports appear on one physical page with only one page number. By default, they
would appear on separate pages.
In addition to showing how to create these tables, this example shows how to
3 use a DATA step to count the number of observations in a data set
3 store that value in a macro variable

3 access that value later in the SAS session.

Collecting the Data

Figure 47.15 on page 1256 shows the survey form that is used to collect data.

Figure 47.15

Completed Survey Form

Customer Questionnaire

ID#:

Please place a check beside all answers that apply.

Why do you buy our products?

Cost

Performance

Reliability

Sales staff

How did you find out about our company?

T.V. / Radio

Newspaper / Magazine

Word of mouth

What makes a sales person effective?

Product knowledge

Personality

Appearance

The TABULATE Procedure

Program

1257

Program

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. The FORMDLIM=
option replaces the character that delimits page breaks with a single blank. By default, a new
physical page starts whenever a page break occurs.
options nodate pageno=1 linesize=80 pagesize=18 formdlim= ;

Create the CUSTOMER_RESPONSE data set. CUSTOMER_RESPONSE contains data from

a customer survey. Each observation in the data set contains information about factors that
inuence one respondents decisions to buy products. A DATA step on page 1380 creates the
data set. Using missing values rather than 0s is crucial for calculating frequency counts in
PROC TABULATE.
data customer_response;
input Customer Factor1-Factor4 Source1-Source3
Quality1-Quality3;
datalines;
1 . . 1 1 1 1 . 1 . .
2 1 1 . 1 1 1 . 1 1 .
3 . . 1 1 1 1 . . . .
. . . more data lines . . .
119 . . . 1 . . . 1 . .
120 1 1 . 1 . . . . 1 .
;

Store the number of observations in a macro variable. The SET statement reads the
descriptor portion of CUSTOMER_RESPONSE at compile time and stores the number of
observations (the number of respondents) in COUNT. The SYMPUT routine stores the value of
COUNT in the macro variable NUM. This variable is available for use by other procedures and
DATA steps for the remainder of the SAS session. The IF 0 condition, which is always false,
ensures that the SET statement, which reads the observations, never executes. (Reading
observations is unnecessary.) The STOP statement ensures that the DATA step executes only
once.
data _null_;
if 0 then set customer_response nobs=count;
call symput(num,left(put(count,4.)));
stop;
run;

1258

Program

Chapter 47

Create the PCTFMT. format. The FORMAT procedure creates a format for percentages. The
PCTFMT. format writes all values with at least one digit to the left of the decimal point and
with one digit to the right of the decimal point. A blank and a percent sign follow the digits.
proc format;
picture pctfmt low-high=009.9 %;
run;

Create the report and use the default table options.

proc tabulate data=customer_response;

Specify the analysis variables. The VAR statement species that PROC TABULATE
calculate statistics on the Factor1, Factor2, Factor3, Factor4, and Customer variables. The
variable Customer must be listed because it is used to calculate the Percent column that is
dened in the TABLE statement.
var factor1-factor4 customer;

Dene the table rows and columns. The TABLE statement creates a row for each factor, a
column for frequency counts, and a column for the percentages. Text in quotation marks
supplies headers for the corresponding row or column. The format modiers F=7. and
F=PCTFMT9. provide formats for values in the associated cells and extend the column widths to
accommodate the column headers.
table factor1=Cost
factor2=Performance
factor3=Reliability
factor4=Sales Staff,
(n=Count*f=7. pctn<customer>=Percent*f=pctfmt9.) ;

Specify the titles.

title Customer Survey Results: Spring 1996;
title3 Factors Influencing the Decision to Buy;
run;

Suppress page numbers. The SAS system option NONUMBER suppresses page numbers for
subsequent pages.
options nonumber;

Create the report and use the default table options.

proc tabulate data=customer_response;

The TABULATE Procedure

Program

1259

Specify the analysis variables. The VAR statement species that PROC TABULATE
calculate statistics on the Source1, Source2, Source3, and Customer variables. The variable
Customer must be in the variable list because it appears in the denominator denition.
var source1-source3 customer;

Dene the table rows and columns. The TABLE statement creates a row for each source of
the company name, a column for frequency counts, and a column for the percentages. Text in
quotation marks supplies a heading for the corresponding row or column.
table source1=TV/Radio
source2=Newspaper
source3=Word of Mouth,
(n=Count*f=7. pctn<customer>=Percent*f=pctfmt9.) ;

Specify the title and footnote. The macro variable NUM resolves to the number of
respondents. The FOOTNOTE statement uses double rather than single quotation marks so
that the macro variable will resolve.
title Source of Company Name;
footnote "Number of Respondents: &num";
run;

Reset the SAS system options. The FORMDLIM= option resets the page delimiter to a page
eject. The NUMBER option resumes the display of page numbers on subsequent pages.
options formdlim= number;

1260

Output

Chapter 47

Output

Customer Survey Results: Spring 1996

Factors Influencing the Decision to Buy

-------------------------------------|
| Count | Percent |
|------------------+-------+---------|
|Cost
|
87|
72.5 %|
|------------------+-------+---------|
|Performance
|
62|
51.6 %|
|------------------+-------+---------|
|Reliability
|
30|
25.0 %|
|------------------+-------+---------|
|Sales Staff
|
120| 100.0 %|
--------------------------------------

Source of Company Name

-------------------------------------|
| Count | Percent |
|------------------+-------+---------|
|TV/Radio
|
92|
76.6 %|
|------------------+-------+---------|
|Newspaper
|
69|
57.5 %|
|------------------+-------+---------|
|Word of Mouth
|
26|
21.6 %|
--------------------------------------

Number of Respondents: 120

Example 11: Reporting on Multiple-Choice Survey Data

Procedure features:

TABLE statement:
N statistic
Other features:

FORMAT procedure
TRANSPOSE procedure
Data set options:
RENAME=

This report of listener preferences shows how many listeners select each type of
programming during each of seven time periods on a typical weekday. The data was
collected by a survey, and the results were stored in a SAS data set. Although this data

The TABULATE Procedure

Collecting the Data

1261

set contains all the information needed for this report, the information is not arranged
in a way that PROC TABULATE can use.
To make this crosstabulation of time of day and choice of radio programming, you
must have a data set that contains a variable for time of day and a variable for
programming preference. PROC TRANSPOSE reshapes the data into a new data set
that contains these variables. Once the data are in the appropriate form, PROC
TABULATE creates the report.

Collecting the Data

Figure 47.16 on page 1261 shows the survey form that is used to collect data.

Figure 47.16

Completed Survey Form

phone_ _ _
LISTENER SURVEY
1. _______

What is your age?

2. _______

What is your gender?

3. _______

On the average WEEKDAY, how many hours do you listen

to the radio?

4. _______

On the average WEEKEND-DAY, how many hours do you

listen to the radio?

Use codes 1-8 for question 5. Use codes 0-8 for 6-19.
0 Do not listen at that time
1
2
3
4
5. _______

Rock
Top 40
Country
Jazz

5
6
7
8

Classical
Easy Listening
News/Information/Talk
Other

What style of music or radio programming do you most

often listen to?

On a typical WEEKDAY,
what kind of radio programming do you listen to

On a typical WEEKEND-DAY,
what kind of radio programming
do you listen to

6. _______ from 6-9 a.m.?

13.

_______ from 6-9 a.m.?

7. _______ from 9 a.m. to noon?

14.

_______ from 9 a.m. to noon?

8. _______ from noon to 1 p.m.?

15.

_______ from noon to 1 p.m.?

9. _______ from 1-4 p.m.?

16.

_______ from 1-4 p.m.?

10. _______ from 4-6 p.m.?

17.

_______ from 4-6 p.m.?

11. _______ from 6-10 p.m.?

18.

_______ from 6-10 p.m.?

12. _______ from 10 p.m. to 2 a.m.?

19.

_______ from 10 p.m. to 2 a.m.?

An external le on page 1405 contains the raw data for the survey. Several lines
from that le appear here.
967
7 5
781
5 0

32 f 5 3 5
5 5 7 0 0 0 8 7 0 0 8 0
30 f 2 3 5
0 0 5 0 0 0 4 7 5 0 0 0

1262

Program

Chapter 47

859 39 f 1 0 5
1 0 0 0 1 0 0 0 0 0 0 0 0 0
. . . more data lines . . .
859 32 m .25 .25 1
1 0 0 0 0 0 0 0 1 0 0 0 0 0

Program

Create the RADIO data set and specify the input le. RADIO contains data from a survey
of 336 listeners. The data set contains information about listeners and their preferences in radio
programming. The INFILE statement species the external le that contains the data.
MISSOVER prevents the input pointer from going to the next record if it fails to nd values in
the current line for all variables that are listed in the INPUT statement.
data radio;
infile input-file missover;

Read the appropriate data line, assign a unique number to each respondent, and
write an observation to RADIO. Each raw-data record contains two lines of information
about each listener. The INPUT statement reads only the information that this example needs.
The / line control skips the rst line of information in each record. The rest of the INPUT
statement reads Time1-Time7 from the beginning of the second line. These variables represent
the listeners radio programming preference for each of seven time periods on weekdays (see
Figure 47.16 on page 1261). The listener=_N_ statement assigns a unique identier to each
listener. An observation is automatically written to RADIO at the end of each iteration.
input /(Time1-Time7) ($1. +1);
listener=_n_;
run;

Create the $TIMEFMT. and $PGMFMT. formats. PROC FORMAT creates formats for the
time of day and the choice of programming.
proc format;
value $timefmt Time1=6-9 a.m.
Time2=9 a.m. to noon
Time3=noon to 1 p.m.
Time4=1-4 p.m.
Time5=4-6 p.m.
Time6=6-10 p.m.
Time7=10 p.m. to 2 a.m.

The TABULATE Procedure

Program

1263

other=* Data Entry Error *;

0="Dont Listen"
1,2=Rock and Top 40
3=Country
4,5,6=Jazz, Classical, and Easy Listening
7=News/ Information /Talk
8=Other
other=*** Data Entry Error ***;

value $pgmfmt

run;

Reshape the data by transposing the RADIO data set. PROC TRANSPOSE creates
RADIO_TRANSPOSED. This data set contains the variable Listener from the original data set.
It also contains two transposed variables: Timespan and Choice. Timespan contains the names
of the variables (Time1-Time7) from the input data set that are transposed to form observations
in the output data set. Choice contains the values of these variables. (See A Closer Look on
page 1264 for a complete explanation of the PROC TRANSPOSE step.)
proc transpose data=radio
out=radio_transposed(rename=(col1=Choice))
name=Timespan;
by listener;
var time1-time7;

Format the transposed variables. The FORMAT statement permanently associates these
formats with the variables in the output data set.
format timespan $timefmt. choice $pgmfmt.;
run;

Create the report and specify the table options. The FORMAT= option species the default
format for the values in each table cell.
proc tabulate data=radio_transposed format=12.;

Specify subgroups for the analysis. The CLASS statement identies Timespan and Choice
as class variables.
class timespan choice;

Dene the table rows and columns. The TABLE statement creates a row for each formatted
value of Timespan and a column for each formatted value of Choice. In each column are values
for the N statistic. Text in quotation marks supplies headings for the corresponding rows or
columns.
table timespan=Time of Day,
choice=Choice of Radio Program*n=Number of Listeners;

1264

Output

Chapter 47

Specify the title.

title Listening Preferences on Weekdays;
run;

Output

Listening Preferences on Weekdays

--------------------------------------------------------------------------------------------------------------|
|

|
Choice of Radio Program
|
|-----------------------------------------------------------------------------|

|
|
|

|Dont Listen|
40
| Country
| Listening |
/Talk
|
Other
|
|------------+------------+------------+------------+------------+------------|
| Number of | Number of | Number of | Number of | Number of | Number of |

|
|
|

|
| Listeners | Listeners | Listeners | Listeners | Listeners | Listeners |
|-------------------------------+------------+------------+------------+------------+------------+------------|
|Time of Day
|
|
|
|
|
|
|
|-------------------------------|
|6-9 a.m.
|

|
34|

|
143|

|
7|

|
39|

|
96|

|
17|

|-------------------------------+------------+------------+------------+------------+------------+------------|
|9 a.m. to noon
|
214|
59|
5|
51|
3|
4|
|-------------------------------+------------+------------+------------+------------+------------+------------|
|noon to 1 p.m.
|
238|
55|
3|
27|
9|
4|
|-------------------------------+------------+------------+------------+------------+------------+------------|
|1-4 p.m.
|
216|
60|
5|
50|
2|
3|
|-------------------------------+------------+------------+------------+------------+------------+------------|
|4-6 p.m.
|
56|
130|
6|
57|
69|
18|
|-------------------------------+------------+------------+------------+------------+------------+------------|
|6-10 p.m.
|
202|
54|
9|
44|
20|
7|
|-------------------------------+------------+------------+------------+------------+------------+------------|
|10 p.m. to 2 a.m.
|
264|
29|
3|
36|
2|
2|
---------------------------------------------------------------------------------------------------------------

A Closer Look
Reshape the data
The original input data set has all the information that you need to make the
crosstabular report, but PROC TABULATE cannot use the information in that form.
PROC TRANSPOSE rearranges the data so that each observation in the new data set
contains the variable Listener, a variable for time of day, and a variable for
programming preference. Figure 47.17 on page 1265 illustrates the transposition.
PROC TABULATE uses this new data set to create the crosstabular report.
PROC TRANSPOSE restructures data so that values that were stored in one
observation are written to one variable. You can specify which variables you want to
transpose. This section illustrates how PROC TRANSPOSE reshapes the data. The
following section explains the PROC TRANSPOSE step in this example.
When you transpose with BY processing, as this example does, you create from each
BY group one observation for each variable that you transpose. In this example,
Listener is the BY variable. Each observation in the input data set is a BY group
because the value of Listener is unique for each observation.
This example transposes seven variables, Time1 through Time7. Therefore, the
output data set has seven observations from each BY group (each observation) in the
input data set.

The TABULATE Procedure

Figure 47.17

A Closer Look

1265

Transposing Two Observations

Input Data Set

Time1

Time2

Time3

Time4

Time5

Time6

Time7

Listener

_NAME_

COL1

Time1

Time2

Time3

Time4

Time5

Time6

Time7

Time1

Time2

Time3

Time4

Time5

Time6

Time7

The BY variable is not

transposed. All the

observations created from the
same BY group contain the
same value of Listener.

_NAME_ contains the name

of the variable in the input
data set that was transposed
to create the current
observation in the output
data set.

COL1 contains the values of

Time1Time7.

Output Data Set

Understanding the PROC TRANSPOSE Step

Here is a detailed explanation of the PROC TRANSPOSE step that reshapes the data:

1266

Example 12: Calculating Various Percentage Statistics

Chapter 47

proc transpose data=radio u

out=radio_transposed(rename=(col1=Choice))
name=Timespan;
w
by listener;
x
var time1-time7;
y
format timespan $timefmt. choice $pgmfmt.;
U
run;

u The DATA= option species the input data set.

v The OUT= option species the output data set. The RENAME= data set option
renames the transposed variable from COL1 (the default name) to Choice.
w The NAME= option species the name for the variable in the output data set that
contains the name of the variable that is being transposed to create the current
observation. By default, the name of this variable is _NAME_.
x The BY statement identies Listener as the BY variable.
y The VAR statement identies Time1 through Time7 as the variables to transpose.
U The FORMAT statement assigns formats to Timespan and Choice. The PROC
TABULATE step that creates the report does not need to format Timespan and
Choice because the formats are stored with these variables.

Example 12: Calculating Various Percentage Statistics

Procedure features:

PROC TABULATE statement options:

FORMAT=
TABLE statement:
ALL class variable
COLPCTSUM statistic
concatenation (blank) operator
crossing (*) operator
format modiers
grouping elements (parentheses) operator
labels
REPPCTSUM statistic
ROWPCTSUM statistic
variable list
TABLE statement options:
ROW=FLOAT
RTS=
Other features: FORMAT procedure

This example shows how to use three percentage sum statistics: COLPCTSUM,
REPPCTSUM, and ROWPCTSUM.

The TABULATE Procedure

Program

1267

Create the FUNDRAIS data set. FUNDRAIS contains data on student sales during a school
fund-raiser. A DATA step creates the data set.
data fundrais;
length name $ 8 classrm $ 1;
input @1 team $ @8 classrm $ @10 name $
@19 pencils @23 tablets;
sales=pencils + tablets;
datalines;
BLUE
A ANN
4
8
RED
A MARY
5 10
GREEN A JOHN
6
4
RED
A BOB
2
3
BLUE
B FRED
6
8
GREEN B LOUISE
12
2
BLUE
B ANNETTE
.
9
RED
B HENRY
8 10
GREEN A ANDREW
3
5
RED
A SAMUEL
12 10
BLUE
A LINDA
7 12
GREEN A SARA
4
.
BLUE
B MARTIN
9 13
RED
B MATTHEW
7
6
GREEN B BETH
15 10
RED
B LAURA
4
3
;

Create the PCTFMT. format. The FORMAT procedure creates a format for percentages. The
PCTFMT. format writes all values with at least one digit, a blank, and a percent sign.
proc format;
picture pctfmt low-high=009 %;
run;

Specify the title.

title "Fundraiser Sales";

Create the report and specify the table options. The FORMAT= option species up to
seven digits as the default format for the value in each table cell.
proc tabulate format=7.;

Specify subgroups for the analysis. The CLASS statement identies Team and Classrm as
class variables.
class team classrm;

1268

Output

Chapter 47

Specify the analysis variable. The VAR statement species that PROC TABULATE calculate
statistics on the Sales variable.
var sales;

Dene the table rows. The row dimension of the TABLE statement creates a row for each
formatted value of Team. The last row of the report summarizes sales for all teams.
table (team all),

Dene the table columns. The column dimension of the TABLE statement creates a column
for each formatted value of Classrm. Crossed within each value of Classrm is the analysis
variable (sales) with a blank label. Nested within each column are columns that summarize
sales for the class.

3 The rst nested column, labeled sum, is the sum of sales for the row for the classroom.
3 The second nested column, labeled ColPctSum, is the percentage of the sum of sales for the
row for the classroom in relation to the sum of sales for all teams in the classroom.

3 The third nested column, labeled RowPctSum, is the percentage of the sum of sales for the
row for the classroom in relation to the sum of sales for the row for all classrooms.

3 The fourth nested column, labeled RepPctSum, is the percentage of the sum of sales for the
row for the classroom in relation to the sum of sales for all teams for all classrooms.
The last column of the report summarizes sales for the row for all classrooms.
classrm=Classroom*sales= *(sum
colpctsum*f=pctfmt9.
rowpctsum*f=pctfmt9.
reppctsum*f=pctfmt9.)
all*sales*sum=

Specify the row title space and eliminate blank row headings. RTS= provides 20
characters per line for row headings.
/rts=20;
run;

Output
Fundraiser Sales

-------------------------------------------------------------------------------------------------------|
|
|

|
Classroom
|
|---------------------------------------------------------------------------|
|
A
|
B
|

|
|

|-------------------------------------+-------------------------------------+-------|
| Sum |ColPctSum|RowPctSum|RepPctSum| Sum |ColPctSum|RowPctSum|RepPctSum| Sum |

All

|
|
|

|------------------+-------+---------+---------+---------+-------+---------+---------+---------+-------|
|team
|
|
|
|
|
|
|
|
|
|
|------------------|
|
|
|
|
|
|
|
|
|
|BLUE
|
31|
34 %|
46 %|
15 %|
36|
31 %|
53 %|
17 %|
67|
|------------------+-------+---------+---------+---------+-------+---------+---------+---------+-------|
|GREEN
|
18|
19 %|
31 %|
8 %|
39|
34 %|
68 %|
19 %|
57|
|------------------+-------+---------+---------+---------+-------+---------+---------+---------+-------|
|RED
|
42|
46 %|
52 %|
20 %|
38|
33 %|
47 %|
18 %|
80|
|------------------+-------+---------+---------+---------+-------+---------+---------+---------+-------|
|All
|
91|
100 %|
44 %|
44 %|
113|
100 %|
55 %|
55 %|
204|
--------------------------------------------------------------------------------------------------------

The TABULATE Procedure

Program

1269

A Closer Look
Here are the percentage sum statistic calculations used to produce the output for the
Blue Team in Classroom A:
COLPCTSUM=31/91*100=34%
ROWPCTSUM=31/67*100=46%
REPPCTSUM=31/204*100=15%
Similar calculations were used to produce the output for the remaining teams and
classrooms.

Example 13: Using Denominator Denitions to Display Basic Frequency

Counts and Percentages
Procedure features:

TABLE statement:
ALL class variable
denominator denitions (angle bracket operators)
N statistic
PCTN statistic
Other features:

FORMAT procedure

Crosstabulation tables (also called contingency tables and stub-and-banner reports)

show combined frequency distributions for two or more variables. This table shows
frequency counts for females and males within each of four job classes. The table also
shows the percentage that each frequency count represents of
3 the total women and men in that job class (row percentage)

3 the total for that gender in all job classes (column percentage)
3 the total for all employees.

Program

Create the JOBCLASS data set. JOBCLASS contains encoded information about the gender
and job class of employees at a ctitious company.
data jobclass;
input Gender Occupation @@;

1270

Program

Chapter 47

1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
;

datalines;
1 1 1 1 1
2 1 2 1 2
3 1 3 1 3
1 1 1 1 1
2 1 2 1 3
4 1 4 1 4
1 1 2 1 2
2 1 3 1 3
4 1 4 1 4
1 2 1 2 1
2 2 2 2 2
4 2 4 2 4
3 2 3 2 3
4 2 4 2 1
2 2 2 2 2
3 2 3 2 4
1 2 1 2 1
3 2 3 2 3

1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2

1
2
3
2
3
1
2
3
1
1
3
4
3
1
2
4
2
4

1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2

1
2
3
2
4
1
2
3
3
1
3
4
4
1
2
4
2

1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2

1
2
3
2
4
1
2
4
1
2
3
1
4
1
2
1
2

1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2

1
2
3
2
4
1
2
4
1
2
4
3
4
1
2
1
3

Create the GENDFMT. and OCCUPFMT. formats. PROC FORMAT creates formats for the
variables Gender and Occupation.
proc format;
value gendfmt 1=Female
2=Male
other=*** Data Entry Error ***;
value occupfmt 1=Technical
2=Manager/Supervisor
3=Clerical
4=Administrative
other=*** Data Entry Error ***;
run;

Create the report and specify the table options. The FORMAT= option species the 8.2
format as the default format for the value in each table cell.
proc tabulate data=jobclass format=8.2;

Specify subgroups for the analysis. The CLASS statement identies Gender and Occupation
as class variables.
class gender occupation;

The TABULATE Procedure

Program

1271

Dene the table rows. The TABLE statement creates a set of rows for each formatted value of
Occupation and for all jobs together. Text in quotation marks supplies a header for the
corresponding row.
The asterisk in the row dimension indicates that the statistics that follow in parentheses are
nested within the values of Occupation and All to form sets of rows. Each set of rows includes
four statistics:

3 N, the frequency count. The format modier (F=9.) writes the values of N without the
decimal places that the default format would use. It also extends the column width to nine
characters so that the word Employees ts on one line.

3 the percentage of the row total (row percent).

3 the percentage of the column total (column percent).
3 the overall percent. Text in quotation marks supplies the heading for the corresponding row.
A comma separates the row denition from the column denition.
For detailed explanations of the structure of this table and of the use of denominator denitions,
see A Closer Look on page 1272.
table (occupation=Job Class all=All Jobs)
*(n=Number of employees*f=9.
pctn<gender all>=Percent of row total
pctn<occupation all>=Percent of column total
pctn=Percent of total),

Dene the table columns and specify the amount of space for row headings. The
column dimension creates a column for each formatted value of Gender and for all employees.
Text in quotation marks supplies the heading for the corresponding column. The RTS= option
provides 50 characters per line for row headings.
gender=Gender all=All Employees/ rts=50;

Format the output. The FORMAT statement assigns formats to the variables Gender and
Occupation.
format gender gendfmt. occupation occupfmt.;

Specify the titles.

title Gender Distribution;
title2 within Job Classes;
run;

1272

Output

Chapter 47

Output
Gender Distribution
within Job Classes

-------------------------------------------------------------------------------|
|
Gender
|
|
|
|-------------------|
All
|
|
| Female | Male
|Employees|
|------------------------------------------------+---------+---------+---------|
|Job Class
|
|
|
|
|
|-----------------------+------------------------|
|
|
|
|Technical
|Number of employees
|
16|
18|
34|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
47.06|
52.94|
100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
26.23|
29.03|
27.64|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
13.01|
14.63|
27.64|
|-----------------------+------------------------+---------+---------+---------|
|Manager/Supervisor
|Number of employees
|
20|
15|
35|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
57.14|
42.86|
100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
32.79|
24.19|
28.46|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
16.26|
12.20|
28.46|
|-----------------------+------------------------+---------+---------+---------|
|Clerical
|Number of employees
|
14|
14|
28|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
50.00|
50.00|
100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
22.95|
22.58|
22.76|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
11.38|
11.38|
22.76|
|-----------------------+------------------------+---------+---------+---------|
|Administrative
|Number of employees
|
11|
15|
26|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
42.31|
57.69|
100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
18.03|
24.19|
21.14|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
8.94|
12.20|
21.14|
|-----------------------+------------------------+---------+---------+---------|
|All Jobs
|Number of employees
|
61|
62|
123|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
49.59|
50.41|
100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
100.00|
100.00|
100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
49.59|
50.41|
100.00|
--------------------------------------------------------------------------------

A Closer Look
The part of the TABLE statement that denes the rows of the table uses the PCTN
statistic to calculate three different percentages.
In all calculations of PCTN, the numerator is N, the frequency count for one cell of
the table. The denominator for each occurrence of PCTN is determined by the
denominator denition. The denominator denition appears in angle brackets after the
keyword PCTN. It is a list of one or more expressions. The list tells PROC TABULATE
which frequency counts to sum for the denominator.

The TABULATE Procedure

A Closer Look

1273

Analyzing the Structure of the Table

Taking a close look at the structure of the table helps you understand how PROC
TABULATE uses the denominator denitions. The following simplied version of the
TABLE statement claries the basic structure of the table:
table occupation=Job Class all=All Jobs,
gender=Gender all=All Employees;

The table is a concatenation of four subtables. In this report, each subtable is a

crossing of one class variable in the row dimension and one class variable in the column
dimension. Each crossing establishes one or more categories. A category is a
combination of unique values of class variables, such as female, technical or all,
clerical. Table 47.8 on page 1273 describes each subtable.
Table 47.8

Contents of Subtables

Class variables contributing to the

subtable

Description of frequency counts

Number of
categories

Occupation and Gender

number of females in each job or

number of males in each job

All and Gender

number of females or number of males

Occupation and All

number of people in each job

All and All

number of people in all jobs

Figure 47.18 on page 1274 highlights these subtables and the frequency counts for each
category.

1274

A Closer Look

Chapter 47

Figure 47.18

Illustration of the Four Subtables

Occupation
and All
Occupation and Gender
---------------------------------------------------------------------|
|
Gender
|
|
|-------------------|
|
| Female | Male |
|------------------------------------------------+---------+---------+
|Job Class
|
|
|
|
|-----------------------+------------------------|
|
|
|Technical
|Number of employees
|
16|
18|
|
|------------------------+---------+---------+
|
|Percent of row total
|
47.06|
52.94|
|
|------------------------+---------+---------+
|
|Percent of column total |
26.23|
29.03|
|
|------------------------+---------+---------+
|
|Percent of total
|
13.01|
14.63|
|-----------------------+------------------------+---------+---------+
|Manager/Supervisor
|Number of employees
|
20|
15|
|
|------------------------+---------+---------+
|
|Percent of row total
|
57.14|
42.86|
|
|------------------------+---------+---------+
|
|Percent of column total |
32.79|
24.19|
|
|------------------------+---------+---------+
|
|Percent of total
|
16.26|
12.20|
|-----------------------+------------------------+---------+---------+
|Clerical
|Number of employees
|
14|
14|
|
|------------------------+---------+---------+
|
|Percent of row total
|
50.00|
50.50|
|
|------------------------+---------+---------+
|
|Percent of column total |
22.95|
22.58|
|
|------------------------+---------+---------+
|
|Percent of total
|
11.38|
11.38|
|-----------------------+------------------------+---------+---------+
|Administrative
|Number of employees
|
11|
15|
|
|------------------------+---------+---------+
|
|Percent of row total
|
42.31|
57.69|
|
|------------------------+---------+---------+
|
|Percent of column total |
18.03|
24.19|
|
|------------------------+---------+---------+
|
|Percent of total
|
8.94|
12.20|
|-----------------------+------------------------+---------+---------+

----------|
|
| All |
|Employees|
+---------|
|
|
|
|
|
34|
+---------|
| 100.00|
+---------|
|
27.64|
+---------|
|
27.64|
+---------|
|
35|
+---------|
| 100.00|
+---------|
|
28.46|
+---------|
|
28.46|
+---------|
|
28|
+---------|
| 100.00|
+---------|
|
22.76|
+---------|
|
22.76|
+---------|
|
26|
+---------|
| 100.00|
+---------|
|
21.14|
+---------|
|
21.14|
+---------|

|-----------------------+------------------------+---------+---------+
|All Jobs
|Number of employees
|
61|
62|
|
|------------------------+---------+---------+
|
|Percent of row total
|
49.59|
50.41|
|
|------------------------+---------+---------+
|
|Percent of column total | 100.00| 100.00|
|
|------------------------+---------+---------+
|
|Percent of total
|
49.59|
50.41|
----------------------------------------------------------------------

All and Gender

+---------|
|
123|
+---------|
| 100.00|
+---------|
| 100.00|
+---------|
| 100.00|
-----------

All
and All

Interpreting Denominator Denitions

The following fragment of the TABLE statement denes the denominator denitions for
this report. The PCTN keyword and the denominator denitions are highlighted.
table (occupation=Job Class all=All Jobs)
*(n=Number of employees*f=5.
pctn<gender all>=Row percent
pctn<occupation all>=Column percent
pctn=Percent of total),

Each use of PCTN nests a row of statistics within each value of Occupation and All.
Each denominator denition tells PROC TABULATE which frequency counts to sum for
the denominators in that row. This section explains how PROC TABULATE interprets
these denominator denitions.
Row Percentages

The part of the TABLE statement that calculates the row percentages and that labels
the row is
pctn<gender all>=Row percent

Consider how PROC TABULATE interprets this denominator denition for each
subtable.

The TABULATE Procedure

A Closer Look

1275

Subtable 1: Occupation and Gender

Gender Distribution
within Job Classes
-------------------------------------------------------------------------------|
|
Gender
|
|
|
|-------------------| All |
|
| Female | Male |Employees|
|------------------------------------------------+---------+---------+---------|
|Job Class
|
|
|
|
|
|-----------------------+------------------------|
|
|
|
|Technical
|Number of employees
|
16|
18|
34|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
47.06|
52.94| 100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
26.23|
29.03|
27.64|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
13.01|
14.63|
27.64|
|-----------------------+------------------------+---------+---------+---------|
|Manager/Supervisor
|Number of employees
|
20|
15|
35|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
57.14|
42.86| 100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
32.79|
24.19|
28.46|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
16.26|
12.20|
28.46|
|-----------------------+------------------------+---------+---------+---------|
|Clerical
|Number of employees
|
14|
14|
28|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
50.00|
50.50| 100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
22.95|
22.58|
22.76|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
11.38|
11.38|
22.76|
|-----------------------+------------------------+---------+---------+---------|
|Administrative
|Number of employees
|
11|
15|
26|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
42.31|
57.69| 100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total |
18.03|
24.19|
21.14|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
8.94|
12.20|
21.14|
|-----------------------+------------------------+---------+---------+---------|
|All Jobs
|Number of employees
|
61|
62|
123|
|
|------------------------+---------+---------+---------|
|
|Percent of row total
|
49.59|
50.41| 100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of column total | 100.00| 100.00| 100.00|
|
|------------------------+---------+---------+---------|
|
|Percent of total
|
49.59|
50.41| 100.00|
--------------------------------------------------------------------------------

PROC TABULATE looks at the rst element in the denominator denition, Gender,
and asks if Gender contributes to the subtable. Because Gender does contribute to the
subtable, PROC TABULATE uses it as the denominator denition. This denominator
denition tells PROC TABULATE to sum the frequency counts for all occurrences of
Gender within the same value of Occupation.
For example, the denominator for the category female, technical is the sum of all
frequency counts for all categories in this subtable for which the value of Occupation is
technical. There are two such categories: female, technical and male,
technical. The corresponding frequency counts are 16 and 18. Therefore, the
denominator for this category is 16+18, or 34.

Subtable 2: All and Gender

PROC TABULATE looks at the rst element in the denominator denition, Gender,
and asks if Gender contributes to the subtable. Because Gender does contribute to the
subtable, PROC TABULATE uses it as the denominator denition. This denominator
denition tells PROC TABULATE to sum the frequency counts for all occurrences of
Gender in the subtable.
For example, the denominator for the category all, female is the sum of the
frequency counts for all, female and all, male. The corresponding frequency counts
are 61 and 62. Therefore, the denominator for cells in this subtable is 61+62, or 123.

1276

A Closer Look

Chapter 47

Subtable 3: Occupation and All

PROC TABULATE looks at the rst element in the denominator denition, Gender,
and asks if Gender contributes to the subtable. Because Gender does not contribute to
the subtable, PROC TABULATE looks at the next element in the denominator
denition, which is All. The variable All does contribute to this subtable, so PROC
TABULATE uses it as the denominator denition. All is a reserved class variable with
only one category. Therefore, this denominator denition tells PROC TABULATE to use
the frequency count of All as the denominator.
For example, the denominator for the category clerical, all is the frequency
count for that category, 28.
Note: In these table cells, because the numerator and the denominator are the
same, the row percentages in this subtable are all 100. 4

Subtable 4: All and All

PROC TABULATE looks at the rst element in the denominator denition, Gender,
and asks if Gender contributes to the subtable. Because Gender does not contribute to
the subtable, PROC TABULATE looks at the next element in the denominator
denition, which is All. The variable All does contribute to this subtable, so PROC
TABULATE uses it as the denominator denition. All is a reserved class variable with
only one category. Therefore, this denominator denition tells PROC TABULATE to use
the frequency count of All as the denominator.
There is only one category in this subtable: all, all. The denominator for this
category is 123.
Note: In this table cell, because the numerator and denominator are the same, the
row percentage in this subtable is 100. 4
Column Percentages

The part of the TABLE statement that calculates the column percentages and labels the
row is
pctn<occupation all>=Column percent

Consider how PROC TABULATE interprets this denominator denition for each
subtable.

The TABULATE Procedure

A Closer Look

1277

Subtable 1: Occupation and Gender

PROC TABULATE looks at the rst element in the denominator denition,

Occupation, and asks if Occupation contributes to the subtable. Because Occupation
does contribute to the subtable, PROC TABULATE uses it as the denominator
denition. This denominator denition tells PROC TABULATE to sum the frequency
counts for all occurrences of Occupation within the same value of Gender.
For example, the denominator for the category manager/supervisor, male is the
sum of all frequency counts for all categories in this subtable for which the value of
Gender is male. There are four such categories: technical, male; manager/
supervisor, male; clerical, male; and administrative, male. The corresponding
frequency counts are 18, 15, 14, and 15. Therefore, the denominator for this category is
18+15+14+15, or 62.

Subtable 2: All and Gender

PROC TABULATE looks at the rst element in the denominator denition,

Occupation, and asks if Occupation contributes to the subtable. Because Occupation
does not contribute to the subtable, PROC TABULATE looks at the next element in the
denominator denition, which is All. Because the variable All does contribute to this
subtable, PROC TABULATE uses it as the denominator denition. All is a reserved
class variable with only one category. Therefore, this denominator denition tells PROC
TABULATE to use the frequency count for All as the denominator.
For example, the denominator for the category all, female is the frequency count
for that category, 61.
Note: In these table cells, because the numerator and denominator are the same,
the column percentages in this subtable are all 100. 4

1278

A Closer Look

Chapter 47

Subtable 3: Occupation and All

PROC TABULATE looks at the rst element in the denominator denition,

Occupation, and asks if Occupation contributes to the subtable. Because Occupation
does contribute to the subtable, PROC TABULATE uses it as the denominator
denition. This denominator denition tells PROC TABULATE to sum the frequency
counts for all occurrences of Occupation in the subtable.
For example, the denominator for the category technical, all is the sum of the
frequency counts for technical, all; manager/supervisor, all; clerical, all;
and administrative, all. The corresponding frequency counts are 34, 35, 28, and 26.
Therefore, the denominator for this category is 34+35+28+26, or 123.

Subtable 4: All and All

PROC TABULATE looks at the rst element in the denominator denition,

Occupation, and asks if Occupation contributes to the subtable. Because Occupation
does not contribute to the subtable, PROC TABULATE looks at the next element in the
denominator denition, which is All. Because the variable All does contribute to this
subtable, PROC TABULATE uses it as the denominator denition. All is a reserved
class variable with only one category. Therefore, this denominator denition tells PROC
TABULATE to use the frequency count of All as the denominator.
There is only one category in this subtable: all, all. The frequency count for this
category is 123.
Note: In this calculation, because the numerator and denominator are the same, the
column percentage in this subtable is 100. 4
Total Percentages

The part of the TABLE statement that calculates the total percentages and labels the
row is
pctn=Total percent

If you do not specify a denominator denition, then PROC TABULATE obtains the
denominator for a cell by totaling all the frequency counts in the subtable. Table 47.9
on page 1279 summarizes the process for all subtables in this example.

The TABULATE Procedure

Table 47.9

Program

1279

Denominators for Total Percentages

Class variables contributing to

the subtable

Frequency counts

Total

Occupat and Gender

16, 18, 20, 15 14, 14, 11, 15

123

Occupat and All

34, 35, 28, 26

123

Gender and All

61, 62

123

All and All

123

Consequently, the denominator for total percentages is always 123.

Example 14: Specifying Style Elements for ODS Output

Procedure features:

STYLE= option in
PROC TABULATE statement
CLASSLEV statement
KEYWORD statement
TABLE statement
VAR statement
Other features:

ODS HTML statement

ODS PDF statement
ODS RTF statement
Data set: ENERGYENERGY on page 1387
Formats: REGFMT, DIVFMT, and USETYPE. on page 1233

This example creates HTML, RTF, and PDF les and species style elements for
various table regions.

Program
Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= and PAGESIZE= are
not set for this example because they have no effect on HTML, RTF, and Printer output.
options nodate pageno=1;

Specify the ODS output lenames. By opening multiple ODS destinations, you can produce
multiple output les in a single execution. The ODS HTML statement produces output that is
written in HTML. The ODS PDF statement produces output in Portable Document Format
(PDF). The ODS RTF statement produces output in Rich Text Format (RTF). The output from
PROC TABULATE goes to each of these les.
ods html body=external-HTML-file;
ods pdf file=external-PDF-file;

1280

Program

Chapter 47

ods rtf file=external-RTF-file;

Specify the table options. The STYLE= option in the PROC TABULATE statement species
the style element for the data cells of the table.
proc tabulate data=energy style=[font_weight=bold];

Specify subgroups for the analysis. The STYLE= option in the CLASS statement species
the style element for the class variable name headings.
class region division type / style=[just=center];

Specify the style attributes for the class variable value headings. The STYLE= option in
the CLASSLEV statement species the style element for the class variable level value headings.
classlev region division type / style=[just=left];

Specify the analysis variable and its style attributes. The STYLE= option in the VAR
statement species a style element for the variable name headings.
var expenditures / style=[font_size=3];

Specify the style attributes for keywords, and label the all keyword. The STYLE=
option in the KEYWORD statement species a style element for keywords. The KEYLABEL
statement assigns a label to the keyword.
keyword all sum / style=[font_width=wide];
keylabel all="Total";

Dene the table rows and columns and their style attributes. The STYLE= option in the
dimension expression overrides any other STYLE= specications in PROC TABULATE that
specify attributes for table cells. The STYLE= option after the slash (/) species attributes for
parts of the table other than table cells.
table (region all)*(division all*[style=[background=yellow]]),
(type all)*(expenditures*f=dollar10.) /
style=[bordercolor=blue]

Specify the style attributes for cells with missing values. The STYLE= option in the
MISSTEXT option of the TABLE statement species a style element to use for the text in table
cells that contain missing values.
misstext=[label="Missing" style=[font_weight=light]]

Specify the style attributes for the box above the row titles. The STYLE= option in the
BOX option of the TABLE statement species a style element to use for text in the box above
the row titles.
box=[label="Region by Division by Type"
style=[font_style=italic]];

Format the class variable values. The FORMAT statement assigns formats to Region,
Division, and Type.
format region regfmt. division divfmt. type usetype.;

The TABULATE Procedure

Specify the titles.

title Energy Expenditures;
title2 (millions of dollars);
run;

Close the ODS destinations.

ods html close;
ods pdf close;
ods rtf close;

HTML Output

1281

1282

PDF Output

Chapter 47

PDF Output

The TABULATE Procedure

References

1283

RTF Output

References
Jain, Raj and Chlamtac, Imrich (1985), The P2 Algorithm for Dynamic Calculation of
Quantiles and Histograms without Storing Observations, Communications of the
Association of Computing Machinery, 28:10.

1284

1285

CHAPTER

48
The TEMPLATE Procedure
Information about the TEMPLATE Procedure

1285

Information about the TEMPLATE Procedure

For complete documentation of the TEMPLATE procedure, see SAS Output
Delivery System: Users Guide.

See:

1286

1287

CHAPTER

49
The TIMEPLOT Procedure
Overview: TIMEPLOT Procedure 1287
Syntax: TIMEPLOT Procedure 1289
PROC TIMEPLOT Statement 1290
BY Statement 1291
CLASS Statement 1291
ID Statement 1292
PLOT Statement 1293
Results: TIMEPLOT Procedure 1297
Data Considerations 1297
Procedure Output 1297
Page Layout 1297
Contents of the Listing 1298
ODS Table Names 1298
Missing Values 1298
Examples: TIMEPLOT Procedure 1299
Example 1: Plotting a Single Variable 1299
Example 2: Customizing an Axis and a Plotting Symbol 1301
Example 3: Using a Variable for a Plotting Symbol 1303
Example 4: Superimposing Two Plots 1306
Example 5: Showing Multiple Observations on One Line of a Plot

1308

Overview: TIMEPLOT Procedure

The TIMEPLOT procedure plots one or more variables over time intervals. A listing
of variable values accompanies the plot. Although the plot and the listing are similar to
those produced by the PLOT and PRINT procedures, PROC TIMEPLOT output has
these distinctive features:
3 The vertical axis always represents the sequence of observations in the data set;
thus, if the observations are in order of date or time, then the vertical axis
represents the passage of time.
3 The horizontal axis represents the values of the variable that you are examining.
Like PROC PLOT, PROC TIMEPLOT can overlay multiple plots on one set of axes
so that each line of the plot can contain values for more than one variable.
3 A plot produced by PROC TIMEPLOT may occupy more than one page.
3 Each observation appears sequentially on a separate line of the plot; PROC
TIMEPLOT does not hide observations as PROC PLOT sometimes does.
3 The listing of the plotted values may include variables that do not appear in the
plot.

1288

Overview: TIMEPLOT Procedure

Chapter 49

Output 49.1 illustrates a simple report that you can produce with PROC TIMEPLOT.
This report shows sales of refrigerators for two sales representatives during the rst six
weeks of the year. The statements that produce the output follow. A DATA
stepExample 1 on page 1299 creates the data set SALES.
options linesize=64 pagesize=60 nodate
pageno=1;

proc timeplot data=sales;

plot icebox;
id month week;
title Weekly Sales of Refrigerators;
title2 for the;
title3 First Six Weeks of the Year;
run;

Output 49.1

Simple Report Created with PROC TIMEPLOT

Weekly Sales of Refrigerators
for the
First Six Weeks of the Year

Month

Week

Icebox

1
1
1
1
1
1
1
1
2
2
2
2

1
1
2
2
3
3
4
4
1
1
2
2

3450.94
2520.04
3240.67
2675.42
3160.45
2805.35
3400.24
2870.61
3550.43
2730.09
3385.74
2670.93

min
max
2520.04
3550.43
*-------------------------------*
|
I |
|I
|
|
I
|
|
I
|
|
I
|
|
I
|
|
I
|
|
I
|
|
I|
|
I
|
|
I
|
|
I
|
*-------------------------------*

Output 49.2 is a more complicated report of the same data set that is used to create
Output 49.1. The statements that create this report
3 create one plot for the sale of refrigerators and one for the sale of stoves

3
3
3
3

plot sales for both sales representatives on the same line

identify points on the plots by the rst letter of the sales representatives last name
control the size of the horizontal axis
control formats and labels.

For an explanation of the program that produces this report, see Example 5 on page
1308.

The TIMEPLOT Procedure

Output 49.2

Syntax: TIMEPLOT Procedure

1289

More Complex Report Created with PROC TIMEPLOT

Weekly Appliance Sales for the First Quarter

Month

Week

January
January
January
January
February
February

1
2
3
4
1
2

Seller :Kreitz
Stove

$1,312.61
$222.35
$2,263.33
$1,787.45
$2,910.37
$819.69

Seller :LeGrange
Stove

$728.13
$184.24
$267.35
$274.51
$397.98
$2,242.24

min
max
$184.24
$2,910.37
*-------------------------*
|
L
K
|
|!
|
| L
K
|
| L
K
|
| L
K|
|
K
L
|
*-------------------------*

Weekly Appliance Sales for the First Quarter

Month

January
January
January
January
February
February

1
2
3
4
1
2

Kreitz
Icebox

Week

$3,450.94
$3,240.67
$3,160.45
$3,400.24
$3,550.43
$3,385.74

LeGrange
Icebox

$2,520.04
$2,675.42
$2,805.35
$2,870.61
$2,730.09
$2,670.93

min
max
$2,520.04
$3,550.43
*-------------------------*
|L
K |
|
L
K
|
|
L
K
|
|
L
K
|
|
L
K|
|
L
K
|
*-------------------------*

Syntax: TIMEPLOT Procedure

Requirements:

At least one PLOT statement

Supports the Output Delivery System. See Output Delivery System on page 32
for details.

Tip:

ODS Table Names: See: ODS Table Names on page 1298

Reminder: You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See

Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.

PROC TIMEPLOT <option(s)>;

BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
CLASS variable(s);
ID variable(s);
PLOT plot-request(s)/option(s);

1290

PROC TIMEPLOT Statement

Chapter 49

To do this

Use this statement

Produce a separate plot for each BY group

Group data according to the values of the class

variables

CLASS

Print in the listing the values of the variables that

you identify

Specify the plots to produce

PLOT

PROC TIMEPLOT Statement

PROC TIMEPLOT <option(s)>;

Options
DATA=SAS-data-set

identies the input data set.

MAXDEC=number

species the maximum number of decimal places to print in the listing.

Interaction: A decimal specication in a format overrides a MAXDEC=

specication.
Default: 2

0-12

Range:

Featured in:

Example 4 on page 1306

SPLIT=split-character

species a split character, which controls line breaks in column headings. It also
species that labels be used as column headings. PROC TIMEPLOT breaks a column
heading when it reaches the split character and continues the heading on the next
line. Unless the split character is a blank, it is not part of the column heading. Each
occurrence of the split character counts toward the 256-character maximum for a
label.
Alias:

Default: blank ( )

Note: Column headings can occupy up to three lines. If the column label can be
split into more lines than this xed number, then the split character is used only as a
recommendation on how to split the label. 4
UNIFORM

uniformly scales the horizontal axis across all BY groups. By default, PROC
TIMEPLOT separately determines the scale of the axis for each BY group.
Interaction: UNIFORM also affects the calculation of means for reference lines (see

REF= on page 1296).

The TIMEPLOT Procedure

CLASS Statement

1291

BY Statement
Produces a separate plot for each BY group.
Main discussion:

BY on page 58

BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;

Required Arguments
variable

species the variable that the procedure uses to form BY groups. You can specify
more than one variable. If you do not use the NOTSORTED option in the BY
statement, then either the observations in the data set must be sorted by all the
variables that you specify, or they must be indexed appropriately. These variables are
called BY variables.

Options
DESCENDING

species that the data set is sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data is grouped in another way, for example, chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations that have the same values for the BY
variables are not contiguous, then the procedure treats each contiguous set as a
separate BY group.

CLASS Statement
Groups data according to the values of the class variables.
PROC TIMEPLOT uses the formatted values of the CLASS variables to form
classes. Thus, if a format groups the values, then the procedure uses those groups.

Tip:

Featured in:

Example 5 on page 1308

CLASS variable(s);

1292

ID Statement

Chapter 49

Required Arguments
variable(s)

species one or more variables that the procedure uses to group the data. Variables
in a CLASS statement are called class variables. Class variables can be numeric or
character. Class variables can have continuous values, but they typically have a few
discrete values that dene the classications of the variable. You do not have to sort
the data by class variables.
The values of the class variables appear in the listing. PROC TIMEPLOT prints
and plots one line each time the combination of values of the class variables changes.
Therefore, the output typically is more meaningful if you sort or group the data
according to values of the class variables.

Using Multiple CLASS Statements

You can use any number of CLASS statements. If you use more than one CLASS
statement, then PROC TIMEPLOT simply concatenates all variables from all of the
CLASS statements. The following form of the CLASS statement includes three
variables:
CLASS variable-1 variable-2 variable-3;
It has the same effect as this form:
CLASS variable-1;
CLASS variable-2;
CLASS variable-3;

Using a Symbol Variable

Normally, you use the CLASS statement with a symbol variable (see the discussion of
plot requests on page 1294). In this case, the listing of the plot variable contains a
column for each value of the symbol variable, and each row of the plot contains a point
for each value of the symbol variable. The plotting symbol is the rst character of the
formatted value of the symbol variable. If more than one observation within a class has
the same value of a symbol variable, then PROC TIMEPLOT plots and prints only the
rst occurrence of that value and writes a warning message to the SAS log.

ID Statement
Prints in the listing the values of the variables that you identify.
Featured in:

Example 1 on page 1299

ID variable(s);

The TIMEPLOT Procedure

PLOT Statement

Required Arguments
variable(s)

identies one or more ID variables to print in the listing.

PLOT Statement
Species the plots to produce.
Tip:

Each PLOT statement produces a separate plot.

PLOT plot-request(s)/option(s);
Table 49.1 on page 1293 summarizes the options that are available in the PLOT
statement.
Table 49.1

Summary of Options for the PLOT Statement

To do this

Use this option

Customize the axis

Specify the range of values to plot on the horizontal axis, as well as the
interval represented by each print position on the horizontal axis

AXIS=

Order the values on the horizontal axis with the largest value in the
leftmost position

REVERSE

Control the appearance of the plot

Connect the leftmost plotting symbol to the rightmost plotting symbol with a
line of hyphens (-)

HILOC

Connect the leftmost and rightmost symbols on each line of the plot with a
line of hyphens (-) regardless of whether the symbols are reference symbols
or plotting symbols

JOINREF

Suppress the name of the symbol variable in column headings when you use
a CLASS statement

NOSYMNAME

Suppress the listing of the values of the variables that appear in the PLOT
statement

NPP

Specify the number of print positions to use for the horizontal axis

POS=

Create and customize a reference line

Draw lines on the plot that are perpendicular to the specied values on the
horizontal axis

REF=

Specify the character for drawing reference lines

REFCHAR=

Display multiple plots on the same set of axes

Plot all requests in one PLOT statement on one set of axes

OVERLAY

Specify the character to print if multiple plotting symbols coincide

OVPCHAR=

1293

1294

PLOT Statement

Chapter 49

Required Arguments
plot-request(s)

species the variable or variables to plot and, optionally, the plotting symbol to use.
By default, each plot request produces a separate plot.
A plot request can have the following forms. You can mix different forms of
requests in one PLOT statement (see Example 4 on page 1306).
variable(s)
identies one or more numeric variables to plot. PROC TIMEPLOT uses the rst
character of the variable name as the plotting symbol.
Featured in: Example 1 on page 1299
(variable(s))=plotting-symbol
identies one or more numeric variables to plot and species the plotting symbol
to use for all variables in the list. You can omit the parentheses if you use only one
variable.
Featured in: Example 2 on page 1301
(variable(s))=symbol-variable
identies one or more numeric variables to plot and species a symbol variable.
PROC TIMEPLOT uses the rst nonblank character of the formatted value of the
symbol variable as the plotting symbol for all variables in the list. The plotting
symbol changes from one observation to the next if the value of the symbol
variable changes. You can omit the parentheses if you use only one variable.
Featured in: Example 3 on page 1303

Options
AXIS=axis-specication

species the range of values to plot on the horizontal axis, as well as the interval
represented by each print position on the axis. PROC TIMEPLOT labels the rst and
last ends of the axis, if space permits.
3 For numeric values, axis-specication can be one of the following or a
combination of both:
n< . . .n>
n TO n <BY increment>
The values must be in either ascending or descending order. Use a negative
value for increment to specify descending order. The specied values are spaced
evenly along the horizontal axis even if the values are not uniformly
distributed. Numeric values can be specied in the following ways:
Specication

Comments

axis=1 2 10

Values are 1, 2, and 10.

axis=10 to 100 by 5

Values appear in increments of 5,

starting at 10 and ending at 100.

axis=12 10 to 100 by 5

A combination of the two previous

forms of specication.

The TIMEPLOT Procedure

PLOT Statement

1295

3 For axis variables that contain datetime values, axis-specication is either an

explicit list of values or a starting and an ending value with an increment
specied:
date-time-valuei <. . . date-time-valuei>
date-time-valuei TO date-time-valuei
<BY increment>
date-time-valuei
any SAS date, time, or datetime value described for the SAS functions INTCK
and INTNX. The sufx i is one of the following:
D

date

time

datetime

increment
one of the valid arguments for the INTCK or INTNX functions. For dates,
increment can be one of the following:
DAY
WEEK
MONTH
QTR
YEAR
For datetimes, increment can be one of the following:
DTDAY
DTWEEK
DTMONTH
DTQTR
DTYEAR
For times, increment can be one of the following:
HOUR
MINUTE
SECOND
For example,
axis=01JAN95d to 01JAN96d by month
axis=01JAN95d to 01JAN96d by qtr

For descriptions of individual intervals, see the chapter on dates, times, and
intervals in SAS Language Reference: Concepts.
Note: You must use a FORMAT statement to print the tick-mark values in
an understandable form. 4

1296

PLOT Statement

Chapter 49

Interaction: The value of POS= (see POS= on page 1296) overrides an interval set

with AXIS=.
Tip: If the range that you specify does not include all your data, then PROC
TIMEPLOT uses angle brackets (< or >) on the left or right border of the plot to
indicate a value that is outside the range.
Featured in: Example 2 on page 1301
HILOC

connects the leftmost plotting symbol to the rightmost plotting symbol with a line of
hyphens (-).
Interactions: If you specify JOINREF, then PROC TIMEPLOT ignores HILOC.
JOINREF

connects the leftmost and rightmost symbols on each line of the plot with a line of
hyphens (-), regardless of whether the symbols are reference symbols or plotting
symbols. However, if a line contains only reference symbols, then PROC TIMEPLOT
does not connect the symbols.
Featured in: Example 3 on page 1303
NOSYMNAME

suppresses the name of the symbol variable in column headings when you use a
CLASS statement. If you use NOSYMNAME, then only the value of the symbol
variable appears in the column heading.
Featured in: Example 5 on page 1308
NPP

suppresses the listing of the values of the variables that appear in the PLOT
statement.
Featured in: Example 3 on page 1303
OVERLAY

plots all requests in one PLOT statement on one set of axes. Otherwise, PROC
TIMEPLOT produces a separate plot for each plot request.
Featured in: Example 4 on page 1306
OVPCHAR=character

species the character to print if multiple plotting symbols coincide. If a plotting

symbol and a character in a reference line coincide, then PROC TIMEPLOT prints
the plotting symbol.
Default: at sign (@)
Featured in: Example 5 on page 1308
POS=print-positions-for-plot

species the number of print positions to use for the horizontal axis.
Default: If you omit both POS= and AXIS=, then PROC TIMEPLOT initially
assumes that POS=20. However, if space permits, then this value increases so
that the plot lls the available space.
Interaction: If you specify POS=0 and AXIS=, then the plot lls the available
space. POS= overrides an interval set with AXIS= (see the discussion of AXIS= on
page 1294).
See also: Page Layout on page 1297
Featured in: Example 1 on page 1299
REF=reference-value(s)

draws lines on the plot that are perpendicular to the specied values on the horizontal
axis. The values for reference-value(s) may be constants, or you may use the form

The TIMEPLOT Procedure

Procedure Output

1297

MEAN(variable(s))
If you use this form of REF=, then PROC TIMEPLOT evaluates the mean for each
variable that you list and draws a reference line for each mean.
Interaction: If you use the UNIFORM option in the PROC TIMEPLOT statement,

then the procedure calculates the mean values for the variables over all
observations for all BY groups. If you do not use UNIFORM, then the procedure
calculates the mean for each variable for each BY group.
Interaction: If a plotting symbol and a reference character coincide, then PROC

TIMEPLOT prints the plotting symbol.

Featured in:

Example 3 on page 1303 and Example 4 on page 1306

REFCHAR=character

species the character for drawing reference lines.

Default: vertical bar (|)
Interaction: If you are using the JOINREF or HILOC option, then do not specify a

value for REFCHAR= that is the same as a plotting symbol, because PROC
TIMEPLOT will interpret the plotting symbols as reference characters and will
not connect the symbols as you expect.
Featured in:

Example 3 on page 1303

REVERSE

orders the values on the horizontal axis with the largest value in the leftmost
position.
Featured in:

Example 4 on page 1306

Results: TIMEPLOT Procedure

Data Considerations
The input data set usually contains a date variable to use as either a class or an ID
variable. Although PROC TIMEPLOT does not require an input data set sorted by
date, the output is usually more meaningful if the observations are in chronological
order. In addition, if you use a CLASS statement, then the output is more meaningful if
the input data set groups observations according to combinations of class variable
values. (For more information see CLASS Statement on page 1291.)

Procedure Output
Page Layout
For each plot request, PROC TIMEPLOT prints a listing and a plot. PROC
TIMEPLOT determines the arrangement of the page as follows:

3 If you use POS=, then the procedure

3 determines the size of the plot from the POS= value
3 determines the space for the listing from the width of the columns of printed
values, equally spaced and with a maximum of ve positions between columns

1298

ODS Table Names

Chapter 49

3 centers the output on the page.

3 If you omit POS=, then the procedure
3 determines the width of the plot from the value of the AXIS= option
3 expands the listing to ll the rest of the page.
If there is not enough space to print the listing and the plot for a particular plot
request, then PROC TIMEPLOT produces no output and writes the following error
message to the SAS log:
ERROR:

Too many variables/symbol values

to print.

The error does not affect other plot requests.

Contents of the Listing

The listing in the output contains different information depending on whether or not
you use a CLASS statement. If you do not use a CLASS statement (see Example 1 on
page 1299), then PROC TIMEPLOT prints (and plots) each observation on a separate
line. If you do use a CLASS statement, then the form of the output varies depending on
whether or not you specify a symbol variable (see Using a Symbol Variable on page
1292).

ODS Table Names

The TIMEPLOT procedure assigns a name to each table that it creates. You can use
these names to reference the table when using the Output Delivery System (ODS) to
select tables and create output data sets. For more information, see SAS Output
Delivery System: Users Guide.
Table 49.2

ODS Tables Produced by the TIMEPLOT Procedure

Table Name

Description

The TIMEPLOT procedure

generates the table:

Plot

A single plot

if you do not specify the

OVERLAY option

OverlaidPlot

Two or more plots on a single

set of axes

if you specify the OVERLAY

option

Missing Values
Four types of variables can appear in the listing from PROC TIMEPLOT: plot
variables, ID variables, class variables, and symbol variables (as part of some column
headers). Plot variables and symbol variables can also appear in the plot.
Observations with missing values of a class variable form a class of observations.
In the listing, missing values appear as a period (.), a blank, or a special missing
value (the letters A through Z and the underscore (_) character).
In the plot, PROC TIMEPLOT handles different variables in different ways:

3 An observation or class of observations with a missing value of the plot variable

does not appear in the plot.

The TIMEPLOT Procedure

Program

1299

3 If you use a symbol variable (see the discussion of plot requests on page 1294),
then PROC TIMEPLOT uses a period (.) as the symbol variable on the plot for all
observations that have a missing value for the symbol variable.

Examples: TIMEPLOT Procedure

Example 1: Plotting a Single Variable

Procedure features:

ID statement
PLOT statement arguments:
simple plot request
POS=

This example

3 uses a single PLOT statement to plot sales of refrigerators

3 species the number of print positions to use for the horizontal axis of the plot
3 provides context for the points in the plot by printing in the listing the values of
two variables that are not in the plot.

Program

Create the SALES data set. SALES contains weekly information on the sales of refrigerators
and stoves by two sales representatives.
data sales;
input Month Week Seller $ Icebox Stove;
datalines;
1 1 Kreitz
3450.94 1312.61
1 1 LeGrange 2520.04 728.13
1 2 Kreitz
3240.67 222.35
1 2 LeGrange 2675.42 184.24
1 3 Kreitz
3160.45 2263.33
1 3 LeGrange 2805.35 267.35
1 4 Kreitz
3400.24 1787.45
1 4 LeGrange 2870.61 274.51

1300

Output

Chapter 49

2
2
2
2
;

1
1
2
2

Kreitz
LeGrange
Kreitz
LeGrange

3550.43 2910.37
2730.09 397.98
3385.74 819.69
2670.93 2242.24

Plot sales of refrigerators. The plot variable, Icebox, appears in both the listing and the
output. POS= provides 50 print positions for the horizontal axis.
proc timeplot data=sales;
plot icebox / pos=50;

Label the rows in the listing. The values of the ID variables, Month and Week, are used to
uniquely identify each row of the listing.
id month week;

Specify the titles.

title Weekly Sales of Iceboxes;
title2 for the;
title3 First Six Weeks of the Year;
run;

title Weekly Sales of Refrigerators;
title2 for the;
title3 First Six Weeks of the Year;
run;

The TIMEPLOT Procedure

Example 3: Using a Variable for a Plotting Symbol

1303

Output

The column headers in the listing are the variables names (for Month and Week, which have no
labels) and the variables label (for Icebox, which has a label). The plotting symbol is R (for
Refrigerator).

Weekly Sales of Refrigerators

for the
First Six Weeks of the Year
Month

Week

Refrigerator

January
January
January
January
January
January
January
January
February
February
February
February

1
1
2
2
3
3
4
4
1
1
2
2

3450.94
2520.04
3240.67
2675.42
3160.45
2805.35
3400.24
2870.61
3550.43
2730.09
3385.74
2670.93

min
max
2500
3600
*---------------------------------------------*
|
R
|
| R
|
|
R
|
|
R
|
|
R
|
|
R
|
|
R
|
|
R
|
|
R |
|
R
|
|
R
|
|
R
|
*---------------------------------------------*

Example 3: Using a Variable for a Plotting Symbol

Procedure features:

ID statement
PLOT statement arguments:
using a variable as the plotting symbol
JOINREF
NPP
REF=
REFCHAR=
Data set:

SALES on page 1299

Formats: MONTHFMT. on page 1302

This example

3 species a variable to use as the plotting symbol to distinguish between points for
each of two sales representatives

3 suppresses the printing of the values of the plot variable in the listing
3 draws a reference line to a specied value on the axis and species the character
to use to draw the line

3 connects the leftmost and rightmost symbols on each line of the plot.

1304

Program

Chapter 49

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
adds the SAS data library PROCLIB to the search path that is used to locate formats.
options nodate pageno=1 linesize=80 pagesize=60
fmtsearch=(proclib);

Plot sales of stoves. The PLOT statement species both the plotting variable, Stove, and a
symbol variable, Seller. The plotting symbol is the rst letter of the formatted value of the Seller
(in this case, L or K).
proc timeplot data=sales;
plot stove=seller /

Suppress the appearance of the plotting variable in the listing. The values of the Stove
variable will not appear in the listing.
npp

Create a reference line on the plot. REF= and REFCHAR= draw a line of colons at the sales
target of $1500.
ref=1500 refchar=:

Draw a line between the symbols on each line of the plot. In this plot, JOINREF connects
each plotting symbol to the reference line.
joinref

Customize the horizontal axis. AXIS= sets the minimum value of the horizontal axis to 100
and the maximum value to 3000. BY 50 species that each print position on the axis represents
50 units (in this case, dollars).
axis=100 to 3000 by 50;

The TIMEPLOT Procedure

Output

1305

Label the rows in the listing. The values of the ID variables, Month and Week, are used to
identify each row of the listing.
id month week;

Apply the MONTHFMT. format to the Month variable. The FORMAT statement assigns a
format to use for Month in the report.
format month monthfmt.;

Specify the titles.

title Weekly Sales of Stoves;
title2 Compared to Target Sales of $1500;
title3 K for Kreitz; L for LeGrange;
run;

Output

The plot uses the rst letter of the value of Seller as the plotting symbol.

Weekly Sales of Stoves

Compared to Target Sales of $1500
K for Kreitz; L for LeGrange
Month

Week

January
January
January
January
January
January
January
January
February
February
February
February

1
1
2
2
3
3
4
4
1
1
2
2

min
max
100
3000
*-----------------------------------------------------------*
|
K---:
|
|
L--------------:
|
| K-------------------------:
|
| L-------------------------:
|
|
:--------------K
|
|
L------------------------:
|
|
:-----K
|
|
L------------------------:
|
|
:---------------------------K |
|
L---------------------:
|
|
K-------------:
|
|
:--------------L
|
*-----------------------------------------------------------*

1306

Example 4: Superimposing Two Plots

Chapter 49

Example 4: Superimposing Two Plots

Procedure features:

PROC TIMEPLOT statement options:

MAXDEC=
PLOT statement arguments:
using two types of plot requests
OVERLAY
REF=MEAN(variable(s))
REVERSE
Data set:

SALES on page 1299

This example
3 superimposes two plots on one set of axes
3 species a variable to use as the plotting symbol for one plot and a character to
use as the plotting symbol for the other plot
3 draws a reference line to the mean value of each of the two variables plotted
3 reverses the labeling of the axis so that the largest value is at the far left of the
plot.

Program

Specify the number of decimal places to display. MAXDEC= species the number of
decimal places to display in the listing.
proc timeplot data=sales maxdec=0;

Plot sales of both stoves and refrigerators.The PLOT statement requests two plots. One
plot uses the rst letter of the formatted value of Seller to plot the values of Stove. The other
uses the letter R (to match the label Refrigerators) to plot the value of Icebox.
plot stove=seller icebox=R /

Print both plots on the same set of axes.

overlay

The TIMEPLOT Procedure

Output

1307

Create two reference lines on the plot. REF= draws two reference lines: one perpendicular
to the mean of Stove, the other perpendicular to the mean of Icebox.
ref=mean(stove icebox)

Order the values on the horizontal axis from largest to smallest.

reverse;

Specify the titles.

title Weekly Sales of Stoves and Refrigerators;
title2 for the;
title3 First Six Weeks of the Year;
run;

Output

The column header for the variable Icebox in the listing is the variables label (Refrigerators).
One plot uses the rst letter of the value of Seller as the plotting symbol. The other plot uses
the letter R.

Weekly Sales of Stoves and Refrigerators

for the
First Six Weeks of the Year
Stove

Refrigerators

1313
728
222
184
2263
267
1787
275
2910
398
820
2242

3451
2520
3241
2675
3160
2805
3400
2871
3550
2730
3386
2671

max
min
3550.43
184.24
*--------------------------------------------------*
|R
|
K |
|
|
|
R
|
L
|
|
R |
|
K |
|
|
R
|
L|
|
R |
K
|
|
|
| R
|
L |
| R
|
K
|
|
|
| R
|
L |
|R
| K
|
|
|
|
R
|
L
|
| R
|
|
K
|
|
|
R
L
|
|
*--------------------------------------------------*

1308

Example 5: Showing Multiple Observations on One Line of a Plot

Chapter 49

Example 5: Showing Multiple Observations on One Line of a Plot

Procedure features:

CLASS statement
PLOT statement arguments:
creating multiple plots
NOSYMNAME
OVPCHAR=
SALES on page 1299

Data set:
Formats:

MONTHFMT. on page 1302

This example

3 groups observations for the same month and week so that sales for the two sales
representatives for the same week appear on the same line of the plot

3 species a variable to use as the plotting symbol

3 suppresses the name of the plotting variable from one plot
3 species a size for the plots so that they both occupy the same amount of space.

Program

Declare the PROCLIB SAS data library.

libname proclib SAS-data-library;

Set the SAS system options. The NODATE option suppresses the display of the date and time
in the output. PAGENO= species the starting page number. LINESIZE= species the output
line length, and PAGESIZE= species the number of lines on an output page. FMTSEARCH=
adds the SAS data library PROCLIB to the search path that is used to locate formats.
options nodate pageno=1 linesize=80 pagesize=60
fmtsearch=(proclib);

Specify subgroups for the analysis. The CLASS statement groups all observations with the
same values of Month and Week into one line in the output. Using the CLASS statement with a
symbol variable produces in the listing one column of the plot variable for each value of the
symbol variable.
proc timeplot data=sales;
class month week;

The TIMEPLOT Procedure

Output

1309

Plot sales of stoves and refrigerators. Each PLOT statement produces a separate plot. The
plotting symbol is the rst character of the formatted value of the symbol variable: K for Kreitz;
L for LeGrange. POS= species that each plot uses 25 print positions for the horizontal axis.
OVPCHAR= designates the exclamation point as the plotting symbol when the plotting symbols
coincide. NOSYMNAME suppresses the name of the symbol variable Seller from the second
listing.
plot stove=seller / pos=25 ovpchar=!;
plot icebox=seller / pos=25 ovpchar=! nosymname;

Apply formats to values in the listing. The FORMAT statement assigns formats to use for
Stove, Icebox, and Month in the report. The TITLE statement species a title.
format stove icebox dollar10.2 month monthfmt.;

Specify the title.

title Weekly Appliance Sales for the First Quarter;
run;

Output

Weekly Appliance Sales for the First Quarter

Month

Week

January
January
January
January
February
February

1
2
3
4
1
2

Seller :Kreitz
Stove

$1,312.61
$222.35
$2,263.33
$1,787.45
$2,910.37
$819.69

Seller :LeGrange
Stove

$728.13
$184.24
$267.35
$274.51
$397.98
$2,242.24

min
max
$184.24
$2,910.37
*-------------------------*
|
L
K
|
|!
|
| L
K
|
| L
K
|
| L
K|
|
K
L
|
*-------------------------*

Weekly Appliance Sales for the First Quarter

Month

Week

January
January
January
January
February
February

1
2
3
4
1
2

Kreitz
Icebox

$3,450.94
$3,240.67
$3,160.45
$3,400.24
$3,550.43
$3,385.74

LeGrange
Icebox

$2,520.04
$2,675.42
$2,805.35
$2,870.61
$2,730.09
$2,670.93

min
max
$2,520.04
$3,550.43
*-------------------------*
|L
K |
|
L
K
|
|
L
K
|
|
L
K
|
|
L
K|
|
L
K
|
*-------------------------*

1310

1311

CHAPTER

50
The TRANSPOSE Procedure
Overview: TRANSPOSE Procedure 1311
What Does the TRANSPOSE Procedure Do? 1311
What Types of Transpositions Can PROC TRANSPOSE Perform? 1312
Syntax: TRANSPOSE Procedure 1314
PROC TRANSPOSE Statement 1314
BY Statement 1315
COPY Statement 1317
ID Statement 1318
IDLABEL Statement 1319
VAR Statement 1319
Results: TRANSPOSE Procedure 1320
Output Data Set 1320
Output Data Set Variables 1320
Attributes of Transposed Variables 1321
Names of Transposed Variables 1321
Examples: TRANSPOSE Procedure 1321
Example 1: Performing a Simple Transposition 1321
Example 2: Naming Transposed Variables 1323
Example 3: Labeling Transposed Variables 1324
Example 4: Transposing BY Groups 1325
Example 5: Naming Transposed Variables When the ID Variable Has Duplicate Values
Example 6: Transposing Data for Statistical Analysis 1329

1328

Overview: TRANSPOSE Procedure

What Does the TRANSPOSE Procedure Do?
The TRANSPOSE procedure creates an output data set by restructuring the values
in a SAS data set, transposing selected variables into observations. The TRANSPOSE
procedure can often eliminate the need to write a lengthy DATA step to achieve the
same result. Further, the output data set can be used in subsequent DATA or PROC
steps for analysis, reporting, or further data manipulation.
PROC TRANSPOSE does not produce printed output. To print the output data set
from the PROC TRANSPOSE step, use PROC PRINT, PROC REPORT, or another SAS
reporting tool.

1312

What Types of Transpositions Can PROC TRANSPOSE Perform?

Chapter 50

A transposed variable is a variable that the procedure creates by transposing the

values of an observation in the input data set into values of a variable in the output
data set.

What Types of Transpositions Can PROC TRANSPOSE Perform?

Simple Transposition
The following example illustrates a simple transposition. In the input data set, each
variable represents the scores from one tester. In the output data set, each observation
now represents the scores from one tester. Each value of _NAME_ is the name of a
variable in the input data set that the procedure transposed. Thus, the value of
_NAME_ identies the source of each observation in the output data set. For example,
the values in the rst observation in the output data set come from the values of the
variable Tester1 in the input data set. The statements that produce the output follow.
proc print data=proclib.product noobs;
title The Input Data Set;
run;
proc transpose data=proclib.product
out=proclib.product_transposed;
run;
proc print data=proclib.product_transposed noobs;
title The Output Data Set;
run;

Output 50.1

A Simple Transposition
The Input Data Set
Tester1

Tester2

22
15
17
20
14
15
10
22

Tester3

25
19
19
19
15
17
11
24

21
18
19
16
13
18
9
23

1
Tester4
21
17
19
19
13
19
10
21

The Output Data Set

_NAME_
Tester1
Tester2
Tester3
Tester4

COL1

COL2

COL3

COL4

COL5

COL6

COL7

COL8

22
25
21
21

15
19
18
17

17
19
19
19

20
19
16
19

14
15
13
13

15
17
18
19

10
11
9
10

22
24
23
21

The TRANSPOSE Procedure

What Types of Transpositions Can PROC TRANSPOSE Perform?

1313

Complex Transposition Using BY Groups

The next example, which uses BY groups, is more complex. The input data set
represents measurements of the weight and length of sh at two lakes. The statements
that create the output data set do the following:
3 transpose only the variables that contain the length measurements
3 create six BY groups, one for each lake and date
3 use a data set option to name the transposed variable.

Output 50.2

A Transposition with BY Groups

Input Data Set

L
o
c
a
t
i
o
n
Cole Pond
Cole Pond
Cole Pond
Eagle Lake
Eagle Lake
Eagle Lake

D
a
t
e

L
e
n
g
t
h
1

02JUN95
03JUL95
04AUG95
02JUN95
03JUL95
04AUG95

31
33
29
32
30
33

W
e
i
g
h
t
1
0.25
0.32
0.23
0.35
0.20
0.30

L
e
n
g
t
h
2
32
34
30
32
36
33

W
e
i
g
h
t
2
0.30
0.41
0.25
0.25
0.45
0.28

L
e
n
g
t
h
3

W
e
i
g
h
t
3

32
37
34
33
.
34

0.25
0.48
0.47
0.30
.
0.42

Fish Length Data for Each Location and Date

Location
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Cole Pond
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake
Eagle Lake

Date
02JUN95
02JUN95
02JUN95
02JUN95
03JUL95
03JUL95
03JUL95
03JUL95
04AUG95
04AUG95
04AUG95
04AUG95
02JUN95
02JUN95
02JUN95
02JUN95
03JUL95
03JUL95
03JUL95
03JUL95
04AUG95
04AUG95
04AUG95
04AUG95

_NAME_
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4

L
e
n
g
t
h
4
33
32
32
.
.
.

W
e
i
g
h
t
4
0.30
0.28
0.30
.
.
.

Measurement
31
32
32
33
33
34
37
32
29
30
34
32
32
32
33
.
30
36
.
.
33
33
34
.

For a complete explanation of the SAS program that produces these results, see
Example 4 on page 1325.

1314

Syntax: TRANSPOSE Procedure

Chapter 50

Syntax: TRANSPOSE Procedure

Tip:

Does not support the Output Delivery System

You can use the ATTRIB, FORMAT, LABEL, and WHERE statements. See
Chapter 3, Statements with the Same Function in Multiple Procedures, on page 57 for
details. You can also use any global statements. See Global Statements on page 18 for
a list.

Reminder:

PROC TRANSPOSE <DATA=input-data-set> <LABEL=label> <LET>

<NAME=name> <OUT=output-data-set> <PREFIX=prex>;
BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;
COPY variable(s);
ID variable;
IDLABEL variable;
VAR variable(s);

To do this

Use this statement

Transpose each BY group

Copy variables directly without transposing them

COPY

Specify a variable whose values name the transposed

variables

Create labels for the transposed variables

IDLABEL

List the variables to transpose

VAR

PROC TRANSPOSE Statement

You can use data set options with the DATA= and OUT= options. See Data
Set Options on page 18 for a list.

Reminder:

PROC TRANSPOSE <DATA=input-data-set> <LABEL=label> <LET>

<NAME=name> < OUT=output-data-set> < PREFIX=prex>;

Options
DATA= input-data-set

names the SAS data set to transpose.

Default: most recently created SAS data set

The TRANSPOSE Procedure

BY Statement

1315

LABEL= label

species a name for the variable in the output data set that contains the label of the
variable that is being transposed to create the current observation.
Default: _LABEL_
LET

allows duplicate values of an ID variable. PROC TRANSPOSE transposes the

observation that contains the last occurrence of a particular ID value within the data
set or BY group.
Featured in: Example 5 on page 1328
NAME= name

species the name for the variable in the output data set that contains the name of
the variable that is being transposed to create the current observation.
Default: _NAME_
Featured in:

Example 2 on page 1323

OUT= output-data-set

names the output data set. If output-data-set does not exist, then PROC
TRANSPOSE creates it by using the DATAn naming convention.
Default: DATAn
Featured in:

Example 1 on page 1321

PREFIX= prex

species a prex to use in constructing names for transposed variables in the output
data set. For example, if PREFIX=VAR, then the names of the variables are VAR1,
VAR2, ,VARn.
Interaction: when you use PREFIX= with an ID statement, the value prexes to

the ID value.
Featured in: Example 2 on page 1323

BY Statement
Denes BY groups.
Main discussion:
Featured in:

BY on page 58

Example 4 on page 1325

Restriction: You cannot use PROC TRANSPOSE with a BY statement or an ID

statement with an engine that supports concurrent access if another user is updating
the data set at the same time.

BY <DESCENDING> variable-1
<< DESCENDING> variable-n>
<NOTSORTED>;

1316

BY Statement

Chapter 50

Required Arguments
variable

species the variable that PROC TRANSPOSE uses to form BY groups. You can
specify more than one variable. If you do not use the NOTSORTED option in the BY
statement, then either the observations must be sorted by all the variables that you
specify, or they must be indexed appropriately. Variables in a BY statement are
called BY variables.

Options
DESCENDING

species that the data set is sorted in descending order by the variable that
immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data is grouped in another way, such as chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED. The
procedure denes a BY group as a set of contiguous observations that have the same
values for all BY variables. If observations with the same values for the BY variables
are not contiguous, then the procedure treats each contiguous set as a separate BY
group.

Transpositions with BY Groups

PROC TRANSPOSE does not transpose BY groups. Instead, for each BY group,
PROC TRANSPOSE creates one observation for each variable that it transposes.
Figure 50.1 on page 1317 shows what happens when you transpose a data set with
BY groups. TYPE is the BY variable, and SOLD, NOTSOLD, REPAIRED, and
JUNKED are the variables to transpose.

The TRANSPOSE Procedure

Figure 50.1

COPY Statement

1317

Transposition with BY Groups

TYPE

input
data set

MONTH

sedan
sedan
sports
sports
trucks
trucks

SOLD
26
28
16
19
29
35

jan
feb
jan
feb
jan
feb

TYPE

_NAME_

sedan
sedan
sedan
sedan
sports
sports
sports
sports
trucks
trucks
trucks
trucks

output
data set

SOLD
NOTSOLD
REPAIRED
JUNKED
SOLD
NOTSOLD
REPAIRED
JUNKED
SOLD
NOTSOLD
REPAIRED
JUNKED

NOTSOLD

JUNKED

41
48
15
20
20
22

6
9
6
7
1
3

REPAIRED

4
2
0
1
3
4

COL1

COL2

26
6
41
4
16
6
15
0
29
1
20
3

28
9
48
2
19
7
20
1
35
3
22
4

3 The number of observations in the output data set (12) is the number of BY groups
(3) multiplied by the number of variables that are transposed (4).

3 The BY variable is not transposed.

3 _NAME_ contains the name of the variable in the input data set that was
transposed to create the current observation in the output data set. You can use
the NAME= option to specify another name for the _NAME_ variable.

3 The maximum number of observations in any BY group in the input data set is
two; therefore, the output data set contains two variables, COL1 and COL2. COL1
and COL2 contain the values of SOLD, NOTSOLD, REPAIRED, and JUNKED.

Note: If a BY group in the input data set has more observations than other BY
groups, then PROC TRANSPOSE assigns missing values in the output data set to the
variables that have no corresponding input observations. 4

COPY Statement
Copies variables directly from the input data set to the output data set without transposing them.
Featured in:

Example 6 on page 1329

COPY variable(s);

1318

ID Statement

Chapter 50

Required Argument
variable(s)

names one or more variables that the COPY statement copies directly from the input
data set to the output data set without transposing them.

Details
Because the COPY statement copies variables directly to the output data set, the
number of observations in the output data set is equal to the number of observations in
the input data set.
The procedure pads the output data set with missing values if the number of
observations in the input data set is not equal to the number of variables that it
transposes.

ID Statement
Species a variable in the input data set whose formatted values name the transposed variables
in the output data set.
Example 2 on page 1323
Restriction: You cannot use PROC TRANSPOSE with an ID statement or a BY
statement with an engine that supports concurrent access if another user is updating
the data set at the same time.
Featured in:

ID variable;

Required Argument
variable

names the variable whose formatted values name the transposed variables.

Duplicate ID Values
Typically, each formatted ID value occurs only once in the input data set or, if you
use a BY statement, only once within a BY group. Duplicate values cause PROC
TRANSPOSE to issue a warning message and stop. However, if you use the LET option
in the PROC TRANSPOSE statement, then the procedure issues a warning message
about duplicate ID values and transposes the observation that contains the last
occurrence of the duplicate ID value.

Making Variable Names out of Numeric Values

When you use a numeric variable as an ID variable, PROC TRANSPOSE changes
the formatted ID value into a valid SAS name.
However, SAS variable names cannot begin with a number. Thus, when the rst
character of the formatted value is numeric, the procedure prexes an underscore to the
value, truncating the last character of a 32-character value. Any remaining invalid

The TRANSPOSE Procedure

VAR Statement

1319

characters are replaced by underscores. The procedure truncates to 32 characters any

ID value that is longer than 32 characters when it uses that value to name a
transposed variable.
If the formatted value looks like a numeric constant, then PROC TRANSPOSE
changes the characters +, , and . to P, N, and D, respectively. If the formatted
value has characters that are not numerics, then PROC TRANSPOSE changes the
characters +, , and . to underscores.
Note: If the value of the VALIDVARNAME system option is V6, then PROC
TRANSPOSE truncates transposed variable names to eight characters. 4

Missing Values
If you use an ID variable that contains a missing value, then PROC TRANSPOSE
writes an error message to the log. The procedure does not transpose observations that
have a missing value for the ID variable.

IDLABEL Statement
Creates labels for the transposed variables.
Restriction:
Featured in:

Must appear after an ID statement.

Example 3 on page 1324

IDLABEL variable;

Required Argument
variable

names the variable whose values the procedure uses to label the variables that the
ID statement names. variable can be character or numeric.
Note: To see the effect of the IDLABEL statement, print the output data set with
the PRINT procedure by using the LABEL option, or print the contents of the output
data set by using the CONTENTS statement in the DATASETS procedure. 4

VAR Statement
Lists the variables to transpose.
Featured in:

Example 4 on page 1325 and Example 6 on page 1329

VAR variable(s);

1320

Results: TRANSPOSE Procedure

Chapter 50

Required Argument
variable(s)

names one or more variables to transpose.

Details
3 If you omit the VAR statement, the then TRANSPOSE procedure transposes all
numeric variables in the input data set that are not listed in another statement.

3 You must list character variables in a VAR statement if you want to transpose
them.

Results: TRANSPOSE Procedure

Output Data Set

The TRANSPOSE procedure always produces an output data set, regardless of
whether you specify the OUT= option in the PROC TRANSPOSE statement. PROC
TRANSPOSE does not print the output data set. Use PROC PRINT, PROC REPORT, or
some other SAS reporting tool to print the output data set.

Output Data Set Variables

The output data set contains the following variables:

3 variables that result from transposing the values of each variable into an
observation.

3 a variable that PROC TRANSPOSE creates to identify the source of the values in
each observation in the output data set. This variable is a character variable
whose values are the names of the variables that are transposed from the input
data set. By default, PROC TRANSPOSE names this variable _NAME_. To
override the default name, use the NAME= option. The label for the _NAME_
variable is NAME OF FORMER VARIABLE.

3 variables that PROC TRANSPOSE copies from the input data set when you use
either the BY or COPY statement. These variables have the same names and
values as they do in the input data set.

3 a character variable whose values are the variable labels of the variables that are
being transposed (if any of the variables that the procedure is transposing have
labels). Specify the name of the variable by using the LABEL= option. The default
is _LABEL_.
Note: If the value of the LABEL= option or the NAME= option is the same as a
variable that appears in a BY or COPY statement, then the output data set does
not contain a variable whose values are the names or labels of the transposed
variables.

The TRANSPOSE Procedure

Example 1: Performing a Simple Transposition

1321

Attributes of Transposed Variables

3 All transposed variables are the same type and length.
3 If all variables that the procedure is transposing are numeric, then the transposed
variables are numeric. Thus, if the numeric variable has a character string as a
formatted value, then its unformatted numeric value is transposed.

3 If any variable that the procedure is transposing is character, then all transposed
variables are character. Thus, if you are transposing a numeric variable that has a
character string as a formatted value, then the formatted value is transposed.

3 The length of the transposed variables is equal to the length of the longest
variable that is being transposed.

Names of Transposed Variables

PROC TRANSPOSE names transposed variables by using the following rules:
1 An ID statement species a variable in the input data set whose formatted values
become names for the transposed variables.
2 The PREFIX= option species a prex to use in constructing the names of

transposed variables.
3 If you do not use an ID statement or the PREFIX= option, then PROC

TRANSPOSE looks for an input variable called _NAME_ from which to get the
names of the transposed variables.
4 If you do not use an ID statement or the PREFIX= option, and if the input data

set does not contain a variable named _NAME_, then PROC TRANSPOSE assigns
the names COL1, COL2, , COLn to the transposed variables.

Examples: TRANSPOSE Procedure

Example 1: Performing a Simple Transposition

Procedure features:

PROC TRANSPOSE statement option:

OUT=

This example performs a default transposition and uses no subordinate statements.

1322

Program

Chapter 50

Program

Create the SCORE data set.set SCORE contains students names, their identication
numbers, and their grades on two tests and a nal exam.
data score;
input Student $9. +1 StudentID $ Section $ Test1 Test2 Final;
datalines;
Capalleti 0545 1 94 91 87
Dubose
1252 2 51 65 91
Engles
1167 1 95 97 97
Grant
1230 2 63 75 80
Krupski
2527 2 80 76 71
Lundsford 4860 1 92 40 86
McBane
0674 1 75 78 72
;

Transpose the data set. PROC TRANSPOSE transposes only the numeric variables, Test1,
Test2, and Final, because no VAR statement appears and none of the numeric variables appear
in another statement. OUT= puts the result of the transposition in the data set
SCORE_TRANSPOSED.
proc transpose data=score out=score_transposed;
run;

Print the SCORE_TRANSPOSED data set. The NOOBS option suppresses the printing of
observation numbers.
proc print data=score_transposed noobs;
title Student Test Scores in Variables;
run;

The TRANSPOSE Procedure

Program

1323

Output

In the output data set SCORE_TRANSPOSED, the variables COL1 through COL7 contain the
individual scores for the students. Each observation contains all the scores for one test. The
variable _NAME_ contains the names of the variables from the input data set that were
transposed.

Student Test Scores in Variables

_NAME_
Test1
Test2
Final

COL1

COL2

COL3

COL4

COL5

COL6

COL7

94
91
87

51
65
91

95
97
97

63
75
80

80
76
71

92
40
86

75
78
72

Example 2: Naming Transposed Variables

Procedure features:

PROC TRANSPOSE statement options:

NAME=
PREFIX=
ID statement
Data set: SCORE on page 1322

This example uses the values of a variable and a user-supplied value to name
transposed variables.

Program

Transpose the data set. PROC TRANSPOSE transposes only the numeric variables, Test1,
Test2, and Final, because no VAR statement appears. OUT= puts the result of the transposition
in the IDNUMBER data set. NAME= species Test as the name for the variable that contains
the names of the variables in the input data set that the procedure transposes. The procedure
names the transposed variables by using the value from PREFIX=, sn, and the value of the ID
variable StudentID.
proc transpose data=score out=idnumber name=Test
prefix=sn;

1324

Output

Chapter 50

id studentid;
run;

Print the IDNUMBER data set. The NOOBS option suppresses the printing of observation
numbers.
proc print data=idnumber noobs;
title Student Test Scores;
run;

Output

This is the output data set, IDNUMBER.

Student Test Scores

Test
Test1
Test2
Final

sn0545

sn1252

sn1167

sn1230

sn2527

sn4860

sn0674

94
91
87

51
65
91

95
97
97

63
75
80

80
76
71

92
40
86

75
78
72

Example 3: Labeling Transposed Variables

Procedure features:

PROC TRANSPOSE statement option:

PREFIX=
IDLABEL statement
Data set: SCORE on page 1322

This example uses the values of the variable in the IDLABEL statement to label
transposed variables.

Program

The TRANSPOSE Procedure

Example 4: Transposing BY Groups

1325

Transpose the data set. PROC TRANSPOSE transposes only the numeric variables, Test1,
Test2, and Final, because no VAR statement appears. OUT= puts the result of the transposition
in the IDLABEL data set. NAME= species Test as the name for the variable that contains the
names of the variables in the input data set that the procedure transposes. The procedure
names the transposed variables by using the value from PREFIX=, sn, and the value of the ID
variable StudentID.
proc transpose data=score out=idlabel name=Test
prefix=sn;
id studentid;

Assign labels to the output variables. PROC TRANSPOSE uses the values of the variable
Student to label the transposed variables. The procedure provides the following as the label for
the _NAME_ variable:NAME OF FORMER VARIABLE
idlabel student;
run;

Print the IDLABEL data set. The LABEL option causes PROC PRINT to print variable labels
for column headers. The NOOBS option suppresses the printing of observation numbers.
proc print data=idlabel label noobs;
title Student Test Scores;
run;

Output

This is the output data set, IDLABEL.

Student Test Scores

NAME OF
FORMER
VARIABLE

Capalleti

Test1
Test2
Final

94
91
87

Dubose

Engles

51
65
91

95
97
97

Example 4: Transposing BY Groups

Procedure features:

BY statement
VAR statement

Grant
63
75
80

Krupski
80
76
71

Lundsford
92
40
86

McBane
75
78
72

1326

Program

Chapter 50

Other features: Data set option:

RENAME=

This example illustrates transposing BY groups and selecting variables to transpose.

Program

Create the FISHDATA data set. The data in FISHDATA represents length and weight
measurements of sh that were caught at two ponds on three separate days. The raw data is
sorted by Location and Date.
data fishdata;
infile datalines missover;
input Location & $10. Date date7.
Length1 Weight1 Length2 Weight2 Length3 Weight3
Length4 Weight4;
format date date7.;
datalines;
Cole Pond
2JUN95 31 .25 32 .3 32 .25 33 .3
Cole Pond
3JUL95 33 .32 34 .41 37 .48 32 .28
Cole Pond
4AUG95 29 .23 30 .25 34 .47 32 .3
Eagle Lake 2JUN95 32 .35 32 .25 33 .30
Eagle Lake 3JUL95 30 .20 36 .45
Eagle Lake 4AUG95 33 .30 33 .28 34 .42
;

Transpose the data set. OUT= puts the result of the transposition in the FISHLENGTH data
set. RENAME= renames COL1 in the output data set to Measurement.
proc transpose data=fishdata
out=fishlength(rename=(col1=Measurement));

Specify the variables to transpose. The VAR statement limits the variables that PROC
TRANSPOSE transposes.
var length1-length4;

The TRANSPOSE Procedure

Output

1327

Organize the output data set into BY groups. The BY statement creates BY groups for each
unique combination of values of Location and Date. The procedure does not transpose the BY
variables.
by location date;
run;

Print the FISHLENGTH data set. The NOOBS option suppresses the printing of observation
numbers.
proc print data=fishlength noobs;
title Fish Length Data for Each Location and Date;
run;

Output

This is the output data set, FISHLENGTH. For each BY group in the original data set, PROC
TRANSPOSE creates four observations, one for each variable that it is transposing. Missing
values appear for the variable Measurement (renamed from COL1) when the variables that are
being transposed have no value in the input data set for that BY group. Several observations
have a missing value for Measurement. For example, in the last observation, a missing value
appears because the input data contained no value for Length4 on 04AUG95 at Eagle Lake.

Fish Length Data for Each Location and Date

Date
02JUN95
02JUN95
02JUN95
02JUN95
03JUL95
03JUL95
03JUL95
03JUL95
04AUG95
04AUG95
04AUG95
04AUG95
02JUN95
02JUN95
02JUN95
02JUN95
03JUL95
03JUL95
03JUL95
03JUL95
04AUG95
04AUG95
04AUG95
04AUG95

_NAME_
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4
Length1
Length2
Length3
Length4

Measurement
31
32
32
33
33
34
37
32
29
30
34
32
32
32
33
.
30
36
.
.
33
33
34
.

1328

Example 5: Naming Transposed Variables When the ID Variable Has Duplicate Values

Chapter 50

Example 5: Naming Transposed Variables When the ID Variable Has

Duplicate Values
Procedure features:

PROC TRANSPOSE statement option:

LET

This example shows how to use values of a variable (ID) to name transposed
variables even when the ID variable has duplicate values.

Program

Create the STOCKS data set. STOCKS contains stock prices for two competing kite
manufacturers. The prices are recorded for two days, three times a day: at opening, at noon, and
at closing. Notice that the input data set contains duplicate values for the Date variable.
data stocks;
input Company $14. Date
datalines;
Horizon Kites jun11 opening
Horizon Kites jun11 noon
Horizon Kites jun11 closing
Horizon Kites jun12 opening
Horizon Kites jun12 noon
Horizon Kites jun12 closing
SkyHi Kites
jun11 opening
SkyHi Kites
jun11 noon
SkyHi Kites
jun11 closing
SkyHi Kites
jun12 opening
SkyHi Kites
jun12 noon
SkyHi Kites
jun12 closing
;

$ Time $ Price;
29
27
27
27
28
30
43
43
44
44
45
45

Transpose the data set. LET transposes only the last observation for each BY group. PROC
TRANSPOSE transposes only the Price variable. OUT= puts the result of the transposition in
the CLOSE data set.
proc transpose data=stocks out=close let;

The TRANSPOSE Procedure

Example 6: Transposing Data for Statistical Analysis

Organize the output data set into BY groups. The BY statement creates two BY groups,
one for each company.
by company;

Name the transposed variables. The values of Date are used as names for the transposed
variables.
id date;
run;

Print the CLOSE data set. The NOOBS option suppresses the printing of observation
numbers..
proc print data=close noobs;
title Closing Prices for Horizon Kites and SkyHi Kites;
run;

Output

This is the output data set, CLOSE.

Closing Prices for Horizon Kites and SkyHi Kites

Company
Horizon Kites
SkyHi Kites

_NAME_
Price
Price

jun11
27
44

jun12
30
45

Example 6: Transposing Data for Statistical Analysis

Procedure features:

COPY statement
VAR statement

This example arranges data to make it suitable for either a multivariate or a

univariate repeated-measures analysis.
The data is from Chapter 8, Repeated-Measures Analysis of Variance, in SAS
System for Linear Models, Third Edition.

1329

1330

Program 1

Chapter 50

Program 1

Create the WEIGHTS data set. The data in WEIGHTS represents the results of an exercise
therapy study of three weight-lifting programs: CONT is a control group, RI is a program in
which the number of repetitions is increased, and WI is a program in which the weight is
increased.
data weights;
input Program $ s1-s7;
datalines;
CONT 85 85 86 85 87 86 87
CONT 80 79 79 78 78 79 78
CONT 78 77 77 77 76 76 77
CONT 84 84 85 84 83 84 85
CONT 80 81 80 80 79 79 80
RI
79 79 79 80 80 78 80
RI
83 83 85 85 86 87 87
RI
81 83 82 82 83 83 82
RI
81 81 81 82 82 83 81
RI
80 81 82 82 82 84 86
WI
84 85 84 83 83 83 84
WI
74 75 75 76 75 76 76
WI
83 84 82 81 83 83 82
WI
86 87 87 87 87 87 86
WI
82 83 84 85 84 85 86
;

Create the SPLIT data set. This DATA step rearranges WEIGHTS to create the data set
SPLIT. The DATA step transposes the strength values and creates two new variables: Time and
Subject. SPLIT contains one observation for each repeated measure. SPLIT can be used in a
PROC GLM step for a univariate repeated-measures analysis.
data split;
set weights;
array s{7} s1-s7;
Subject + 1;
do Time=1 to 7;
Strength=s{time};
output;
end;
drop s1-s7;
run;

The TRANSPOSE Procedure

Program 2

1331

Print the SPLIT data set. The NOOBS options suppresses the printing of observation
numbers. The OBS= data set option limits the printing to the rst 15 observations. SPLIT has
105 observations.
proc print data=split(obs=15) noobs;
title SPLIT Data Set;
title2 First 15 Observations Only;
run;

Output 1
SPLIT Data Set
First 15 Observations Only
Program
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT

Subject
1
1
1
1
1
1
1
2
2
2
2
2
2
2
3

Time

Strength

1
2
3
4
5
6
7
1
2
3
4
5
6
7
1

85
85
86
85
87
86
87
80
79
79
78
78
79
78
78

Program 2

Set the SAS system options.

options nodate pageno=1 linesize=80 pagesize=40;

Transpose the SPLIT data set. PROC TRANSPOSE transposes SPLIT to create TOTSPLIT.
The TOTSPLIT data set contains the same variables as SPLIT and a variable for each strength
measurement (Str1-Str7). TOTSPLIT can be used for either a multivariate repeated-measures
analysis or a univariate repeated-measures analysis.
proc transpose data=split out=totsplit prefix=Str;

Organize the output data set into BY groups, and populate each BY group with
untransposed values.The variables in the BY and COPY statements are not transposed.
TOTSPLIT contains the variables Program, Subject, Time, and Strength with the same values
that are in SPLIT. The BY statement creates the rst observation in each BY group, which
contains the transposed values of Strength. The COPY statement creates the other observations
in each BY group by copying the values of Time and Strength without transposing them.

1332

Output 2

Chapter 50

by program subject;
copy time strength;

Specify the variable to transpose. The VAR statement species the Strength variable as the
only variable to be transposed.
var strength;
run;

Print the TOTSPLIT data set. The NOOBS options suppresses the printing of observation
numbers. The OBS= data set option limits the printing to the rst 15 observations. SPLIT has
105 observations.
proc print data=totsplit(obs=15) noobs;
title TOTSPLIT Data Set;
title2 First 15 Observations Only;
run;

Output 2

The variables in TOTSPLIT with missing values are used only in a multivariate
repeated-measures analysis. The missing values do not preclude this data set from being used
in a repeated-measures analysis because the MODEL statement in PROC GLM ignores
observations with missing values.

TOTSPLIT Data Set

First 15 Observations Only
Program Subject Time Strength
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT
CONT

1
1
1
1
1
1
1
2
2
2
2
2
2
2
3

1
2
3
4
5
6
7
1
2
3
4
5
6
7
1

85
85
86
85
87
86
87
80
79
79
78
78
79
78
78

_NAME_
Strength

Strength

Str1 Str2 Str3 Str4 Str5 Str6 Str7

85
.
.
.
.
.
.
80
.
.
.
.
.
.
78

85
.
.
.
.
.
.
79
.
.
.
.
.
.
77

86
.
.
.
.
.
.
79
.
.
.
.
.
.
77

85
.
.
.
.
.
.
78
.
.
.
.
.
.
77

87
.
.
.
.
.
.
78
.
.
.
.
.
.
76

86
.
.
.
.
.
.
79
.
.
.
.
.
.
76

87
.
.
.
.
.
.
78
.
.
.
.
.
.
77

1333

CHAPTER

51
The TRANTAB Procedure
Information about the TRANTAB Procedure

1333

Information about the TRANTAB Procedure

See: For documentation of the TRANTAB procedure, see SAS National Language
Support (NLS): Users Guide.

1334

1335

CHAPTER

52
The UNIVARIATE Procedure
Information about the UNIVARIATE Procedure

1335

Information about the UNIVARIATE Procedure

See:

book.

The documentation for the UNIVARIATE procedure has moved to Volume 3 of this

1336

1337

P A R T

Appendices
Appendix

1. . . . . . . . . SAS Elementary Statistics Procedures

Appendix

2. . . . . . . . . Operating Environment-Specic Procedures

Appendix

3. . . . . . . . . Raw Data and DATA Steps

Appendix

4. . . . . . . . . Recommended Reading

1377

1419

1339
1375

1338

1339

APPENDIX

1
SAS Elementary Statistics
Procedures
Overview 1339
Keywords and Formulas 1340
Simple Statistics 1340
Descriptive Statistics 1342
Quantile and Related Statistics 1345
Hypothesis Testing Statistics 1347
Condence Limits for the Mean 1347
Using Weights 1348
Data Requirements for Summarization Procedures
Statistical Background 1348
Populations and Parameters 1348
Samples and Statistics 1349
Measures of Location 1350
The Mean 1350
The Median 1350
The Mode 1350
Percentiles 1350
Quantiles 1350
Measures of Variability 1354
The Range 1354
The Interquartile Range 1355
The Variance 1355
The Standard Deviation 1355
Coefcient of Variation 1355
Measures of Shape 1355
Skewness 1355
Kurtosis 1356
The Normal Distribution 1356
Sampling Distribution of the Mean 1359
Testing Hypotheses 1369
Dening a Hypothesis 1369
Signicance and Power 1370
Students t Distribution 1371
Probability Values 1372
References 1373

1348

Overview
This appendix provides a brief description of some of the statistical concepts
necessary for you to interpret the output of base SAS procedures for elementary

1340

Keywords and Formulas

Appendix 1

statistics. In addition, this appendix lists statistical notation, formulas, and standard
keywords used for common statistics in base SAS procedures. Brief examples illustrate
the statistical concepts.
Table A1.1 on page 1341 lists the most common statistics and the procedures that
compute them.

Keywords and Formulas

Simple Statistics
The base SAS procedures use a standardized set of keywords to refer to statistics.
You specify these keywords in SAS statements to request the statistics to be displayed
or stored in an output data set.
In the following notation, summation is over observations that contain nonmissing
values of the analyzed variable and, except where shown, over nonmissing weights and
frequencies of one or more:

xi
fi
wi

n
x

is the nonmissing value of the analyzed variable for observation i.

is the frequency that is associated with xi if you use a FREQ statement. If you
for all i.
omit the FREQ statement, then fi

is the weight that is associated with xi if you use a WEIGHT statement. The base
procedures automatically exclude the values of xi with missing weights from the
analysis.
By default, the base procedures treat a negative weight as if it is equal to zero.
However, if you use the EXCLNPWGT option in the PROC statement, then the
procedure also excludes those values of xi with nonpositive weights. Note that
most SAS/STAT procedures, such as PROC TTEST and PROC GLM, exclude
values with nonpositive weights by default.
If you omit the WEIGHT statement, then wi
for all i.

is the number of nonmissing values of xi ,

fi . If you use the EXCLNPWGT
option and the WEIGHT statement, then n is the number of nonmissing values
with positive weights.
is the mean

X X
w i xi =

is the variance

wi (xi 0 x)2

SAS Elementary Statistics Procedures

Simple Statistics

where d is the variance divisor (the VARDEF= option) that you specify in the
PROC statement. Valid values are as follows:
When VARDEF=

d equals .

. .

n01

Pw
Pw 0 1

DF
WEIGHT

WDF

The default is DF.

zi
is the standardized variable

(xi

0 x) =s

The standard keywords and formulas for each statistic follow. Some formulas use
keywords to designate the corresponding statistic.
Table A1.1

The Most Common Simple Statistics

Statistic

PROC
MEANS and
SUMMARY

PROC
UNIVARIATE

PROC
PROC
TABULATE REPORT

PROC
CORR

PROC
SQL

Number of missing values

Number of nonmissing
values

Number of observations

Sum of weights

Mean

Sum

Extreme values

Minimum

Maximum

Range

Uncorrected sum of
squares

Corrected sum of squares

Variance

X
X

Covariance
Standard deviation

X
X

1341

1342

Descriptive Statistics

Appendix 1

PROC
MEANS and
SUMMARY

Statistic

PROC
UNIVARIATE

PROC
PROC
TABULATE REPORT

PROC
CORR

PROC
SQL

Standard error of the

mean

Coefcient of variation

Skewness

Kurtosis

Condence Limits
of the mean
of the variance

of quantiles

Median

Mode

Percentiles/Deciles/
Quartiles

t test
for mean=0

for mean= 0

Nonparametric tests for

When VARDEF=N, the skewness is computed as

3
zi

and the weighted skewness is computed as

n
1
=
n

X
X

^
0 x) =j )3
3=2
3
^
wi ((xi 0 x) = )
((xi

3
The formula is invariant under the transformation wi = zwi ; z > 0. When you
use VARDEF=WDF or VARDEF=WEIGHT, the skewnessis set to missing.

Note: PROC MEANS and PROC TABULATE do not compute weighted

skewness. 4
STDDEV|STD
is the standard deviation s and is computed as the square root of the variance, s2 .
STDERR | STDMEAN
is the standard error of the mean, computed as

qX
w

when VARDEF=DF, which is the default. Otherwise, STDERR is set to missing.

SAS Elementary Statistics Procedures

SUM
is the sum, computed as

Xw x
i

SUMWGT
is the sum of the weights,

W , computed as

USS
is the uncorrected sum of squares, computed as

Xw x
i

2
i

VAR
is the variance 2 .

Quantile and Related Statistics

The keywords for quantiles and related statistics are
MEDIAN
is the middle value.
P1
st
is the 1 percentile.
P5
th
is the 5 percentile.
P10
th
is the 10 percentile.
P90
th
is the 90 percentile.
P95
th
is the 95 percentile.
P99
th
is the 99 percentile.
Q1
th
is the lower quartile (25 percentile).
Q3
th
is the upper quartile (75 percentile).
QRANGE
is interquartile range and is calculated as

Q3 0 Q 1

Quantile and Related Statistics

1345

1346

Quantile and Related Statistics

Appendix 1

You use the QNTLDEF= option (PCTLDEF= in PROC UNIVARIATE) to specify the
method that the procedure uses to compute percentiles. Let n be the number of
nonmissing values for a variable, and let x1 ; x2 ; . . . ; xn represent the ordered values of
the variable such that x1 is the smallest value, x2 is next smallest value, and xn is the
largest value. For the tth percentile between 0 and 1, let p = t=100. Then dene j as
the integer part of np and g as the fractional part of np or (n + 1) p, so that

np = j + g
(n + 1) p = j + g

; ; ;

when QNTLDEF = 1 2 3 or 5
when QNTLDEF = 4

Here, QNTLDEF= species the method that the procedure uses to compute the tth
percentile, as shown in the table that follows.
When you use the WEIGHT statement, the tth percentile is computed as

8
> 1 xi
>2
<
> xi+1
>
:
(

where

xi+1 )

if
if

i
Pw

j =1
i

j =1

j < pW <

wj is the weight associated with xi and W

i+1

j =1

n
Pw

i=1

i is the sum of the weights.

When the observations have identical weights, the weighted percentiles are the same as
the unweighted percentiles with QNTLDEF=5.
Table A1.2

Methods for Computing Quantile Statistics

QNTLDEF= Description

Formula

xnp

weighted average at

observation numbered closest to

y = (1 0 g ) xj + gxj +1
where xo is taken to be x1
y = xi

y = xj

if g
even

y = xj +1
where i is the integer part of
3

empirical distribution function

weighted average aimed at

x(n+1)p

g 6=

if g
odd

np + 1
2

y = xj
y = xj +1
y = (1 0 g ) xj + gxj +1
where xn+1 is taken to be xn

1
2
1
2 and

= 1 and j is
2

g=0
if g > 0
if

j is

SAS Elementary Statistics Procedures

QNTLDEF= Description
5

Condence Limits for the Mean

Formula

=0
if g > 0

= 1 (xj + xj+1)
2
y = xj +1
y

empirical distribution function with

averaging

1347

Hypothesis Testing Statistics

The keywords for hypothesis testing statistics are
T
is the Students t statistic to test the null hypothesis that the population mean is
equal to 0 and is calculated as

0
pP w
s=
x

By default, 0 is equal to zero. You can use the MU0= option in the PROC
UNIVARIATE statement to specify 0 . You must use VARDEF=DF, which is the
default variance divisor, otherwise T is set to missing.
By default, when you use a WEIGHT statement, the procedure counts the xi
values with nonpositive weights in the degrees of freedom. Use the EXCLNPWGT
option in the PROC statement to exclude values with nonpositive weights. Most
SAS/STAT procedures, such as PROC TTEST and PROC GLM automatically
exclude values with nonpositive weights.
PROBT
is the two-tailed p-value for Students t statistic, T, with n
degrees of freedom.
This is the probability under the null hypothesis of obtaining a more extreme
value of T than is observed in this sample.

Condence Limits for the Mean

The keywords for condence limits are
CLM
is the two-sided condence limit for the mean. A two-sided
condence interval for the mean has upper and lower limits

100 (1 0 )percent

s
6 t(10=2;n01) pP

q P
where s is
(x 0 x) , t
1

n 1

1 0 =2) critical value of the

(10=2;n01) is the (

degrees of freedom, and is the value of the

Students t statistics with n
ALPHA= option which by default is 0.05. Unless you use VARDEF=DF, which is
the default variance divisor, CLM is set to missing.

LCLM
is the one-sided condence limit below the mean. The one-sided
percent condence interval for the mean has the lower limit

100 (1 0 )

1348

Using Weights

Appendix 1

x 0 t(10;n01)

s
pP w

Unless you use VARDEF=DF, which is the default variance divisor, LCLM is set to
missing.
UCLM
is the one-sided condence limit above the mean. The one-sided
100 (1 )percent condence interval for the mean has the upper limit

x + t(10;n01)

s
pP w

Unless you use VARDEF=DF, which is the default variance divisor, UCLM is set to
missing.

Using Weights
For more information on using weights and an example, see WEIGHT on page 63.

Data Requirements for Summarization Procedures

The following are the minimal data requirements to compute unweighted statistics
and do not describe recommended sample sizes. Statistics are reported as missing if
VARDEF=DF (the default) and these requirements are not met:

3 N and NMISS are computed regardless of the number of missing or nonmissing

observations.

3 SUM, MEAN, MAX, MIN, RANGE, USS, and CSS require at least one nonmissing
observation.

3
3
3
3
3
3

VAR, STD, STDERR, CV, T, and PRT require at least two nonmissing observations.
SKEWNESS requires at least three nonmissing observations.
KURTOSIS requires at least four nonmissing observations.
SKEWNESS, KURTOSIS, T, and PROBT require that STD is greater than zero.
CV requires that MEAN is not equal to zero.
CLM, LCLM, UCLM, STDERR, T, and PROBT require that VARDEF=DF.

Statistical Background
Populations and Parameters
Usually, there is a clearly dened set of elements in which you are interested. This
set of elements is called the universe, and a set of values associated with these elements
is called a population of values. The statistical term population has nothing to do with
people per se. A statistical population is a collection of values, not a collection of people.
For example, a universe is all the students at a particular school, and there could be
two populations of interest: one of height values and one of weight values. Or, a

SAS Elementary Statistics Procedures

Samples and Statistics

1349

universe is the set of all widgets manufactured by a particular company, while the
population of values could be the length of time each widget is used before it fails.
A population of values can be described in terms of its cumulative distribution
function, which gives the proportion of the population less than or equal to each
possible value. A discrete population can also be described by a probability function,
which gives the proportion of the population equal to each possible value. A continuous
population can often be described by a density function, which is the derivative of the
cumulative distribution function. A density function can be approximated by a
histogram that gives the proportion of the population lying within each of a series of
intervals of values. A probability density function is like a histogram with an innite
number of innitely small intervals.
In technical literature, when the term distribution is used without qualication, it
generally refers to the cumulative distribution function. In informal writing,
distribution sometimes means the density function instead. Often the word distribution
is used simply to refer to an abstract population of values rather than some concrete
population. Thus, the statistical literature refers to many types of abstract distributions,
such as normal distributions, exponential distributions, Cauchy distributions, and so
on. When a phrase such as normal distribution is used, it frequently does not matter
whether the cumulative distribution function or the density function is intended.
It may be expedient to describe a population in terms of a few measures that
summarize interesting features of the distribution. One such measure, computed from
the population values, is called a parameter. Many different parameters can be dened
to measure different aspects of a distribution.
The most commonly used parameter is the (arithmetic) mean. If the population
contains a nite number of values, then the population mean is computed as the sum of
all the values in the population divided by the number of elements in the population.
For an innite population, the concept of the mean is similar but requires more
complicated mathematics.
E(x) denotes the mean of a population of values symbolized by x, such as height,
where E stands for expected value. You can also consider expected values of 0 1
derived
functions of the original values. For example, if x represents height, then E x 2 is the
expected value of height squared, that is, the mean value of the population obtained by
squaring every value in the population of heights.

Samples and Statistics

It is often impossible to measure all of the values in a population. A collection of
measured values is called a sample. A mathematical function of a sample of values is
called a statistic. A statistic is to a sample as a parameter is to a population. It is
customary to denote statistics by Roman letters and parameters by Greek letters. For
example, the population mean is often written as , whereas the sample mean is
written as x. The eld of statistics is largely concerned with the study of the behavior of

sample statistics.
Samples can be selected in a variety of ways. Most SAS procedures assume that the
data constitute a simple random sample, which means that the sample was selected in
such a way that all possible samples were equally likely to be selected.
Statistics from a sample can be used to make inferences, or reasonable guesses,
about the parameters of a population. For example, if you take a random sample of 30
students from the high school, then the mean height for those 30 students is a
reasonable guess, or estimate, of the mean height of all the students in the high school.
Other statistics, such as the standard error, can provide information about how good an
estimate is likely to be.
For any population parameter, several statistics can estimate it. Often, however,
there is one particular statistic that is customarily used to estimate a given parameter.

1350

Measures of Location

Appendix 1

For example, the sample mean is the usual estimator of the population mean. In the
case of the mean, the formulas for the parameter and the statistic are the same. In
other cases, the formula for a parameter may be different from that of the most
commonly used estimator. The most commonly used estimator is not necessarily the
best estimator in all applications.

Measures of Location
Measures of location include the mean, the median, and the mode. These measures
describe the center of a distribution. In the denitions that follow, notice that if the
entire sample changes by adding a xed amount to each observation, then these
measures of location are shifted by the same xed amount.

The Mean
The population mean

= E (x ) is usually estimated by the sample mean x.

The Median
The population median is the central value, lying above and below half of the
population values. The sample median is the middle value when the data are arranged
in ascending or descending order. For an even number of observations, the midpoint
between the two middle values is usually reported as the median.

The Mode
The mode is the value at which the density of the population is at a maximum. Some
densities have more than one local maximum (peak) and are said to be multimodal.
The sample mode is the value that occurs most often in the sample. By default, PROC
UNIVARIATE reports the lowest such value if there is a tie for the most-often-occurring
sample value. PROC UNIVARIATE lists all possible modes when you specify the
MODES option in the PROC statement. If the population is continuous, then all sample
values occur once, and the sample mode has little use.

Percentiles
Percentiles, including quantiles, quartiles, and the median, are useful for a detailed
study of a distribution. For a set of measurements arranged in order of magnitude, the
pth percentile is the value that has p percent of the measurements below it and (100p)
percent above it. The median is the 50th percentile. Because it may not be possible to
divide your data so that you get exactly the desired percentile, the UNIVARIATE
procedure uses a more precise denition.
The upper quartile of a distribution is the value below which 75 percent of the
measurements fall (the 75th percentile). Twenty-ve percent of the measurements fall
below the lower quartile value.

Quantiles
In the following example, SAS articially generates the data with a pseudorandom
number function. The UNIVARIATE procedure computes a variety of quantiles and
measures of location, and outputs the values to a SAS data set. A DATA step then uses
the SYMPUT routine to assign the values of the statistics to macro variables. The

SAS Elementary Statistics Procedures

Quantiles

1351

macro %FORMGEN uses these macro variables to produce value labels for the
FORMAT procedure. PROC CHART uses the resulting format to display the values of
the statistics on a histogram.
options nodate pageno=1 linesize=80 pagesize=52;
title Example of Quantiles and Measures of Location;
data random;
drop n;
do n=1 to 1000;
X=floor(exp(rannor(314159)*.8+1.8));
output;
end;
run;
proc univariate data=random nextrobs=0;
var x;
output out=location
mean=Mean mode=Mode median=Median
q1=Q1 q3=Q3 p5=P5 p10=P10 p90=P90 p95=P95
max=Max;
run;

proc print data=location noobs;

run;

data _null_;
set location;
call symput(MEAN,round(mean,1));
call symput(MODE,mode);
call symput(MEDIAN,round(median,1));
call symput(Q1,round(q1,1));
call symput(Q3,round(q3,1));
call symput(P5,round(p5,1));
call symput(P10,round(p10,1));
call symput(P90,round(p90,1));
call symput(P95,round(p95,1));
call symput(MAX,min(50,max));
run;
%macro formgen;
%do i=1 %to &max;
%let value=&i;
%if &i=&p5
%if &i=&p10
%if &i=&q1
%if &i=&mode
%if &i=&median
%if &i=&mean
%if &i=&q3
%if &i=&p90
%if &i=&p95

%then
%then
%then
%then
%then
%then
%then
%then
%then

%let
%let
%let
%let
%let
%let
%let
%let
%let

value=&value
value=&value
value=&value
value=&value
value=&value
value=&value
value=&value
value=&value
value=&value

P5;
P10;
Q1;
Mode;
Median;
Mean;
Q3;
P90;
P95;

1352

Quantiles

Appendix 1

%if &i=&max
&i="&value"
%end;
%mend;

%then %let value=>=&value;

proc format print;

value stat %formgen;
run;
options pagesize=42 linesize=80;
proc chart data=random;
vbar x / midpoints=1 to &max by 1;
format x stat.;
footnote P5 = 5TH PERCENTILE;
footnote2 P10 = 10TH PERCENTILE;
footnote3 P90 = 90TH PERCENTILE;
footnote4 P95 = 95TH PERCENTILE;
footnote5 Q1 = 1ST QUARTILE ;
footnote6 Q3 = 3RD QUARTILE ;

SAS Elementary Statistics Procedures

run;
Example of Quantiles and Measures of Location

The UNIVARIATE Procedure

Variable: X
Moments
N
Mean
Std Deviation
Skewness
Uncorrected SS
Coeff Variation

1000
7.605
7.38169794
2.73038523
112271
97.0637467

Sum Weights
Sum Observations
Variance
Kurtosis
Corrected SS
Std Error Mean

1000
7605
54.4894645
11.1870588
54434.975
0.23342978

Basic Statistical Measures

Location
Mean
Median
Mode

Variability

7.605000
5.000000
3.000000

Std Deviation
Variance
Range
Interquartile Range

7.38170
54.48946
62.00000
6.00000

Tests for Location: Mu0=0

Test

-Statistic-

-----p Value------

Students t
Sign
Signed Rank

t
M
S

Pr > |t|
Pr >= |M|
Pr >= |S|

32.57939
494.5
244777.5

<.0001
<.0001
<.0001

Quantiles (Definition 5)
Quantile

Estimate

100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

62.0
37.5
21.5
16.0
9.0
5.0
3.0
2.0
1.0
0.0
0.0

Example of Quantiles and Measures of Location

Mean

Max

P95

P90

7.605

21.5

Median
5

Q1
3

P10
2

P5
1

Mode
3

Quantiles

1353

1354

Measures of Variability

Appendix 1

Example of Quantiles and Measures of Location

Frequency
120 + *
| *
| **
| ***
90 +*****
|*****
|*******
|*******
60 +*******
|*********
|*********
|*********
30 +************
|************
*
|**************** *
|*********************** * *
--------------------------------------------------1234567891111111111222222222233333333334444444444>
0123456789012345678901234567890123456789=
5
PPQ M MQ
0
511 e e3
P
P
0 d a
9
9
i n
0
5
M a
o n
d
e
X Midpoint

P5
P10
P90
P95
Q1
Q3

= 5TH PERCENTILE
= 10TH PERCENTILE
= 90TH PERCENTILE
= 95TH PERCENTILE
= 1ST QUARTILE
= 3RD QUARTILE

Measures of Variability
Another group of statistics is important in studying the distribution of a population.
These statistics measure the variability, also called the spread, of values. In the
denitions given in the sections that follow, notice that if the entire sample is changed
by the addition of a xed amount to each observation, then the values of these statistics
are unchanged. If each observation in the sample is multiplied by a constant, however,
then the values of these statistics are appropriately rescaled.

The Range
The sample range is the difference between the largest and smallest values in the
sample. For many populations, at least in statistical theory, the range is innite, so the
sample range may not tell you much about the population. The sample range tends to
increase as the sample size increases. If all sample values are multiplied by a constant,
then the sample range is multiplied by the same constant.

SAS Elementary Statistics Procedures

Measures of Shape

1355

The Interquartile Range

The interquartile range is the difference between the upper and lower quartiles. If
all sample values are multiplied by a constant, then the sample interquartile range is
multiplied by the same constant.

The Variance
The population variance, usually denoted by 2 , is the expected value of the squared
difference of the values from the population mean:

2 = E (x

0 )2

The sample variance is denoted by s2 . The difference between a value and the mean
is called a deviation from the mean. Thus, the variance approximates the mean of the
squared deviations.
When all the values lie close to the mean, the variance is small but never less than
zero. When values are more scattered, the variance is larger. If all sample values are
multiplied by a constant, then the sample variance is multiplied by the square of the
constant.
are used in the denominator. The VARDEF=
Sometimes values other than n
option controls what divisor the procedure uses.

The Standard Deviation

The standard deviation is the square root of the variance, or root-mean-square
deviation from the mean, in either a population or a sample. The usual symbols are
for the population and s for a sample. The standard deviation is expressed in the same
units as the observations, rather than in squared units. If all sample values are
multiplied by a constant, then the sample standard deviation is multiplied by the same
constant.

Coefcient of Variation
The coefcient of variation is a unitless measure of relative variability. It is dened
as the ratio of the standard deviation to the mean expressed as a percentage. The
coefcient of variation is meaningful only if the variable is measured on a ratio scale. If
all sample values are multiplied by a constant, then the sample coefcient of variation
remains unchanged.

Measures of Shape

Skewness
The variance is a measure of the overall size of the deviations from the mean. Since
the formula for the variance squares the deviations, both positive and negative
deviations contribute to the variance in the same way. In many distributions, positive
deviations may tend to be larger in magnitude than negative deviations, or vice versa.
Skewness is a measure of the tendency of the deviations to be larger in one direction
than in the other. For example, the data in the last example are skewed to the right.

1356

The Normal Distribution

Appendix 1

Population skewness is dened as

E (x 0 )3 = 3
Because the deviations are cubed rather than squared, the signs of the deviations are
maintained. Cubing the deviations also emphasizes the effects of large deviations. The
formula includes a divisor of 3 to remove the effect of scale, so multiplying all values
by a constant does not change the skewness. Skewness can thus be interpreted as a
tendency for one tail of the population to be heavier than the other. Skewness can be
positive or negative and is unbounded.

Kurtosis
The heaviness of the tails of a distribution affects the behavior of many statistics.
Hence it is useful to have a measure of tail heaviness. One such measure is kurtosis.
The population kurtosis is usually dened as

E (x 0 )4
4

Note:

Some statisticians omit the subtraction of 3.

Because the deviations are raised to the fourth power, positive and negative
deviations make the same contribution, while large deviations are strongly emphasized.
Because of the divisor 4 , multiplying each value by a constant has no effect on kurtosis.
Population kurtosis must lie between 2 and + , inclusive. If M3 represents
population skewness and M4 represents population kurtosis, then

M4 > (M3 )2

Statistical literature sometimes reports that kurtosis measures the peakedness of a

density. However, heavy tails have much more inuence on kurtosis than does the shape
of the distribution near the mean (Kaplansky 1945; Ali 1974; Johnson, et al. 1980).
Sample skewness and kurtosis are rather unreliable estimators of the corresponding
parameters in small samples. They are better estimators when your sample is very
large. However, large values of skewness or kurtosis may merit attention even in small
samples because such values indicate that statistical methods that are based on
normality assumptions may be inappropriate.

The Normal Distribution

One especially important family of theoretical distributions is the normal or Gaussian
distribution. A normal distribution is a smooth symmetric function often referred to as
"bell-shaped." Its skewness and kurtosis are both zero. A normal distribution can be
completely specied by only two parameters: the mean and the standard deviation.
Approximately 68 percent of the values in a normal population are within one standard
deviation of the population mean; approximately 95 percent of the values are within

SAS Elementary Statistics Procedures

The Normal Distribution

1357

two standard deviations of the mean; and about 99.7 percent are within three standard
deviations. Use of the term normal to describe this particular kind of distribution does
not imply that other kinds of distributions are necessarily abnormal or pathological.
Many statistical methods are designed under the assumption that the population
being sampled is normally distributed. Nevertheless, most real-life populations do not
have normal distributions. Before using any statistical method based on normality
assumptions, you should consult the statistical literature to nd out how sensitive the
method is to nonnormality and, if necessary, check your sample for evidence of
nonnormality.
In the following example, SAS generates a sample from a normal distribution with a
mean of 50 and a standard deviation of 10. The UNIVARIATE procedure performs tests
for location and normality. Because the data are from a normal distribution, all p-values
from the tests for normality are greater than 0.15. The CHART procedure displays a
histogram of the observations. The shape of the histogram is a belllike, normal density.
options nodate pageno=1 linesize=80 pagesize=52;
title 10000 Obs Sample from a Normal Distribution;
title2 with Mean=50 and Standard Deviation=10;
data normaldat;
drop n;
do n=1 to 10000;
X=10*rannor(53124)+50;
output;
end;
run;
proc univariate data=normaldat nextrobs=0 normal
mu0=50 loccount;
var x;
run;

proc format;
picture msd
20=20 3*Std
30=30 2*Std
40=40 1*Std
50=50 Mean
60=60 1*Std
70=70 2*Std
80=80 3*Std
other= ;
run;
options linesize=80

(noedit)
(noedit)
(noedit)
(noedit)
(noedit)
(noedit)
(noedit)

pagesize=42;

proc chart;
vbar x / midpoints=20 to 80 by 2;
format x msd.;
run;

1358

The Normal Distribution

Appendix 1

10000 Obs Sample from a Normal Distribution

with Mean=50 and Standard Deviation=10

The UNIVARIATE Procedure

Variable: X
Moments
N
Mean
Std Deviation
Skewness
Uncorrected SS
Coeff Variation

10000
50.0323744
9.92013874
-0.019929
26016378
19.8274395

Sum Weights
Sum Observations
Variance
Kurtosis
Corrected SS
Std Error Mean

10000
500323.744
98.4091525
-0.0163755
983993.116
0.09920139

Basic Statistical Measures

Location
Mean
Median
Mode

Variability

50.03237
50.06492
.

Std Deviation
Variance
Range
Interquartile Range

9.92014
98.40915
76.51343
13.28179

Tests for Location: Mu0=50

Test

-Statistic-

-----p Value------

Students t
Sign
Signed Rank

t
M
S

Pr > |t|
Pr >= |M|
Pr >= |S|

0.32635
26
174063

0.7442
0.6101
0.5466

Location Counts: Mu0=50.00

Count

Value

Num Obs > Mu0

Num Obs ^= Mu0
Num Obs < Mu0

5026
10000
4974

Tests for Normality

Test

--Statistic---

-----p Value------

Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

D
W-Sq
A-Sq

Pr > D
Pr > W-Sq
Pr > A-Sq

0.006595
0.049963
0.371151

>0.1500
>0.2500
>0.2500

SAS Elementary Statistics Procedures

10000 Obs Sample from a Normal Distribution

with Mean=50 and Standard Deviation=10

Sampling Distribution of the Mean

1359

The UNIVARIATE Procedure

Variable: X
Quantiles (Definition 5)
Quantile

Estimate

100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

90.2105
72.6780
66.2221
62.6678
56.7280
50.0649
43.4462
37.1139
33.5454
26.9189
13.6971

10000 Obs Sample from a Normal Distribution

with Mean=50 and Standard Deviation=10

Frequency
|
*
800 +
***
|
****
|
******
|
*******
600 +
*******
|
**********
|
***********
|
***********
400 +
************
|
*************
|
***************
|
*****************
200 +
******************
|
*******************
|
**********************
| ***************************
-------------------------------2
3
4
5
6
7
8
0
0
0
0
0
0
0
3
*
S
t
d

2
*
S
t
d

1
*
S
t
d

M
e
a
n

1
*
S
t
d

2
*
S
t
d

3
*
S
t
d

X Midpoint

Sampling Distribution of the Mean

If you repeatedly draw samples of size n from a population and compute the mean of
each sample, then the sample means themselves have a distribution. Consider a new
population consisting of the means of all the samples that could possibly be drawn from
the original population. The distribution of this new population is called a sampling
distribution.

1360

Sampling Distribution of the Mean

Appendix 1

It can be proven mathematically that if the original population has mean and
standard deviation , then the sampling distribution of the mean also has mean , but
its standard deviation is = n. The standard deviation of the sampling distribution of
the mean is called the standard error of the mean. The standard error of the mean
provides an indication of the accuracy of a sample mean as an estimator of the
population mean.
If the original population has a normal distribution, then the sampling distribution of
the mean is also normal. If the original distribution is not normal but does not have
excessively long tails, then the sampling distribution of the mean can be approximated
by a normal distribution for large sample sizes.
The following example consists of three separate programs that show how the
sampling distribution of the mean can be approximated by a normal distribution as the
sample size increases. The rst DATA step uses the RANEXP function to create a
sample of 1000 observations from an exponential distribution.The theoretical
population mean is 1.00, while the sample mean is 1.01, to two decimal places. The
population standard deviation is 1.00; the sample standard deviation is 1.04.
This is an example of a nonnormal distribution. The population skewness is 2.00,
which is close to the sample skewness of 1.97. The population kurtosis is 6.00, but the
sample kurtosis is only 4.80.

options nodate pageno=1 linesize=80 pagesize=42;

title 1000 Observation Sample;
title2 from an Exponential Distribution;
data expodat;
drop n;
do n=1 to 1000;
X=ranexp(18746363);
output;
end;
run;
proc format;
value axisfmt
.05=0.05
.55=0.55
1.05=1.05
1.55=1.55
2.05=2.05
2.55=2.55
3.05=3.05
3.55=3.55
4.05=4.05
4.55=4.55
5.05=5.05
5.55=5.55
other= ;
run;
proc chart data=expodat ;
vbar x / axis=300
midpoints=0.05 to 5.55 by .1;
format x axisfmt.;
run;

SAS Elementary Statistics Procedures

options pagesize=64;
proc univariate data=expodat noextrobs=0 normal
mu0=1;
var x;

Sampling Distribution of the Mean

1361

1362

Sampling Distribution of the Mean

Appendix 1

run;
1000 Observation Sample
from an Exponential Distribution

Frequency
300 +
|
|
|
|
250 +
|
|
|
|
200 +
|
|
|
|
150 +
|
|
|
|
100 +*
|*
|*** *
|*****
|***** *
50 +********
|***********
|************ *
|*************** **
*
|************************* *** *** *
*
*
--------------------------------------------------------0
0
1
1
2
2
3
3
4
4
5
5
.
.
.
.
.
.
.
.
.
.
.
.
0
5
0
5
0
5
0
5
0
5
0
5
5
5
5
5
5
5
5
5
5
5
5
5
X Midpoint

1000 Observation Sample

from an Exponential Distribution

The UNIVARIATE Procedure

Variable: X
Moments
N
Mean
Std Deviation
Skewness
Uncorrected SS
Coeff Variation

1000
1.01176214
1.04371187
1.96963112
2111.90777
103.15783

Sum Weights
Sum Observations
Variance
Kurtosis
Corrected SS
Std Error Mean

1000
1011.76214
1.08933447
4.80150594
1088.24514
0.03300507

Basic Statistical Measures

Location
Mean
Median
Mode

1.011762
0.689502
.

Variability
Std Deviation
Variance
Range
Interquartile Range

1.04371
1.08933
6.63851
1.06252

SAS Elementary Statistics Procedures

Sampling Distribution of the Mean

1363

Tests for Location: Mu0=1

Test

-Statistic-

-----p Value------

Students t
Sign
Signed Rank

t
M
S

Pr > |t|
Pr >= |M|
Pr >= |S|

0.356374
-140
-50781

0.7216
<.0001
<.0001

Tests for Normality

Test

--Statistic---

-----p Value------

Shapiro-Wilk
Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

W
D
W-Sq
A-Sq

Pr
Pr
Pr
Pr

0.801498
0.166308
9.507975
54.5478

<
>
>
>

W
D
W-Sq
A-Sq

<0.0001
<0.0100
<0.0050
<0.0050

Quantiles (Definition 5)
Quantile
100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

Estimate
6.63906758
5.04491651
3.13482318
2.37803632
1.35733401
0.68950221
0.29481436
0.10219011
0.05192799
0.01195590
0.00055441

The next DATA step generates 1000 different samples from the same exponential
distribution. Each sample contains ten observations. The MEANS procedure computes
the mean of each sample. In the data set that is created by PROC MEANS, each
observation represents the mean of a sample of ten observations from an exponential
distribution. Thus, the data set is a sample from the sampling distribution of the mean
for an exponential population.
PROC UNIVARIATE displays statistics for this sample of means. Notice that the
mean of the sample of means is .99, almost the same as the mean of the original
population. Theoretically, the standard deviation of the sampling distribution is
= n
: =
: , whereas the standard deviation of this sample from
thesampling distribution is .30. The skewness (.55) and kurtosis (-.006) are closer to
zero in the sample from the sampling distribution than in the original sample from the
exponential distribution. This is so because the sampling distribution is closer to a
normal distribution than is the original exponential distribution. The CHART
procedure displays a histogram of the 1000-sample means. The shape of the histogram
is much closer to a belllike, normal density, but it is still distinctly lopsided.

p = 1 00 p10 = 32

options nodate pageno=1 linesize=80 pagesize=48;

title 1000 Sample Means with 10 Obs per Sample;
title2 Drawn from an Exponential Distribution;
data samp10;
drop n;
do Sample=1 to 1000;
do n=1 to 10;

1364

Sampling Distribution of the Mean

Appendix 1

X=ranexp(433879);
output;
end;
end;
proc means data=samp10 noprint;
output out=mean10 mean=Mean;
var x;
by sample;
run;

proc format;
value axisfmt
.05=0.05
.55=0.55
1.05=1.05
1.55=1.55
2.05=2.05
other= ;
run;
proc chart data=mean10;
vbar mean/axis=300
midpoints=0.05 to 2.05 by .1;
format mean axisfmt.;
run;

options pagesize=64;
proc univariate data=mean10 noextrobs=0 normal
mu0=1;
var mean;

SAS Elementary Statistics Procedures

Sampling Distribution of the Mean

run;
1000 Sample Means with 10 Obs per Sample
Drawn from an Exponential Distribution

Frequency
300 +
|
|
|
|
250 +
|
|
|
|
200 +
|
|
|
|
150 +
|
*
|
* *
*
|
* * * *
|
* * * *
100 +
* * * *
|
* * * * *
|
* * * * * *
|
* * * * * *
|
* * * * * * *
*
50 +
* * * * * * * * * *
|
* * * * * * * * * *
|
* * * * * * * * * * *
|
* * * * * * * * * * * *
|
* * * * * * * * * * * * * * * *
-------------------------------------------0
0
1
1
2
.
.
.
.
.
0
5
0
5
0
5
5
5
5
5
Mean Midpoint

1000 Sample Means with 10 Obs per Sample

Drawn from an Exponential Distribution

The UNIVARIATE Procedure

Variable: Mean
Moments
N
Mean
Std Deviation
Skewness
Uncorrected SS
Coeff Variation

1000
0.9906857
0.30732649
0.54575615
1075.81327
31.0215931

Sum Weights
Sum Observations
Variance
Kurtosis
Corrected SS
Std Error Mean

1000
990.685697
0.09444957
-0.0060892
94.3551193
0.00971852

Basic Statistical Measures

Location
Mean
Median
Mode

0.990686
0.956152
.

Variability
Std Deviation
Variance
Range
Interquartile Range

0.30733
0.09445
1.79783
0.41703

1365

1366

Sampling Distribution of the Mean

Appendix 1

Tests for Location: Mu0=1

Test

-Statistic-

-----p Value------

Students t
Sign
Signed Rank

t
M
S

Pr > |t|
Pr >= |M|
Pr >= |S|

-0.95841
-53
-22687

0.3381
0.0009
0.0129

Tests for Normality

Test

--Statistic---

-----p Value------

Shapiro-Wilk
Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

W
D
W-Sq
A-Sq

Pr
Pr
Pr
Pr

0.9779
0.055498
0.953926
5.945023

<
>
>
>

W
D
W-Sq
A-Sq

<0.0001
<0.0100
<0.0050
<0.0050

Quantiles (Definition 5)
Quantile

Estimate

100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

2.053899
1.827503
1.557175
1.416611
1.181006
0.956152
0.763973
0.621787
0.553568
0.433820
0.256069

In the following DATA step, the size of each sample from the exponential distribution
is increased to 50. The standard deviation of the sampling distribution is smaller than
in the previous example because the size of each sample is larger. Also, the sampling
distribution is even closer to a normal distribution, as can be seen from the histogram
and the skewness.
options nodate pageno=1 linesize=80 pagesize=48;
title 1000 Sample Means with 50 Obs per Sample;
title2 Drawn from an Exponential Distribution;
data samp50;
drop n;
do sample=1 to 1000;
do n=1 to 50;
X=ranexp(72437213);
output;
end;
end;
proc means data=samp50 noprint;
output out=mean50 mean=Mean;
var x;
by sample;
run;

SAS Elementary Statistics Procedures

proc format;
value axisfmt
.05=0.05
.55=0.55
1.05=1.05
1.55=1.55
2.05=2.05
2.55=2.55
other= ;
run;
proc chart data=mean50;
vbar mean / axis=300
midpoints=0.05 to 2.55 by .1;
format mean axisfmt.;
run;

options pagesize=64;
proc univariate data=mean50 nextrobs=0 normal
mu0=1;
var mean;

Sampling Distribution of the Mean

1367

1368

Sampling Distribution of the Mean

Appendix 1

run;
1000 Sample Means with 50 Obs per Sample
Drawn from an Exponential Distribution

Frequency
300 +
|
|
|
*
|
* *
250 +
* *
|
* *
|
* *
|
* *
|
* *
200 +
* *
|
* *
|
* * *
|
* * *
|
* * *
150 +
* * * *
|
* * * *
|
* * * *
|
* * * *
|
* * * *
100 +
* * * *
|
* * * *
|
* * * *
|
* * * * *
|
* * * * * *
50 +
* * * * * *
|
* * * * * *
|
* * * * * *
|
* * * * * * *
|
* * * * * * * *
-----------------------------------------------------0
0
1
1
2
2
.
.
.
.
.
.
0
5
0
5
0
5
5
5
5
5
5
5
Mean Midpoint

1000 Sample Means with 50 Obs per Sample

Drawn from an Exponential Distribution

The UNIVARIATE Procedure

Variable: Mean
Moments
N
Mean
Std Deviation
Skewness
Uncorrected SS
Coeff Variation

1000
0.99679697
0.13815404
0.19062633
1012.67166
13.8597969

Sum Weights
Sum Observations
Variance
Kurtosis
Corrected SS
Std Error Mean

1000
996.796973
0.01908654
-0.1438604
19.067451
0.00436881

Basic Statistical Measures

Location
Mean
Median
Mode

0.996797
0.996023
.

Variability
Std Deviation
Variance
Range
Interquartile Range

0.13815
0.01909
0.87040
0.18956

SAS Elementary Statistics Procedures

Testing Hypotheses

1369

Tests for Location: Mu0=1

Test

-Statistic-

-----p Value------

Students t
Sign
Signed Rank

t
M
S

Pr > |t|
Pr >= |M|
Pr >= |S|

-0.73316
-13
-10767

0.4636
0.4292
0.2388

Tests for Normality

Test

--Statistic---

-----p Value------

Shapiro-Wilk
Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

W
D
W-Sq
A-Sq

Pr
Pr
Pr
Pr

0.996493
0.023687
0.084468
0.66039

<
>
>
>

W
D
W-Sq
A-Sq

0.0247
>0.1500
0.1882
0.0877

Quantiles (Definition 5)
Quantile

Estimate

100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

1.454957
1.337016
1.231508
1.179223
1.086515
0.996023
0.896953
0.814906
0.780783
0.706588
0.584558

Testing Hypotheses
Dening a Hypothesis
The purpose of the statistical methods that have been discussed so far is to estimate
a population parameter by means of a sample statistic. Another class of statistical
methods is used for testing hypotheses about population parameters or for measuring
the amount of evidence against a hypothesis.
Consider the universe of students in a college. Let the variable X be the number of
pounds by which a students weight deviates from the ideal weight for a person of the
same sex, height, and build. You want to nd out whether the population of students is,
on the average, underweight or overweight. To this end, you have taken a random
sample of X values from nine students, with results as given in the following DATA step:
title Deviations from Normal Weight;
data x;
input X @@;
datalines;
-7 -2 1 3 6 10 15 21 30
;

You can dene several hypotheses of interest. One hypothesis is that, on the average,
the students are of exactly ideal weight. If represents the population mean of the X
values, then you can write this hypothesis, called the null hypothesis, as 0
.

H : =0

1370

Testing Hypotheses

Appendix 1

The other two hypotheses, called alternative hypotheses, are that the students are
, and that the students are overweight on the
underweight on the average, 1
.
average, 2
The null hypothesis is so called because in many situations it corresponds to the
assumption of no effect or no difference. However, this interpretation is not
appropriate for all testing problems. The null hypothesis is like a straw man that can
be toppled by statistical evidence. You decide between the alternative hypotheses
according to which way the straw man falls.
A naive way to approach this problem would be to look at the sample mean and
decide among the three hypotheses according to the following rule:
3 If
, then decide on 1
.
, then decide on 0
.
3 If
3 If
, then decide on 2
.

H : <0

H : >0

x<0

x=0

x>0

H : <0
H : =0
H : >0

The trouble with this approach is that there may be a high probability of making an
incorrect decision. If H0 is true, then you are nearly certain to make a wrong decision
because the chances of being exactly zero are almost nil. If is slightly less than
zero, so that H1 is true, then there may be nearly a 50 percent chance that will be
greater than zero in repeated sampling, so the chances of incorrectly choosing H2 would
also be nearly 50 percent. Thus, you have a high probability of making an error if is
near zero. In such cases, there is not enough evidence to make a condent decision, so
the best response may be to reserve judgment until you can obtain more evidence.
The question is, how far from zero must be for you to be able to make a condent
decision? The answer can be obtained by considering the sampling distribution of . If
X has a roughly normal distribution, then has an approximately normal sampling
distribution. The mean of the sampling distribution of is . Assume temporarily that
, the standard deviation of X, is known to be 12. Then the standard error of for
.
samples of nine observations is
You know that about 95 percent of the values from a normal distribution are within
two standard deviations of the mean, so about 95 percent of the possible samples of
nine X values have a sample mean between
and
, or between 8
and 8. Consider the chances of making an error with the following decision rule:
, then decide on 1
.
3 If
, then reserve judgment.
3 If
3 If
, then decide on 2
.

x

x

p
=pn = 12= 9 = 4

x < 08

08 x 8

x>8

x

0 0 2 (4)
H : <0

H : >0

0 + 2 (4)

If H0 is true, then in about 95 percent of the possible samples will be between the
critical values
and 8, so you will reserve judgment. In these cases the statistical
evidence is not strong enough to fell the straw man. In the other 5 percent of the
samples you will make an error; in 2.5 percent of the samples you will incorrectly
choose H1, and in 2.5 percent you will incorrectly choose H2.
The price you pay for controlling the chances of making an error is the necessity of
reserving judgment when there is not sufcient statistical evidence to reject the null
hypothesis.

Signicance and Power

The probability of rejecting the null hypothesis if it is true is called the Type I error
rate of the statistical test and is typically denoted as . In this example, an value less
than
or greater than 8 is said to be statistically signicant at the 5 percent level.
You can adjust the type I error rate according to your needs by choosing different critical
values. For example, critical values of 4 and 4 would produce a signicance level of
about 32 percent, while 12 and 12 would give a type I error rate of about 0.3 percent.
The decision rule is a two-tailed test because the alternative hypotheses allow for
population means either smaller or larger than the value specied in the null

SAS Elementary Statistics Procedures

Testing Hypotheses

1371

hypothesis. If you were interested only in the possibility of the students being
overweight on the average, then you could use a one-tailed test:

3 If x

3 If x

8, then reserve judgment.

> 8, then decide on H2 : > 0.

For this one-tailed test, the type I error rate is 2.5 percent, half that of the two-tailed
test.
The probability of rejecting the null hypothesis if it is false is called the power of the
statistical test and is typically denoted as 1 . is called the Type II error rate,
which is the probability of not rejecting a false null hypothesis. The power depends on
the true value of the parameter. In the example, assume the population mean is 4. The
power for detecting H2 is the probability of getting a sample mean greater than 8. The
critical value 8 is one standard error higher than the population mean 4. The chance of
getting a value at least one standard deviation greater than the mean from a normal
distribution is about 16 percent, so the power for detecting the alternative hypothesis
H2 is about 16 percent. If the population mean were 8, then the power for H2 would be
50 percent, whereas a population mean of 12 would yield a power of about 84 percent.
The smaller the type I error rate is, the less the chance of making an incorrect
decision, but the higher the chance of having to reserve judgment. In choosing a type I
error rate, you should consider the resulting power for various alternatives of interest.

Students t Distribution
In practice, you usually cannot use any decision rule that uses a critical value based
on because you do not usually know the value of . You can, however, use s as an
estimate of . Consider the following statistic:

x 0 0

p
s= n

This t statistic is the difference between the sample mean and the hypothesized
mean 0 divided by the estimated standard error of the mean.
If the null hypothesis is true and the population is normally distributed, then the t
statistic has what is called a Students t distribution with n 1 degrees of freedom.
This distribution looks very similar to a normal distribution, but the tails of the
Students t distribution are heavier. As the sample size gets larger, the sample standard
deviation becomes a better estimator of the population standard deviation, and the t
distribution gets closer to a normal distribution.
You can base a decision rule on the t statistic:

0
0

3 If t < 2:3, then decide on H1 : < 0.

3 If 2:3 t 2:3, then reserve judgment.
3 If t > 2:3, then decide on H0 : > 0.
The value 2.3 was obtained from a table of Students t distribution to give a type I
error rate of 5 percent for 8 (that is, 9 1 = 8) degrees of freedom. Most common
statistics texts contain a table of Students t distribution. If you do not have a statistics
text handy, then you can use the DATA step and the TINV function to print any values
from the t distribution.
By default, PROC UNIVARIATE computes a t statistic for the null hypothesis that
0 = 0, along with related statistics. Use the MU0= option in the PROC statement to
specify another value for the null hypothesis.
This example uses the data on deviations from normal weight, which consist of nine
observations. First, PROC MEANS computes the t statistic for the null hypothesis that

1372

Testing Hypotheses

Appendix 1

= 0.

Then, the TINV function in a DATA step computes the value of Students t
distribution for a two-tailed test at the 5 percent level of signicance and 8 degrees of
freedom.
data devnorm;
title Deviations from Normal Weight;
input X @@;
datalines;
-7 -2 1 3 6 10 15 21 30
;
proc means data=devnorm maxdec=3 n mean
std stderr t probt;
run;
title Students t Critical Value;
data _null_;
file print;
t=tinv(.975,8);
put t 5.3;
run;

Deviations from Normal Weight

The MEANS Procedure

Analysis Variable : X
N
Mean
Std Dev
Std Error t Value Pr > |t|
-------------------------------------------------------------9
8.556
11.759
3.920
2.18
0.0606
--------------------------------------------------------------

Students t Critical Value

2.306

In the current example, the value of the t statistic is 2.18, which is less than the critical
t value of 2.3 (for a 5 percent signicance level and 8 degrees of freedom). Thus, at a 5
percent signicance level you must reserve judgment. If you had elected to use a 10
percent signicance level, then the critical value of the t distribution would have been
1.86 and you could have rejected the null hypothesis. The sample size is so small,
however, that the validity of your conclusion depends strongly on how close the
distribution of the population is to a normal distribution.

Probability Values
Another way to report the results of a statistical test is to compute a probability
value or p-value. A p-value gives the probability in repeated sampling of obtaining a
statistic as far in the direction(s) specied by the alternative hypothesis as is the value
actually observed. A two-tailed p-value for a t statistic is the probability of obtaining an
absolute t value that is greater than the observed absolute t value. A one-tailed p-value
for a t statistic for the alternative hypothesis
0 is the probability of obtaining a t

SAS Elementary Statistics Procedures

References

1373

value greater than the observed t value. Once the p-value is computed, you can perform
a hypothesis test by comparing the p-value with the desired signicance level. If the
p-value is less than or equal to the type I error rate of the test, then the null hypothesis
can be rejected. The two-tailed p-value, labeled Pr > |t| in the PROC MEANS output,
is .0606, so the null hypothesis could be rejected at the 10 percent signicance level but
not at the 5 percent level.
A p-value is a measure of the strength of the evidence against the null hypothesis.
The smaller the p-value, the stronger the evidence for rejecting the null hypothesis.
Note: For a more thorough discussion, consult an introductory statistics textbook
such as Mendenhall and Beaver (1998); Ott and Mendenhall (1994); or Snedecor and
Cochran (1989). 4

References
Ali, M.M. (1974), Stochastic Ordering and Kurtosis Measure, Journal of the
American Statistical Association, 69, 543545.
Johnson, M.E., Tietjen, G.L., and Beckman, R.J. (1980), A New Family of
Probability Distributions With Applications to Monte Carlo Studies, Journal of
the American Statistical Association, 75, 276-279.
Kaplansky, I. (1945), A Common Error Concerning Kurtosis, Journal of the
American Statistical Association, 40, 259-263.
Mendenhall, W. and Beaver, R.. (1998), Introduction to Probability and Statistics,
10th Edition, Belmont, CA: Wadsworth Publishing Company.
Ott, R. and Mendenhall, W. (1994) Understanding Statistics, 6th Edition, North
Scituate, MA: Duxbury Press.
Schlotzhauer, S.D. and Littell, R.C. (1997), SAS System for Elementary Statistical
Analysis, Second Edition, Cary, NC: SAS Institute Inc.
Snedecor, G.W. and Cochran, W.C. (1989), Statistical Methods, 8th Edition, Ames, IA:
Iowa State University Press.

1374

1375

APPENDIX

2
Operating Environment-Specic
Procedures
Descriptions of Operating Environment-Specic Procedures

1375

Descriptions of Operating Environment-Specic Procedures

The following table gives a brief description and the relevant releases for some
common operating environment-specic procedures. All of these procedures are
described in more detail in operating environment-companion documentation.
Table A2.1

Host-Specic Procedures

Procedure

Description

Releases

BMDP

Calls any BMDP program to analyze data in a SAS data set.

All

CONVERT

Converts BMDP, OSIRIS, and SPSS system les to SAS data

sets.

All

C16PORT

Converts a 16-bit SAS data library or catalog created in Release

6.08 to a transport le, which you can then convert to a 32-bit
format for use in the current release of SAS by using the
CIMPORT procedure.

6.10 - 6.12

FSDEVICE

Creates, copies, modies, deletes, or renames device descriptions

in a catalog.

All

PDS

Lists, deletes, or renames the members of a partitioned data set.

6.09E

PDSCOPY

Copies partitioned data sets from disk to disk, disk to tape, tape
to tape, or tape to disk.

6.09E

RELEASE

volume to the SAS procedure output le.

6.09E

1376

1377

APPENDIX

3
Raw Data and DATA Steps
Overview 1377
CENSUS 1377
CHARITY 1378
CUSTOMER_RESPONSE 1380
DJIA 1383
EDUCATION 1384
EMPDATA 1385
ENERGY 1387
GROC 1388
MATCH_11 1388
PROCLIB.DELAY 1390
PROCLIB.EMP95 1391
PROCLIB.EMP96 1392
PROCLIB.INTERNAT 1393
PROCLIB.LAKES 1393
PROCLIB.MARCH 1394
PROCLIB.PAYLIST2 1395
PROCLIB.PAYROLL 1395
PROCLIB.PAYROLL2 1398
PROCLIB.SCHEDULE 1399
PROCLIB.STAFF 1402
PROCLIB.SUPERV 1405
RADIO 1405

Overview
The programs for examples in this document generally show you how to create the
data sets that are used. Some examples show only partial data. For these examples,
the complete data is shown in this appendix.

CENSUS
data census;
input Density CrimeRate State $ 14-27 PostalCode $ 29-30;
datalines;
263.3 4575.3 Ohio

1378

CHARITY

62.1 7017.1

Appendix 3

Washington

103.4 5161.9 South Carolina SC

53.4 3438.6

Mississippi

180.0 8503.2 Florida

80.8 2190.7

West Virginia

428.7 5477.6 Maryland

71.2 4707.5

Missouri

43.9 4245.2

Arkansas

7.3 6371.4

Nevada

264.3 3163.2 Pennsylvania

11.5 4156.3

Idaho

44.1 6025.6

Oklahoma

51.2 4615.8

Minnesota

55.2 4271.2

Vermont

27.4 6969.9

Oregon

205.3 5416.5 Illinois

94.1 5792.0

Georgia

9.1 2678.0

South Dakota

9.4 2833.0

North Dakota

102.4 3371.7 New Hampshire

54.3 7722.4

Texas

76.6 4451.4

Alabama

307.6 4938.8 Delaware

151.4 6506.4 California

1992 Edward

53.76 31

Monroe

1992 Fiona

48.55 13

Monroe

1992 Gert

24.00 16

Monroe

1992 Harold

27.55 17

Monroe

1992 Ima

15.98

Monroe

1992 Jack

20.00 23

Monroe

1992 Katie

22.11

Monroe

1992 Lisa

18.34 17

Monroe

1992 Tonya

55.16 40

Monroe

1992 Max

26.77 34

Monroe

1992 Ned

28.43 22

Monroe

1992 Opal

32.66 14

Monroe

1993 Patsy

18.33 18

Monroe

1993 Quentin 16.89 15

23.76 16
5

6.89 23

Raw Data and DATA Steps

Monroe

1993 Randall 12.98 17

Monroe

1993 Sam

15.88

Monroe

1993 Tyra

21.88 23

Monroe

1993 Myrtle

47.33 26

Monroe

1993 Frank

41.11 22

Monroe

1993 Cameron 65.44 14

Monroe

1993 Vern

Monroe

1993 Wendell 23.00 10

Monroe

1993 Bob

26.88

Monroe

1993 Leah

28.99 23

Monroe

1994 Becky

30.33 26

Monroe

1994 Sally

35.75 27

Monroe

1994 Edgar

27.11 12

Monroe

1994 Dawson

17.24 16

Monroe

1994 Lou

Monroe

1994 Damien

18.74 17

Monroe

1994 Mona

27.43

Monroe

1994 Della

56.78 15

Monroe

1994 Monique 29.88 19

Monroe

1994 Carl

31.12 25

Monroe

1994 Reba

35.16 22

Monroe

1994 Dax

27.65 23

Monroe

1994 Gary

23.11 15

Monroe

1994 Suzie

26.65 11

Monroe

1994 Benito

47.44 18

Monroe

1994 Thomas

21.99 23

Monroe

1994 Annie

24.99 27

Monroe

1994 Paul

27.98 22

Monroe

1994 Alex

24.00 16

Monroe

1994 Lauren

15.00 17

Monroe

1994 Julia

12.98 15

Monroe

1994 Keith

11.89 19

Monroe

1994 Jackie

26.88 22

Monroe

1994 Pablo

13.98 28

Monroe

1994 L.T.

56.87 33

Monroe

1994 Willard 78.65 24

Monroe

1994 Kathy

Monroe

1994 Abby

17.89 11

5.12 16

32.88 11
35.88 10

Kennedy 1992 Arturo

34.98 14

Kennedy 1992 Grace

27.55 25

Kennedy 1992 Winston 23.88 22

Kennedy 1992 Vince

12.88 21

Kennedy 1992 Claude

15.62

Kennedy 1992 Mary

28.99 34

Kennedy 1992 Abner

25.89 22

Kennedy 1992 Jay

35.89 35

Kennedy 1992 Alicia

28.77 26

Kennedy 1992 Freddy

29.00 27

Kennedy 1992 Eloise

31.67 25

Kennedy 1992 Jenny

43.89 22

Kennedy 1992 Thelma

52.63 21

Kennedy 1992 Tina

19.67 21

Kennedy 1992 Eric

24.89 12

Kennedy 1993 Bubba

37.88 12

CHARITY

1379

1380

CUSTOMER_RESPONSE

Kennedy 1993 G.L.

28.89 21

Kennedy 1993 Clay

26.44 21

Kennedy 1993 Leeann

Appendix 3

25.89 21

Kennedy 1993 Bert

27.17 17

Kennedy 1993 Georgia 38.90 11

Kennedy 1993 Bill

42.23 25

Kennedy 1993 Holly

18.67 27

Kennedy 1993 Benny

19.09 25

Kennedy 1993 Cammie

28.77 28

Kennedy 1993 Amy

27.08 31

Kennedy 1993 Doris

22.22 24

Kennedy 1993 Robbie

19.80 24

Kennedy 1993 Ted

27.07 25

Kennedy 1993 Sarah

24.44 12

Kennedy 1993 Megan

28.89 11

Kennedy 1993 Jeff

31.11 12

Kennedy 1993 Taz

30.55 11

Kennedy 1993 George

27.56 11

Kennedy 1993 Heather 38.67 15

Kennedy 1994 Nancy

29.90 26

Kennedy 1994 Rusty

30.55 28

Kennedy 1994 Mimi

37.67 22

Kennedy 1994 J.C.

23.33 27

Kennedy 1994 Clark

27.90 25

Kennedy 1994 Rudy

27.78 23

Kennedy 1994 Samuel

34.44 18

Kennedy 1994 Forrest 28.89 26

Kennedy 1994 Luther
Kennedy 1994 Trey
Kennedy 1994 Albert

72.22 24
6.78 18
23.33 19

Kennedy 1994 Che-Min 26.66 33

Kennedy 1994 Preston 32.22 23
Kennedy 1994 Larry

40.00 26

Kennedy 1994 Anton

35.99 28

Kennedy 1994 Sid

27.45 25

Kennedy 1994 Will

28.88 21

Kennedy 1994 Morty

34.44 25

CUSTOMER_RESPONSE
data customer_response;
input Customer Factor1-Factor4 Source1-Source3
Quality1-Quality3;
datalines;
1 . . 1 1 1 1 . 1 . .
2 1 1 . 1 1 1 . 1 1 .
3 . . 1 1 1 1 . . . .
4 1 1 . 1 . 1 . . . 1
5 . 1 . 1 1 . . . . 1
6 . 1 . 1 1 . . . . .
7 . 1 . 1 1 . . 1 . .

Raw Data and DATA Steps

8 1 . . 1 1 1 . 1 1 .
9 1 1 . 1 1 . . . . 1
10 1 . . 1 1 1 . 1 1 .
11 1 1 1 1 . 1 . 1 1 1
12 1 1 . 1 1 1 . . . .
13 1 1 . 1 . 1 . 1 1 .
14 1 1 . 1 1 1 . . . .
15 1 1 . 1 . 1 . 1 1 1
16 1 . . 1 1 . . 1 . .
17 1 1 . 1 1 1 . . 1 .
18 1 1 . 1 1 1 1 . . 1
19 . 1 . 1 1 1 1 . 1 .
20 1 . . 1 1 1 . 1 1 1
21 . . . 1 1 1 . 1 . .
22 . . . 1 1 1 . 1 1 .
23 1 . . 1 . . . . . 1
24 . 1 . 1 1 . . 1 . 1
25 1 1 . 1 1 . . . 1 1
26 1 1 . 1 1 . . 1 . .
27 1 . . 1 1 . . . 1 .
28 1 1 . 1 . . . 1 1 1
29 1 . . 1 1 1 . 1 . 1
30 1 . 1 1 1 . . 1 1 .
31 . . . 1 1 . . 1 1 .
32 1 1 1 1 1 . . 1 1 1
33 1 . . 1 1 . . 1 . 1
34 . . 1 1 . . . 1 1 .
35 1 1 1 1 1 . 1 1 . .
36 1 1 1 1 . 1 . 1 . .
37 1 1 . 1 . . . 1 . .
38 . . . 1 1 1 . 1 . .
39 1 1 . 1 1 . . 1 . 1
40 1 . . 1 . . 1 1 . 1
41 1 . . 1 1 1 1 1 . 1
42 1 1 1 1 . . 1 1 . .
43 1 . . 1 1 1 . 1 . .
44 1 . 1 1 . 1 . 1 . 1
45 . . . 1 . . 1 . . 1
46 . . . 1 1 . . . 1 .
47 1 1 . 1 . . 1 1 . .
48 1 . 1 1 1 . 1 1 . .
49 . . 1 1 1 1 . 1 . 1
50 . 1 . 1 1 . . 1 1 .
51 1 . 1 1 1 1 . . . .
52 1 1 1 1 1 1 . 1 . .
53 . 1 1 1 . 1 . 1 1 1
54 1 . . 1 1 . . 1 1 .
55 1 1 . 1 1 1 . 1 . .
56 1 . . 1 1 . . 1 1 .
57 1 1 . 1 1 . 1 . . 1
58 . 1 . 1 . 1 . . 1 1
59 1 1 1 1 . . 1 1 1 .
60 . 1 1 1 1 1 . . 1 1
61 1 1 1 1 1 1 . 1 . .

CUSTOMER_RESPONSE

1381

1382

CUSTOMER_RESPONSE

62 1 1 . 1 1 . . 1 1 .
63 . . . 1 . . . 1 1 1
64 1 . . 1 1 1 . 1 . .
65 1 . . 1 1 1 . 1 . .
66 1 . . 1 1 1 1 1 1 .
67 1 1 . 1 1 1 . 1 1 .
68 1 1 . 1 1 1 . 1 1 .
69 1 1 . 1 1 . 1 . . .
70 . . . 1 1 1 . 1 . .
71 1 . . 1 1 . 1 . . 1
72 1 . 1 1 1 1 . . 1 .
73 1 1 . 1 . 1 . 1 1 .
74 1 1 1 1 1 1 . 1 . .
75 . 1 . 1 1 1 . . 1 .
76 1 1 . 1 1 1 . 1 1 1
77 . . . 1 1 1 . . . .
78 1 1 1 1 1 1 . 1 1 .
79 1 . . 1 1 1 . 1 1 .
80 1 1 1 1 1 . 1 1 . 1
81 1 1 . 1 1 1 1 1 1 .
82 . . . 1 1 1 1 . . .
83 1 1 . 1 1 1 . 1 1 .
84 1 . . 1 1 . . 1 1 .
85 . . . 1 . 1 . 1 . .
86 1 . . 1 1 1 . 1 1 1
87 1 1 . 1 1 1 . 1 . .
88 . . . 1 . 1 . . . .
89 1 . . 1 . 1 . . 1 1
90 1 1 . 1 1 1 . 1 . 1
91 . . . 1 1 . . . 1 .
92 1 . . 1 1 1 . 1 1 .
93 1 . . 1 1 . . 1 1 .
94 1 . . 1 1 1 1 1 . .
95 1 . . 1 . 1 1 1 1 .
96 1 . 1 1 1 1 . . 1 .
97 1 1 . 1 1 . . . 1 .
98 1 . 1 1 1 1 1 1 . .
99 1 1 . 1 1 1 1 1 1 .
100 1 . 1 1 1 . . . 1 1
101 1 . 1 1 1 1 . . . .
102 1 . . 1 1 . 1 1 . .
103 1 1 . 1 1 1 . 1 . .
104 . . . 1 1 1 . 1 1 1
105 1 . 1 1 1 . . 1 . 1
106 1 1 1 1 1 1 1 1 1 1
107 1 1 1 1 . . . 1 . 1
108 1 . . 1 . 1 1 1 . .
109 . 1 . 1 1 . . 1 1 .
110 1 . . 1 . . . . . .
111 1 . . 1 1 1 . 1 1 .
112 1 1 . 1 1 1 . . . 1
113 1 1 . 1 1 . 1 1 1 .
114 1 1 . 1 1 . . . . .
115 1 1 . 1 1 . . 1 . .

Appendix 3

Raw Data and DATA Steps

116 . 1 . 1 1 1 1 1 . .
117 . 1 . 1 1 1 . . . .
118 . 1 1 1 1 . . 1 1 .
119 . . . 1 . . . 1 . .
120 1 1 . 1 . . . . 1 .
;

DJIA
data djia;
input Year @7 HighDate date7. High @24 LowDate date7. Low;
format highdate lowdate date7.;
datalines;
1954

31DEC54

404.39

11JAN54

279.87

1955

30DEC55

488.40

17JAN55

388.20

1956

06APR56

521.05

23JAN56

462.35

1957

12JUL57

520.77

22OCT57

419.79

1958

31DEC58

583.65

25FEB58

436.89

1959

31DEC59

679.36

09FEB59

574.46

1960

05JAN60

685.47

25OCT60

568.05

1961

13DEC61

734.91

03JAN61

610.25

1962

03JAN62

726.01

26JUN62

535.76

1963

18DEC63

767.21

02JAN63

646.79

1964

18NOV64

891.71

02JAN64

768.08

1965

31DEC65

969.26

28JUN65

840.59

1966

09FEB66

995.15

07OCT66

744.32

1967

25SEP67

943.08

03JAN67

786.41

1968

03DEC68

985.21

21MAR68

825.13

1969

14MAY69

968.85

17DEC69

769.93

1970

29DEC70

842.00

06MAY70

631.16

1971

28APR71

950.82

23NOV71

797.97

1972

11DEC72 1036.27

26JAN72

889.15

1973

11JAN73 1051.70

05DEC73

788.31

1974

13MAR74

891.66

06DEC74

577.60

1975

15JUL75

881.81

02JAN75

632.04

1976

21SEP76 1014.79

02JAN76

858.71

1977

03JAN77

999.75

02NOV77

800.85

1978

08SEP78

907.74

28FEB78

742.12

1979

05OCT79

897.61

07NOV79

796.67

1980

20NOV80 1000.17

21APR80

759.13

1981

27APR81 1024.05

25SEP81

824.01

1982

27DEC82 1070.55

12AUG82

776.92

1983

29NOV83 1287.20

03JAN83 1027.04

1984

06JAN84 1286.64

24JUL84 1086.57

1985

16DEC85 1553.10

04JAN85 1184.96

1986

02DEC86 1955.57

22JAN86 1502.29

1987

25AUG87 2722.42

19OCT87 1738.74

1988

21OCT88 2183.50

20JAN88 1879.14

1989

09OCT89 2791.41

03JAN89 2144.64

1990

16JUL90 2999.75

11OCT90 2365.10

1991

31DEC91 3168.83

09JAN91 2470.30

1992

01JUN92 3413.21

09OCT92 3136.58

DJIA

1383

1384

EDUCATION

Appendix 3

1993

29DEC93 3794.33

20JAN93 3241.95

1994

31JAN94 3978.36

04APR94 3593.35

EDUCATION
data education;
input State $14. +1 Code $ DropoutRate Expenditures MathScore Region $;
label dropoutrate=Dropout Percentage - 1989
expenditures=Expenditure Per Pupil - 1989
mathscore=8th Grade Math Exam - 1990;
datalines;
Alabama

AL 22.3 3197 252 SE

Alaska

AK 35.8 7716 .

Arizona

AZ 31.2 3902 259 W

Arkansas

AR 11.5 3273 256 SE

California

CA 32.7 4121 256 W

Colorado

CO 24.7 4408 267 W

Connecticut

CT 16.8 6857 270 NE

Delaware

DE 28.5 5422 261 NE

Florida

FL 38.5 4563 255 SE

Georgia

GA 27.9 3852 258 SE

Hawaii

HI 18.3 4121 251 W

Idaho

ID 21.8 2838 272 W

Illinois

IL 21.5 4906 260 MW

Indiana

IN 13.8 4284 267 MW

Iowa

IA 13.6 4285 278 MW

Kansas

KS 17.9 4443 .

Kentucky

KY 32.7 3347 256 SE

Louisiana

LA 43.1 3317 246 SE

Maine

ME 22.5 4744 .

Maryland

MD 26.0 5758 260 NE

Massachusetts

MA 28.0 5979 .

Michigan

MI 29.3 5116 264 MW

Minnesota

MN 11.4 4755 276 MW

Mississippi

MS 39.9 2874 .

Missouri

MO 26.5 4263 .

Montana

MT 15.0 4293 280 W

Nebraska

NE 13.9 4360 276 MW

Nevada

NV 28.1 3791 .

New Hampshire

NH 25.9 4807 273 NE

New Jersey

NE 20.4 7549 269 NE

New Mexico

NM 28.5 3473 256 W

New York

NY 35.0 .

261 NE

North Carolina NC 31.2 3874 250 SE

North Dakota

ND 12.1 3952 281 MW

Ohio

OH 24.4 4649 264 MW

Raw Data and DATA Steps

EMPDATA
data empdata;
input IdNumber $ 1-4 LastName $ 9-19 FirstName $ 20-29
City $ 30-42 State $ 43-44 /
Gender $ 1 JobCode $ 9-11 Salary 20-29 @30 Birth date7.
@43 Hired date7. HomePhone $ 54-65;
format birth hired date7.;
datalines;
1919

Adams

Gerald

Stamford

TA2

34376

15SEP48

07JUN75

1653

Alexander

Susan

Bridgeport

ME2

35108

18OCT52

12AUG78

1400

Apple

Troy

New York

ME1

29769

08NOV55

19OCT78

1350

Arthur

Barbara

New York

FA3

32886

03SEP53

01AUG78

1401

Avery

Jerry

Paterson

TA3

38822

16DEC38

20NOV73

1499

Barefoot

Joseph

Princeton

ME3

43025

29APR42

10JUN68

1101

Baucom

Walter

New York

SCP

18723

09JUN50

04OCT78

1333

Blair

Justin

Stamford

PT2

88606

02APR49

13FEB69

1402

Blalock

Ralph

New York

TA2

32615

20JAN51

05DEC78

1479

Bostic

Marie

New York

TA3

38785

25DEC56

08OCT77

1403

Bowden

Earl

Bridgeport

ME1

28072

31JAN57

24DEC79

1739

Boyce

Jonathan

New York

PT1

66517

28DEC52

30JAN79

1658

Bradley

Jeremy

New York

SCP

17943

11APR55

03MAR80

1428

Brady

Christine Stamford

PT1

68767

07APR58

19NOV79

1782

Brown

Jason

Stamford

ME2

35345

07DEC58

25FEB80

1244

Bryant

Leonard

New York

ME2

36925

03SEP51

20JAN76

1383

Burnette

Thomas

New York

BCK

25823

28JAN56

23OCT80

1574

Cahill

Marshall

New York

FA2

28572

30APR48

23DEC80

1789

Caraway

Davis

New York

SCP

18326

28JAN45

14APR66

1404

Carter

Donald

New York

PT2

91376

27FEB41

04JAN68

1437

Carter

Dorothy

Bridgeport

33104

23SEP48

03SEP72

1639

Carter

Karen

Stamford

203/781-1255

203/675-7715

212/586-0808

718/383-1549

201/732-8787

201/812-5665

212/586-8060

203/781-1777

718/384-2849

718/384-8816

203/675-3434

212/587-1247

212/587-3622

203/781-1212

203/781-0019

718/383-3334

718/384-3569

718/383-2338

212/587-9000

718/384-2946

203/675-4117

EMPDATA

1385

1386

EMPDATA

Appendix 3

40260

29JUN45

31JAN72

1269

Caston

Franklin

Stamford

NA1

41690

06MAY60

01DEC80

1065

Chapman

Neil

New York

ME2

35090

29JAN32

10JAN75

1876

Chin

Jack

New York

TA3

39675

23MAY46

30APR73

1037

Chow

Jane

Stamford

TA1

28558

13APR52

16SEP80

1129

Cook

Brenda

New York

ME2

34929

11DEC49

20AUG79

1988

Cooper

Anthony

New York

FA3

32217

03DEC47

21SEP72

1405

Davidson

Jason

Paterson

SCP

18056

08MAR54

29JAN80

1430

Dean

Sandra

Bridgeport

TA2

32925

03MAR50

30APR75

1983

Dean

Sharon

New York

FA3

33419

03MAR50

30APR75

1134

Delgado

Maria

Stamford

TA2

33462

08MAR57

24DEC76

1118

Dennis

Roger

New York

PT3

111379

19JAN32

21DEC68

1438

Donaldson

Karen

Stamford

TA3

39223

18MAR53

21NOV75

1125

Dunlap

Donna

New York

FA2

28888

11NOV56

14DEC75

1475

Eaton

Alicia

New York

FA2

27787

18DEC49

16JUL78

1117

Edgerton

Joshua

New York

TA3

39771

08JUN51

16AUG80

1935

Fernandez

Katrina

Bridgeport

NA2

51081

31MAR42

19OCT69

1124

Fields

Diana

White Plains NY

FA1

23177

13JUL46

04OCT78

1422

Fletcher

Marie

Princeton

FA1

22454

07JUN52

09APR79

1616

Flowers

Annette

New York

TA2

34137

04MAR58

07JUN81

1406

Foster

Gerald

Bridgeport

ME2

35185

11MAR49

20FEB75

1120

Garcia

Jack

New York

ME1

28619

14SEP60

10OCT81

1094

Gomez

Alan

Bridgeport

FA1

22268

05APR58

20APR79

1389

Gordon

Levi

New York

BCK

25028

18JUL47

21AUG78

1905

Graham

Alvin

New York

PT1

65111

19APR60

01JUN80

1407

Grant

Daniel

Mt. Vernon

PT1

68096

26MAR57

21MAR78

1114

Green

Janice

New York

TA2

32928

21SEP57

30JUN75

203/781-8839

203/781-3335

718/384-5618

212/588-5634

203/781-8868

718/383-2313

212/587-1228

201/732-2323

203/675-1647

718/384-1647

203/781-1528

718/383-1122

203/781-2229

718/383-2094

718/383-2828

212/588-1239

203/675-2962

914/455-2998

201/812-0902

718/384-3329

203/675-6363

718/384-4930

203/675-7181

718/384-9326

212/586-8815

914/468-1616

212/588-1092

Raw Data and DATA Steps

ENERGY

data energy;
length State $2;
input Region Division state $ Type Expenditures;
datalines;
1 1 ME 1 708
1 1 ME 2 379
1 1 NH 1 597
1 1 NH 2 301
1 1 VT 1 353
1 1 VT 2 188
1 1 MA 1 3264
1 1 MA 2 2498
1 1 RI 1 531
1 1 RI 2 358
1 1 CT 1 2024
1 1 CT 2 1405
1 2 NY 1 8786
1 2 NY 2 7825
1 2 NJ 1 4115
1 2 NJ 2 3558
1 2 PA 1 6478
1 2 PA 2 3695
4 3 MT 1 322
4 3 MT 2 232
4 3 ID 1 392
4 3 ID 2 298
4 3 WY 1 194
4 3 WY 2 184
4 3 CO 1 1215
4 3 CO 2 1173
4 3 NM 1 545
4 3 NM 2 578
4 3 AZ 1 1694
4 3 AZ 2 1448
4 3 UT 1 621
4 3 UT 2 438
4 3 NV 1 493
4 3 NV 2 378
4 4 WA 1 1680
4 4 WA 2 1122
4 4 OR 1 1014
4 4 OR 2 756
4 4 CA 1 10643
4 4 CA 2 10114
4 4 AK 1 349
4 4 AK 2 329
4 4 HI 1 273
4 4 HI 2 298
;

ENERGY

1387

1388

GROC

Appendix 3

GROC
data groc;
input Region $9. Manager $ Department $ Sales;
datalines;
Southeast

Hayes

Paper

250

Southeast

Hayes

Produce

100

Southeast

Hayes

Canned

120

Southeast

Hayes

Meat

Southeast

Michaels

Paper

Southeast

Michaels

Produce

300

Southeast

Michaels

Canned

220

Southeast

Michaels

Meat

Northwest

Jeffreys

Paper

Northwest

Jeffreys

Produce

600

Northwest

Jeffreys

Canned

420

Northwest

Jeffreys

Meat

Northwest

Duncan

Paper

Northwest

Duncan

Produce

250

Northwest

Duncan

Canned

230

Northwest

Duncan

Meat

Northwest

Aikmann

Paper

Northwest

Aikmann

Produce

205

Northwest

Aikmann

Canned

420

Northwest

Aikmann

Meat

Southwest

Royster

Paper

Southwest

Royster

Produce

130

Southwest

Royster

Canned

120

Southwest

Royster

Meat

Southwest

Patel

Paper

Southwest

Patel

Produce

350

Southwest

Patel

Canned

225

Southwest

Patel

Meat

Northeast

Rice

Paper

Northeast

Rice

Produce

Northeast

Rice

Canned

Northeast

Rice

Meat

Northeast

Fuller

Paper

200

Northeast

Fuller

Produce

300

Northeast

Fuller

Canned

420

Northeast

Fuller

Meat

125

70
60

30
45

76
53

50
40

420
86

MATCH_11
data match_11;
input Pair Low Age Lwt Race Smoke Ptd Ht UI @@;
select(race);
when (1) do;

Raw Data and DATA Steps

race1=0;
race2=0;
end;
when (2) do;
race1=1;
race2=0;
end;
when (3) do;
race1=0;
race2=1;
end;
end;
datalines;
1

0 14 135 1 0 0 0 0

1 14 101 3 1 1 0 0

0 15

98 2 0 0 0 0

1 15 115 3 0 0 0 1

0 16

95 3 0 0 0 0

1 16 130 3 0 0 0 0

0 17 103 3 0 0 0 0

1 17 130 3 1 1 0 1

0 17 122 1 1 0 0 0

1 17 110 1 1 0 0 0

0 17 113 2 0 0 0 0

1 17 120 1 1 0 0 0

0 17 113 2 0 0 0 0

1 17 120 2 0 0 0 0

0 17 119 3 0 0 0 0

1 17 142 2 0 0 1 0

0 18 100 1 1 0 0 0

1 18 148 3 0 0 0 0

10 0 18

90 1 1 0 0 1

10 1 18 110 2 1 1 0 0

11 0 19 150 3 0 0 0 0

11 1 19

12 0 19 115 3 0 0 0 0

12 1 19 102 1 0 0 0 0

13 0 19 235 1 1 0 1 0

13 1 19 112 1 1 0 0 1

14 0 20 120 3 0 0 0 1

14 1 20 150 1 1 0 0 0

15 0 20 103 3 0 0 0 0

15 1 20 125 3 0 0 0 1

16 0 20 169 3 0 1 0 1

16 1 20 120 2 1 0 0 0

17 0 20 141 1 0 1 0 1

17 1 20

18 0 20 121 2 1 0 0 0

18 1 20 109 3 0 0 0 0

19 0 20 127 3 0 0 0 0

19 1 20 121 1 1 1 0 1

20 0 20 120 3 0 0 0 0

20 1 20 122 2 1 0 0 0

21 0 20 158 1 0 0 0 0

21 1 20 105 3 0 0 0 0

22 0 21 108 1 1 0 0 1

22 1 21 165 1 1 0 1 0

23 0 21 124 3 0 0 0 0

23 1 21 200 2 0 0 0 0

24 0 21 185 2 1 0 0 0

24 1 21 103 3 0 0 0 0

25 0 21 160 1 0 0 0 0

25 1 21 100 3 0 1 0 0

26 0 21 115 1 0 0 0 0

26 1 21 130 1 1 0 1 0

27 0 22

91 1 1 1 0 1

80 3 1 0 0 1

95 3 0 0 1 0

27 1 22 130 1 1 0 0 0

28 0 22 158 2 0 1 0 0

28 1 22 130 1 1 1 0 1

29 0 23 130 2 0 0 0 0

29 1 23

30 0 23 128 3 0 0 0 0

30 1 23 187 2 1 0 0 0

31 0 23 119 3 0 0 0 0

31 1 23 120 3 0 0 0 0

32 0 23 115 3 1 0 0 0

32 1 23 110 1 1 1 0 0

33 0 23 190 1 0 0 0 0

33 1 23

34 0 24

90 1 1 1 0 0

34 1 24 128 2 0 1 0 0

35 0 24 115 1 0 0 0 0

35 1 24 132 3 0 0 1 0

36 0 24 110 3 0 0 0 0

36 1 24 155 1 1 1 0 0

37 0 24 115 3 0 0 0 0

37 1 24 138 1 0 0 0 0

38 0 24 110 3 0 1 0 0

38 1 24 105 2 1 0 0 0

39 0 25 118 1 1 0 0 0

39 1 25 105 3 0 1 1 0

40 0 25 120 3 0 0 0 1

40 1 25

41 0 25 155 1 0 0 0 0

41 1 25 115 3 0 0 0 0

97 3 0 0 0 1

94 3 1 0 0 0

85 3 0 0 0 1

MATCH_11

1389

1390

PROCLIB.DELAY

Appendix 3

42 0 25 125 2 0 0 0 0

42 1 25

92 1 1 0 0 0

43 0 25 140 1 0 0 0 0

43 1 25

89 3 0 1 0 0

44 0 25 241 2 0 0 1 0

44 1 25 105 3 0 1 0 0

45 0 26 113 1 1 0 0 0

45 1 26 117 1 1 1 0 0

46 0 26 168 2 1 0 0 0

46 1 26

47 0 26 133 3 1 1 0 0

47 1 26 154 3 0 1 1 0

48 0 26 160 3 0 0 0 0

48 1 26 190 1 1 0 0 0

49 0 27 124 1 1 0 0 0

49 1 27 130 2 0 0 0 1

50 0 28 120 3 0 0 0 0

50 1 28 120 3 1 1 0 1

51 0 28 130 3 0 0 0 0

51 1 28

52 0 29 135 1 0 0 0 0

52 1 29 130 1 0 0 0 1

53 0 30

96 3 0 0 0 0

95 1 1 0 0 0

53 1 30 142 1 1 1 0 0

54 0 31 215 1 1 0 0 0

54 1 31 102 1 1 1 0 0

55 0 32 121 3 0 0 0 0

55 1 32 105 1 1 0 0 0

56 0 34 170 1 0 1 0 0

56 1 34 187 2 1 0 1 0

PROCLIB.DELAY
data proclib.delay;
input flight $3. +5 date date7. +2 orig $3. +3 dest $3. +3
delaycat $15. +2 destype $15. +8 delay;
informat date date7.;
format date date7.;
datalines;
114

01MAR94

LGA

LAX

1-10 Minutes

Domestic

202

01MAR94

LGA

ORD

No Delay

Domestic

-5

219

01MAR94

LGA

LON

11+ Minutes

International

622

01MAR94

LGA

FRA

No Delay

International

-5

132

01MAR94

LGA

YYZ

11+ Minutes

International

271

01MAR94

LGA

PAR

1-10 Minutes

International

302

01MAR94

LGA

WAS

No Delay

Domestic

114

02MAR94

LGA

LAX

No Delay

Domestic

202

02MAR94

LGA

ORD

1-10 Minutes

Domestic

219

02MAR94

LGA

LON

11+ Minutes

International

622

02MAR94

LGA

FRA

No Delay

International

132

02MAR94

LGA

YYZ

1-10 Minutes

International

271

02MAR94

LGA

PAR

1-10 Minutes

International

302

02MAR94

LGA

WAS

No Delay

Domestic

114

03MAR94

LGA

LAX

No Delay

Domestic

-1

202

03MAR94

LGA

ORD

No Delay

Domestic

-1

219

03MAR94

LGA

LON

1-10 Minutes

International

622

03MAR94

LGA

FRA

No Delay

International

-2

132

03MAR94

LGA

YYZ

1-10 Minutes

International

271

03MAR94

LGA

PAR

1-10 Minutes

International

302

03MAR94

LGA

WAS

1-10 Minutes

Domestic

114

04MAR94

LGA

LAX

11+ Minutes

Domestic

202

04MAR94

LGA

ORD

No Delay

Domestic

-5

219

04MAR94

LGA

LON

1-10 Minutes

International

622

04MAR94

LGA

FRA

11+ Minutes

International

132

04MAR94

LGA

YYZ

No Delay

International

-5

271

04MAR94

LGA

PAR

1-10 Minutes

International

5
-2

Raw Data and DATA Steps

302

04MAR94

LGA

WAS

1-10 Minutes

Domestic

114

05MAR94

LGA

LAX

No Delay

Domestic

-2

202

05MAR94

LGA

ORD

1-10 Minutes

Domestic

219

05MAR94

LGA

LON

1-10 Minutes

International

622

05MAR94

LGA

FRA

No Delay

International

-6

132

05MAR94

LGA

YYZ

1-10 Minutes

International

271

05MAR94

LGA

PAR

1-10 Minutes

International

114

06MAR94

LGA

LAX

No Delay

Domestic

-1

202

06MAR94

LGA

ORD

No Delay

Domestic

-3

219

06MAR94

LGA

LON

11+ Minutes

International

132

06MAR94

LGA

YYZ

1-10 Minutes

International

302

06MAR94

LGA

WAS

1-10 Minutes

Domestic

114

07MAR94

LGA

LAX

No Delay

Domestic

-1

202

07MAR94

LGA

ORD

No Delay

Domestic

-2

219

07MAR94

LGA

LON

11+ Minutes

International

622

07MAR94

LGA

FRA

11+ Minutes

International

132

07MAR94

LGA

YYZ

No Delay

International

-2

271

07MAR94

LGA

PAR

1-10 Minutes

International

302

07MAR94

LGA

WAS

No Delay

Domestic

PROCLIB.EMP95
data proclib.emp95;
input #1 idnum $4. @6 name $15.
#2 address $42.
#3 salary 6.;
datalines;
2388 James Schmidt
100 Apt. C Blount St. SW Raleigh NC 27693
92100
2457 Fred Williams
99 West Lane

Garner NC 27509

33190
2776 Robert Jones
12988 Wellington Farms Ave. Cary NC 27512
29025
8699 Jerry Capalleti
222 West L St. Oxford NC 27587
39985
2100 Lanny Engles
293 Manning Pl. Raleigh NC 27606
30998
9857 Kathy Krupski
1000 Taft Ave. Morrisville NC 27508
38756
0987 Dolly Lunford
2344 Persimmons Branch

Apex NC 27505

44010
3286 Hoa Nguyen
2818 Long St. Cary NC 27513
87734

PROCLIB.EMP95

1391

1392

PROCLIB.EMP96

Appendix 3

6579 Bryan Samosky

3887 Charles Ave. Garner NC 27508
50234
3888 Kim Siu
5662 Magnolia Blvd Southeast Cary NC 27513
77558
;

PROCLIB.EMP96
data proclib.emp96;
input #1 idnum $4. @6 name $15.
#2 address $42.
#3 salary 6.;
datalines;
2388 James Schmidt
100 Apt. C Blount St. SW Raleigh NC 27693
92100
2457 Fred Williams
99 West Lane

Garner NC 27509

33190
2776 Robert Jones
12988 Wellington Farms Ave. Cary NC 27511
29025
8699 Jerry Capalleti
222 West L St. Oxford NC 27587
39985
3278 Mary Cravens
211 N. Cypress St. Cary NC 27512
35362
2100 Lanny Engles
293 Manning Pl. Raleigh NC 27606
30998
9857 Kathy Krupski
100 Taft Ave. Morrisville NC 27508
40456
0987 Dolly Lunford
2344 Persimmons Branch Trail Apex NC 27505
45110
3286 Hoa Nguyen
2818 Long St. Cary NC 27513
89834
6579 Bryan Samosky
3887 Charles Ave. Garner NC 27508
50234
3888 Kim Siu
5662 Magnolia Blvd Southwest Cary NC 27513
79958
6544 Roger Monday

Raw Data and DATA Steps

3004 Crepe Myrtle Court Raleigh NC 27604

47007
;

PROCLIB.INTERNAT
data proclib.internat;
input flight $3.

+5 date date7. +2 dest $3. +8 boarded;

informat date date7.;

format date date7.;
datalines;
219

01MAR94

LON

198

622

01MAR94

FRA

207

132

01MAR94

YYZ

115

271

01MAR94

PAR

138

219

02MAR94

LON

147

622

02MAR94

FRA

176

132

02MAR94

YYZ

106

271

02MAR94

PAR

172

219

03MAR94

LON

197

622

03MAR94

FRA

180

132

03MAR94

YYZ

271

03MAR94

PAR

147

219

04MAR94

LON

232

622

04MAR94

FRA

137

132

04MAR94

YYZ

117

271

04MAR94

PAR

146

219

05MAR94

LON

160

622

05MAR94

FRA

185

132

05MAR94

YYZ

157

271

05MAR94

PAR

177

219

06MAR94

LON

163

132

06MAR94

YYZ

150

219

07MAR94

LON

241

622

07MAR94

FRA

210

132

07MAR94

YYZ

164

271

07MAR94

PAR

155

PROCLIB.LAKES
data proclib.lakes;
input region $ 1-2 lake $ 5-13 pol_a1 pol_a2 pol_b1-pol_b4;
datalines;
NE

Carr

0.24

0.99

0.95

0.36

0.44

0.67

Duraleigh

0.34

0.01

0.48

0.58

0.12

0.56

Charlie

0.40

0.48

0.29

0.56

0.52

0.95

Farmer

0.60

0.65

0.25

0.20

0.30

0.64

Canyon

0.63

0.44

0.20

0.98

0.19

0.01

Morris

0.85

0.95

0.80

0.67

0.32

0.81

PROCLIB.LAKES

1393

1394

PROCLIB.MARCH

Appendix 3

Golf

0.69

0.37

0.08

0.72

0.71

0.32

Falls

0.01

0.02

0.59

0.58

0.67

0.02

Pleasant

0.16

0.96

0.71

0.35

0.48

Juliette

0.82

0.35

0.09

0.03

0.59

0.90

Massey

1.01

0.77

0.45

0.32

0.55

0.66

Delta

0.84

1.05

0.90

0.09

0.64

0.03

Alumni

0.45

0.32

0.45

0.44

0.55

0.12

New Dam

0.80

0.70

0.31

0.98

1.00

0.22

Border

0.51

0.04

0.55

0.35

0.45

0.78

Red

0.22

0.09

0.02

0.10

0.32

0.01

PROCLIB.MARCH
data proclib.march;
input flight $3. +5 date date7. +3 depart time5. +2 orig $3.
+3 dest $3.

+7 miles +6 boarded +6 capacity;

format date date7. depart time5.;

informat date date7. depart time5.;
datalines;
114

01MAR94

7:10

LGA

LAX

2475

172

210

202

01MAR94

10:43

LGA

ORD

740

151

210

219

01MAR94

9:31

LGA

LON

3442

198

250

622

01MAR94

12:19

LGA

FRA

3857

207

250

132

01MAR94

15:35

LGA

YYZ

366

115

178

271

01MAR94

13:17

LGA

PAR

3635

138

250

302

01MAR94

20:22

LGA

WAS

229

105

180

114

02MAR94

7:10

LGA

LAX

2475

119

210

202

02MAR94

10:43

LGA

ORD

740

120

210

219

02MAR94

9:31

LGA

LON

3442

147

250

622

02MAR94

12:19

LGA

FRA

3857

176

250

132

02MAR94

15:35

LGA

YYZ

366

106

178

302

02MAR94

20:22

LGA

WAS

229

180

271

02MAR94

13:17

LGA

PAR

3635

104

250

114

03MAR94

7:10

LGA

LAX

2475

197

210

202

03MAR94

10:43

LGA

ORD

740

118

210

219

03MAR94

9:31

LGA

LON

3442

197

250

622

03MAR94

12:19

LGA

FRA

3857

180

250

132

03MAR94

15:35

LGA

YYZ

366

178

271

03MAR94

13:17

LGA

PAR

3635

147

250

302

03MAR94

20:22

LGA

WAS

229

123

180

114

04MAR94

7:10

LGA

LAX

2475

178

210

202

04MAR94

10:43

LGA

ORD

740

148

210

219

04MAR94

9:31

LGA

LON

3442

232

250

622

04MAR94

12:19

LGA

FRA

3857

137

250

132

04MAR94

15:35

LGA

YYZ

366

117

178

271

04MAR94

13:17

LGA

PAR

3635

146

250

302

04MAR94

20:22

LGA

WAS

229

115

180

114

05MAR94

7:10

LGA

LAX

2475

117

210

202

05MAR94

10:43

LGA

ORD

740

104

210

219

05MAR94

9:31

LGA

LON

3442

160

250

622

05MAR94

12:19

LGA

FRA

3857

185

250

Raw Data and DATA Steps

132

05MAR94

15:35

LGA

YYZ

366

157

05MAR94

13:17

LGA

PAR

3635

177

06MAR94

7:10

LGA

LAX

2475

128

210

202

06MAR94

10:43

LGA

ORD

740

115

210

219

06MAR94

9:31

LGA

LON

3442

163

250

132

06MAR94

15:35

LGA

YYZ

366

150

178

302

06MAR94

20:22

LGA

WAS

229

180

114

07MAR94

7:10

LGA

LAX

2475

160

210

202

07MAR94

10:43

LGA

ORD

740

175

210

219

07MAR94

9:31

LGA

LON

3442

241

250

622

07MAR94

12:19

LGA

FRA

3857

210

250

132

07MAR94

15:35

LGA

YYZ

366

164

178

271

07MAR94

13:17

LGA

PAR

3635

155

250

302

07MAR94

20:22

LGA

WAS

229

135

1395

250

114

PROCLIB.PAYROLL

178

271

180

PROCLIB.PAYLIST2
proc sql;
create table proclib.paylist2
(IdNum char(4),
Gender char(1),
Jobcode char(3),
Salary num,
Birth num informat=date7.
format=date7.,
Hired num informat=date7.
format=date7.);

insert into proclib.paylist2

values(1919,M,TA2,34376,12SEP66d,04JUN87d)
values(1653,F,ME2,31896,15OCT64d,09AUG92d)
values(1350,F,FA3,36886,31AUG55d,29JUL91d)
values(1401,M,TA3,38822,13DEC55d,17NOV93d)
values(1499,M,ME1,23025,26APR74d,07JUN92d);

title PROCLIB.PAYLIST2 Table;

select * from proclib.paylist2;

PROCLIB.PAYROLL
This data set (table) is updated in Example 3 on page 1129 and its updated data is
used in subsequent examples.
data proclib.payroll;
input IdNumber $4. +3 Gender $1. +4 Jobcode $3. +9 Salary 5.
+2 Birth date7. +2 Hired date7.;
informat birth date7. hired date7.;
format birth date7. hired date7.;
datalines;
1919

TA2

34376

12SEP60

04JUN87

1396

PROCLIB.PAYROLL

Appendix 3

1653

ME2

35108

15OCT64

09AUG90

1400

ME1

29769

05NOV67

16OCT90

1350

FA3

32886

31AUG65

29JUL90

1401

TA3

38822

13DEC50

17NOV85

1499

ME3

43025

26APR54

07JUN80

1101

SCP

18723

06JUN62

01OCT90

1333

PT2

88606

30MAR61

10FEB81

1402

TA2

32615

17JAN63

02DEC90

1479

TA3

38785

22DEC68

05OCT89

1403

ME1

28072

28JAN69

21DEC91

1739

PT1

66517

25DEC64

27JAN91

1658

SCP

17943

08APR67

29FEB92

1428

PT1

68767

04APR60

16NOV91

1782

ME2

35345

04DEC70

22FEB92

1244

ME2

36925

31AUG63

17JAN88

1383

BCK

25823

25JAN68

20OCT92

1574

FA2

28572

27APR60

20DEC92

1789

SCP

18326

25JAN57

11APR78

1404

PT2

91376

24FEB53

01JAN80

1437

FA3

33104

20SEP60

31AUG84

1639

TA3

40260

26JUN57

28JAN84

1269

NA1

41690

03MAY72

28NOV92

1065

ME2

35090

26JAN44

07JAN87

1876

TA3

39675

20MAY58

27APR85

1037

TA1

28558

10APR64

13SEP92

1129

ME2

34929

08DEC61

17AUG91

1988

FA3

32217

30NOV59

18SEP84

1405

SCP

18056

05MAR66

26JAN92

1430

TA2

32925

28FEB62

27APR87

1983

FA3

33419

28FEB62

27APR87

1134

TA2

33462

05MAR69

21DEC88

1118

PT3

111379

16JAN44

18DEC80

1438

TA3

39223

15MAR65

18NOV87

1125

FA2

28888

08NOV68

11DEC87

1475

FA2

27787

15DEC61

13JUL90

1117

TA3

39771

05JUN63

13AUG92

1935

NA2

51081

28MAR54

16OCT81

1124

FA1

23177

10JUL58

01OCT90

1422

FA1

22454

04JUN64

06APR91

1616

TA2

34137

01MAR70

04JUN93

1406

ME2

35185

08MAR61

17FEB87

1120

ME1

28619

11SEP72

07OCT93

1094

FA1

22268

02APR70

17APR91

1389

BCK

25028

15JUL59

18AUG90

1905

PT1

65111

16APR72

29MAY92

1407

PT1

68096

23MAR69

18MAR90

1114

TA2

32928

18SEP69

27JUN87

1410

PT2

84685

03MAY67

07NOV86

1439

PT1

70736

06MAR64

10SEP90

1409

ME3

41551

19APR50

22OCT81

1408

TA2

34138

29MAR60

14OCT87

1121

ME1

29112

26SEP71

07DEC91

1991

TA1

27645

07MAY72

12DEC92

1102

TA2

34542

01OCT59

15APR91

Raw Data and DATA Steps

1356

ME2

36869

26SEP57

22FEB83

1545

PT1

66130

12AUG59

29MAY90

1292

ME2

36691

28OCT64

02JUL89

1440

ME2

35757

27SEP62

09APR91

1368

FA2

27808

11JUN61

03NOV84

1369

TA2

33705

28DEC61

13MAR87

1411

FA2

27265

27MAY61

01DEC89

1113

FA1

22367

15JAN68

17OCT91

1704

BCK

25465

30AUG66

28JUN87

1900

ME2

35105

25MAY62

27OCT87

1126

TA3

40899

28MAY63

21NOV80

1677

BCK

26007

05NOV63

27MAR89

1441

FA2

27158

19NOV69

23MAR91

1421

TA2

33155

08JAN59

28FEB90

1119

TA1

26924

20JUN62

06SEP88

1834

BCK

26896

08FEB72

02JUL92

1777

PT3

109630

23SEP51

21JUN81

1663

BCK

26452

11JAN67

11AUG91

1106

PT2

89632

06NOV57

16AUG84

1103

FA1

23738

16FEB68

23JUL92

1477

FA2

28566

21MAR64

07MAR88

1476

TA2

34803

30MAY66

17MAR87

1379

ME3

42264

08AUG61

10JUN84

1104

SCP

17946

25APR63

10JUN91

1009

TA1

28880

02MAR59

26MAR92

1412

ME1

27799

18JUN56

05DEC91

1115

FA3

32699

22AUG60

29FEB80

1128

TA2

32777

23MAY65

20OCT90

1442

PT2

84536

05SEP66

12APR88

1417

NA2

52270

27JUN64

07MAR89

1478

PT2

84203

09AUG59

24OCT90

1673

BCK

25477

27FEB70

15JUL91

1839

NA1

43433

29NOV70

03JUL93

1347

TA3

40079

21SEP67

06SEP84

1423

ME2

35773

14MAY68

19AUG90

1200

ME1

27816

10JAN71

14AUG92

1970

FA1

22615

25SEP64

12MAR91

1521

ME3

41526

12APR63

13JUL88

1354

SCP

18335

29MAY71

16JUN92

1424

FA2

28978

04AUG69

11DEC89

1132

FA1

22413

30MAY72

22OCT93

1845

BCK

25996

20NOV59

22MAR80

1556

PT1

71349

22JUN64

11DEC91

1413

FA2

27435

16SEP65

02JAN90

1123

TA1

28407

31OCT72

05DEC92

1907

TA2

33329

15NOV60

06JUL87

1436

TA2

34475

11JUN64

12MAR87

1385

ME3

43900

16JAN62

01APR86

1432

ME2

35327

03NOV61

10FEB85

1111

NA1

40586

14JUL73

31OCT92

1116

FA1

22862

28SEP69

21MAR91

1352

NA2

53798

02DEC60

16OCT86

1555

FA2

27499

16MAR68

04JUL92

1038

TA1

26533

09NOV69

23NOV91

PROCLIB.PAYROLL

1397

1398

PROCLIB.PAYROLL2

Appendix 3

1420

ME3

43071

19FEB65

22JUL87

1561

TA2

34514

30NOV63

07OCT87

1434

FA2

28622

11JUL62

28OCT90

1414

FA1

23644

24MAR72

12APR92

1112

TA1

26905

29NOV64

07DEC92

1390

FA2

27761

19FEB65

23JUN91

1332

NA1

42178

17SEP70

04JUN91

1890

PT2

91908

20JUL51

25NOV79

1429

TA1

27939

28FEB60

07AUG92

1107

PT2

89977

09JUN54

10FEB79

1908

TA2

32995

10DEC69

23APR90

1830

PT2

84471

27MAY57

29JAN83

1882

ME3

41538

10JUL57

21NOV78

1050

ME2

35167

14JUL63

24AUG86

1425

FA1

23979

28DEC71

28FEB93

1928

PT2

89858

16SEP54

13JUL90

1480

TA3

39583

03SEP57

25MAR81

1100

BCK

25004

01DEC60

07MAY88

1995

ME1

28810

24AUG73

19SEP93

1135

FA2

27321

20SEP60

31MAR90

1415

FA2

28278

09MAR58

12FEB88

1076

PT1

66558

14OCT55

03OCT91

1426

TA2

32991

05DEC66

25JUN90

1564

SCP

18833

12APR62

01JUL92

1221

FA2

27896

22SEP67

04OCT91

1133

TA1

27701

13JUL66

12FEB92

1435

TA3

38808

12MAY59

08FEB80

1418

ME1

28005

29MAR57

06JAN92

1017

TA3

40858

28DEC57

16OCT81

1443

NA1

42274

17NOV68

29AUG91

1131

TA2

32575

26DEC71

19APR91

1427

TA2

34046

31OCT70

30JAN90

1036

TA3

39392

19MAY65

23OCT84

1130

FA1

23916

16MAY71

05JUN92

1127

TA2

33011

09NOV64

07DEC86

1433

FA3

32982

08JUL66

17JAN87

1431

FA3

33230

09JUN64

05APR88

1122

FA2

27956

01MAY63

27NOV88

1105

ME2

34805

01MAR62

13AUG90

PROCLIB.PAYROLL2
data proclib.payroll2;
input idnum $4. +3 gender $1. +4 jobcode $3. +9 salary 5.
+2 birth date7. +2 hired date7.;
informat birth date7. hired date7.;
format birth date7. hired date7.;
datalines;
1639

TA3

42260

26JUN57

28JAN84

1065

ME3

38090

26JAN44

07JAN87

1561

TA3

36514

30NOV63

07OCT87

Raw Data and DATA Steps

1221

FA3

29896

22SEP67

04OCT91

1447

FA1

22123

07AUG72

29OCT92

1998

SCP

23100

10SEP70

02NOV92

1036

TA3

42465

19MAY65

23OCT84

1106

PT3

94039

06NOV57

16AUG84

1129

ME3

36758

08DEC61

17AUG91

1350

FA3

36098

31AUG65

29JUL90

1369

TA3

36598

28DEC61

13MAR87

1076

PT1

69742

14OCT55

03OCT91

PROCLIB.SCHEDULE
data proclib.schedule;
input flight $3. +5 date date7. +2 dest $3. +3 idnum $4.;
format date date7.;
informat date date7.;
datalines;
132

01MAR94

YYZ

1739

132

01MAR94

YYZ

1478

132

01MAR94

YYZ

1130

132

01MAR94

YYZ

1390

132

01MAR94

YYZ

1983

132

01MAR94

YYZ

1111

219

01MAR94

LON

1407

219

01MAR94

LON

1777

219

01MAR94

LON

1103

219

01MAR94

LON

1125

219

01MAR94

LON

1350

219

01MAR94

LON

1332

271

01MAR94

PAR

1439

271

01MAR94

PAR

1442

271

01MAR94

PAR

1132

271

01MAR94

PAR

1411

271

01MAR94

PAR

1988

271

01MAR94

PAR

1443

622

01MAR94

FRA

1545

622

01MAR94

FRA

1890

622

01MAR94

FRA

1116

622

01MAR94

FRA

1221

622

01MAR94

FRA

1433

622

01MAR94

FRA

1352

132

02MAR94

YYZ

1556

132

02MAR94

YYZ

1478

132

02MAR94

YYZ

1113

132

02MAR94

YYZ

1411

132

02MAR94

YYZ

1574

132

02MAR94

YYZ

1111

219

02MAR94

LON

1407

219

02MAR94

LON

1118

219

02MAR94

LON

1132

219

02MAR94

LON

1135

PROCLIB.SCHEDULE

1399

1400

PROCLIB.SCHEDULE

219

02MAR94

LON

1441

219

02MAR94

LON

1332

271

02MAR94

PAR

1739

271

02MAR94

PAR

1442

271

02MAR94

PAR

1103

271

02MAR94

PAR

1413

271

02MAR94

PAR

1115

271

02MAR94

PAR

1443

622

02MAR94

FRA

1439

622

02MAR94

FRA

1890

622

02MAR94

FRA

1124

622

02MAR94

FRA

1368

622

02MAR94

FRA

1477

622

02MAR94

FRA

1352

132

03MAR94

YYZ

1739

132

03MAR94

YYZ

1928

132

03MAR94

YYZ

1425

132

03MAR94

YYZ

1135

132

03MAR94

YYZ

1437

132

03MAR94

YYZ

1111

219

03MAR94

LON

1428

219

03MAR94

LON

1442

219

03MAR94

LON

1130

219

03MAR94

LON

1411

219

03MAR94

LON

1115

219

03MAR94

LON

1332

271

03MAR94

PAR

1905

271

03MAR94

PAR

1118

271

03MAR94

PAR

1970

271

03MAR94

PAR

1125

271

03MAR94

PAR

1983

271

03MAR94

PAR

1443

622

03MAR94

FRA

1545

622

03MAR94

FRA

1830

622

03MAR94

FRA

1414

622

03MAR94

FRA

1368

622

03MAR94

FRA

1431

622

03MAR94

FRA

1352

132

04MAR94

YYZ

1428

132

04MAR94

YYZ

1118

132

04MAR94

YYZ

1103

132

04MAR94

YYZ

1390

132

04MAR94

YYZ

1350

132

04MAR94

YYZ

1111

219

04MAR94

LON

1739

219

04MAR94

LON

1478

219

04MAR94

LON

1130

219

04MAR94

LON

1125

219

04MAR94

LON

1983

219

04MAR94

LON

1332

271

04MAR94

PAR

1407

271

04MAR94

PAR

1410

271

04MAR94

PAR

1094

271

04MAR94

PAR

1411

Appendix 3

Raw Data and DATA Steps

271

04MAR94

PAR

1115

271

04MAR94

PAR

1443

622

04MAR94

FRA

1545

622

04MAR94

FRA

1890

622

04MAR94

FRA

1116

622

04MAR94

FRA

1221

622

04MAR94

FRA

1433

622

04MAR94

FRA

1352

132

05MAR94

YYZ

1556

132

05MAR94

YYZ

1890

132

05MAR94

YYZ

1113

132

05MAR94

YYZ

1475

132

05MAR94

YYZ

1431

132

05MAR94

YYZ

1111

219

05MAR94

LON

1428

219

05MAR94

LON

1442

219

05MAR94

LON

1422

219

05MAR94

LON

1413

219

05MAR94

LON

1574

219

05MAR94

LON

1332

271

05MAR94

PAR

1739

271

05MAR94

PAR

1928

271

05MAR94

PAR

1103

271

05MAR94

PAR

1477

271

05MAR94

PAR

1433

271

05MAR94

PAR

1443

622

05MAR94

FRA

1545

622

05MAR94

FRA

1830

622

05MAR94

FRA

1970

622

05MAR94

FRA

1441

622

05MAR94

FRA

1350

622

05MAR94

FRA

1352

132

06MAR94

YYZ

1333

132

06MAR94

YYZ

1890

132

06MAR94

YYZ

1414

132

06MAR94

YYZ

1475

132

06MAR94

YYZ

1437

132

06MAR94

YYZ

1111

219

06MAR94

LON

1106

219

06MAR94

LON

1118

219

06MAR94

LON

1425

219

06MAR94

LON

1434

219

06MAR94

LON

1555

219

06MAR94

LON

1332

132

07MAR94

YYZ

1407

132

07MAR94

YYZ

1118

132

07MAR94

YYZ

1094

132

07MAR94

YYZ

1555

132

07MAR94

YYZ

1350

132

07MAR94

YYZ

1111

219

07MAR94

LON

1905

219

07MAR94

LON

1478

219

07MAR94

LON

1124

219

07MAR94

LON

1434

PROCLIB.SCHEDULE

1401

1402

PROCLIB.STAFF

219

07MAR94

LON

1983

219

07MAR94

LON

1332

271

07MAR94

PAR

1410

271

07MAR94

PAR

1777

271

07MAR94

PAR

1103

271

07MAR94

PAR

1574

271

07MAR94

PAR

1115

271

07MAR94

PAR

1443

622

07MAR94

FRA

1107

622

07MAR94

FRA

1890

622

07MAR94

FRA

1425

622

07MAR94

FRA

1475

622

07MAR94

FRA

1433

622

07MAR94

FRA

Appendix 3

1352

PROCLIB.STAFF
data proclib.staff;
input idnum $4. +3 lname $15. +2 fname $15. +2 city $15. +2
state $2. +5 hphone $12.;
datalines;
1919

ADAMS

GERALD

STAMFORD

203/781-1255

1653

ALIBRANDI

MARIA

BRIDGEPORT

203/675-7715

1400

ALHERTANI

ABDULLAH

NEW YORK

212/586-0808

1350

ALVAREZ

MERCEDES

NEW YORK

718/383-1549

1401

ALVAREZ

CARLOS

PATERSON

201/732-8787

1499

BAREFOOT

JOSEPH

PRINCETON

201/812-5665

1101

BAUCOM

WALTER

NEW YORK

212/586-8060

1333

BANADYGA

JUSTIN

STAMFORD

203/781-1777

1402

BLALOCK

RALPH

NEW YORK

718/384-2849

1479

BALLETTI

MARIE

NEW YORK

718/384-8816

1403

BOWDEN

EARL

BRIDGEPORT

203/675-3434

1739

BRANCACCIO

JOSEPH

NEW YORK

212/587-1247

1658

BREUHAUS

JEREMY

NEW YORK

212/587-3622

1428

BRADY

CHRISTINE

STAMFORD

203/781-1212

1782

BREWCZAK

JAKOB

STAMFORD

203/781-0019

1244

BUCCI

ANTHONY

NEW YORK

718/383-3334

1383

BURNETTE

THOMAS

NEW YORK

718/384-3569

1574

CAHILL

MARSHALL

NEW YORK

718/383-2338

1789

CARAWAY

DAVIS

NEW YORK

212/587-9000

1404

COHEN

LEE

NEW YORK

718/384-2946

1437

CARTER

DOROTHY

BRIDGEPORT

203/675-4117

1639

CARTER-COHEN

KAREN

STAMFORD

203/781-8839

1269

CASTON

FRANKLIN

STAMFORD

203/781-3335

1065

COPAS

FREDERICO

NEW YORK

718/384-5618

1876

CHIN

JACK

NEW YORK

212/588-5634

1037

CHOW

JANE

STAMFORD

203/781-8868

1129

COUNIHAN

BRENDA

NEW YORK

718/383-2313

1988

COOPER

ANTHONY

NEW YORK

212/587-1228

1405

DACKO

JASON

PATERSON

201/732-2323

1430

DABROWSKI

SANDRA

BRIDGEPORT

203/675-1647

Raw Data and DATA Steps

1983

DEAN

SHARON

NEW YORK

718/384-1647

1134

DELGADO

MARIA

STAMFORD

203/781-1528

1118

DENNIS

ROGER

NEW YORK

718/383-1122

1438

DABBOUSSI

KAMILLA

STAMFORD

203/781-2229

1125

DUNLAP

DONNA

NEW YORK

718/383-2094

1475

ELGES

MARGARETE

NEW YORK

718/383-2828

1117

EDGERTON

JOSHUA

NEW YORK

212/588-1239

1935

FERNANDEZ

KATRINA

BRIDGEPORT

203/675-2962

1124

FIELDS

DIANA

WHITE PLAINS

914/455-2998

1422

FUJIHARA

KYOKO

PRINCETON

201/812-0902

1616

FUENTAS

CARLA

NEW YORK

718/384-3329

1406

FOSTER

GERALD

BRIDGEPORT

203/675-6363

1120

GARCIA

JACK

NEW YORK

718/384-4930

1094

GOMEZ

ALAN

BRIDGEPORT

203/675-7181

1389

GOLDSTEIN

LEVI

NEW YORK

718/384-9326

1905

GRAHAM

ALVIN

NEW YORK

212/586-8815

1407

GREGORSKI

DANIEL

MT. VERNON

914/468-1616

1114

GREENWALD

JANICE

NEW YORK

212/588-1092

1410

HARRIS

CHARLES

STAMFORD

203/781-0937

1439

HASENHAUER

CHRISTINA

BRIDGEPORT

203/675-4987

1409

HAVELKA

RAYMOND

STAMFORD

203/781-9697

1408

HENDERSON

WILLIAM

PRINCETON

201/812-4789

1121

HERNANDEZ

ROBERTO

NEW YORK

718/384-3313

1991

HOWARD

GRETCHEN

BRIDGEPORT

203/675-0007

1102

HERMANN

JOACHIM

WHITE PLAINS

914/455-0976

1356

HOWARD

MICHAEL

NEW YORK

212/586-8411

1545

HERRERO

CLYDE

STAMFORD

203/781-1119

1292

HUNTER

HELEN

BRIDGEPORT

203/675-4830

1440

JACKSON

LAURA

STAMFORD

203/781-0088

1368

JEPSEN

RONALD

STAMFORD

203/781-8413

1369

JONSON

ANTHONY

NEW YORK

212/587-5385

1411

JOHNSEN

JACK

PATERSON

201/732-3678

1113

JOHNSON

LESLIE

NEW YORK

718/383-3003

1704

JONES

NATHAN

NEW YORK

718/384-0049

1900

KING

WILLIAM

NEW YORK

718/383-3698

1126

KIMANI

ANNE

NEW YORK

212/586-1229

1677

KRAMER

JACKSON

BRIDGEPORT

203/675-7432

1441

LAWRENCE

KATHY

PRINCETON

201/812-3337

1421

LEE

RUSSELL

MT. VERNON

914/468-9143

1119

JEFF

NEW YORK

212/586-2344

1834

LEBLANC

RUSSELL

NEW YORK

718/384-0040

1777

LUFKIN

ROY

NEW YORK

718/383-4413

1663

MARKS

JOHN

NEW YORK

212/587-7742

1106

MARSHBURN

JASPER

STAMFORD

203/781-1457

1103

MCDANIEL

RONDA

NEW YORK

212/586-0013

1477

MEYERS

PRESTON

BRIDGEPORT

203/675-8125

1476

MONROE

JOYCE

STAMFORD

203/781-2837

1379

MORGAN

ALFRED

STAMFORD

203/781-2216

1104

MORGAN

CHRISTOPHER

NEW YORK

718/383-9740

1009

MORGAN

GEORGE

NEW YORK

212/586-7753

1412

MURPHEY

JOHN

PRINCETON

201/812-4414

1115

MURPHY

ALICE

NEW YORK

718/384-1982

1128

NELSON

FELICIA

BRIDGEPORT

203/675-1166

1442

NEWKIRK

SANDRA

PRINCETON

201/812-3331

PROCLIB.STAFF

1403

1404

PROCLIB.STAFF

Appendix 3

1417

NEWKIRK

WILLIAM

PATERSON

201/732-6611

1478

NEWTON

JAMES

NEW YORK

212/587-5549

1673

NICHOLLS

HENRY

STAMFORD

203/781-7770

1839

NORRIS

DIANE

NEW YORK

718/384-1767

1347

ONEAL

BRYAN

NEW YORK

718/384-0230

1423

OSWALD

LESLIE

MT. VERNON

914/468-9171

1200

OVERMAN

MICHELLE

STAMFORD

203/781-1835

1970

PARKER

ANNE

NEW YORK

718/383-3895

1521

PARKER

JAY

NEW YORK

212/587-7603

1354

PARKER

MARY

WHITE PLAINS

914/455-2337

1424

PATTERSON

RENEE

NEW YORK

212/587-8991

1132

PEARCE

CAROL

NEW YORK

718/384-1986

1845

PEARSON

JAMES

NEW YORK

718/384-2311

1556

PENNINGTON

MICHAEL

NEW YORK

718/383-5681

1413

PETERS

RANDALL

PRINCETON

201/812-2478

1123

PETERSON

SUZANNE

NEW YORK

718/383-0077

1907

PHELPS

WILLIAM

STAMFORD

203/781-1118

1436

PORTER

SUSAN

NEW YORK

718/383-5777

1385

RAYNOR

MILTON

BRIDGEPORT

203/675-2846

1432

REED

MARILYN

MT. VERNON

914/468-5454

1111

RHODES

JEREMY

PRINCETON

201/812-1837

1116

RICHARDS

CASEY

NEW YORK

212/587-1224

1352

RIVERS

SIMON

NEW YORK

718/383-3345

1555

RODRIGUEZ

JULIA

BRIDGEPORT

203/675-2401

1038

RODRIGUEZ

MARIA

BRIDGEPORT

203/675-2048

1420

ROUSE

JEREMY

PATERSON

201/732-9834

1561

SANDERS

RAYMOND

NEW YORK

212/588-6615

1434

SANDERSON

EDITH

STAMFORD

203/781-1333

1414

SANDERSON

NATHAN

BRIDGEPORT

203/675-1715

1112

SANYERS

RANDY

NEW YORK

718/384-4895

1390

SMART

JONATHAN

NEW YORK

718/383-1141

1332

STEPHENSON

ADAM

BRIDGEPORT

203/675-1497

1890

STEPHENSON

ROBERT

NEW YORK

718/384-9874

1429

THOMPSON

ALICE

STAMFORD

203/781-3857

1107

THOMPSON

WAYNE

NEW YORK

718/384-3785

1908

TRENTON

MELISSA

NEW YORK

212/586-6262

1830

TRIPP

KATHY

BRIDGEPORT

203/675-2479

1882

TUCKER

ALAN

NEW YORK

718/384-0216

1050

TUTTLE

THOMAS

WHITE PLAINS

914/455-2119

1425

UNDERWOOD

JENNY

STAMFORD

203/781-0978

1928

UPCHURCH

LARRY

WHITE PLAINS

914/455-5009

1480

UPDIKE

THERESA

NEW YORK

212/587-8729

1100

VANDEUSEN

RICHARD

NEW YORK

212/586-2531

1995

VARNER

ELIZABETH

NEW YORK

718/384-7113

1135

VEGA

ANNA

NEW YORK

718/384-5913

1415

VEGA

FRANKLIN

NEW YORK

718/384-2823

1076

VENTER

RANDALL

NEW YORK

718/383-2321

1426

VICK

THERESA

PRINCETON

201/812-2424

1564

WALTERS

ANNE

NEW YORK

212/587-3257

1221

WALTERS

DIANE

NEW YORK

718/384-1918

1133

WANG

CHIN

NEW YORK

212/587-1956

1435

WARD

ELAINE

NEW YORK

718/383-4987

1418

WATSON

BERNARD

NEW YORK

718/383-1298

1017

WELCH

DARIUS

NEW YORK

212/586-5535

Raw Data and DATA Steps

1443

WELLS

AGNES

STAMFORD

1131

WELLS

NADINE

NEW YORK

WHALEY

CAROLYN

MT. VERNON

WONG

LESLIE

NEW YORK

212/587-2570

1130

WOOD

DEBORAH

NEW YORK

212/587-0013

1127

WOOD

SANDRA

NEW YORK

212/587-2881

1433

YANCEY

ROBIN

PRINCETON

201/812-1874

1431

YOUNG

DEBORAH

STAMFORD

203/781-2987

1122

YOUNG

JOANN

NEW YORK

718/384-2021

1105

YOUNG

LAWRENCE

NEW YORK

1405

914/468-4528

1036

RADIO

718/383-1045

1427

203/781-5546

718/384-0008

PROCLIB.SUPERV

data proclib.superv;
input supid $4. +8 state $2. +5

jobcat

$2.;

label supid=Supervisor Id jobcat=Job Category;

datalines;
1677

1834

1431

1433

1983

1385

1420

1882

1935

1417

1352

1106

1442

1118

1405

1564

1639

1401

5 4 1 5 1 0 0 1 1 0 0 0 0 0
782 38 m 30 8 5
7 5 5 5 5 0 0 4 4 4 4 4 0 0
222 33 m 3 3 1
1 1 1 1 1 1 1 1 4 1 1 1 1 1
467 24 f 2 4 1
0 0 1 0 1 0 0 0 1 1 1 0 0 0
467 34 f 1 1 1
1 0 0 0 1 0 0 1 1 0 0 0 0 0
781 53 f 2 1 5
5 0 0 0 5 5 0 0 0 0 5 5 5 0
222 30 m 2 5 3
6 3 3 3 6 0 0 0 3 3 3 3 0 0
688 26 f 2 2 1
1 0 0 0 1 0 0 0 1 0 1 1 0 0
222 29 m 8 5 1
1 6 0 6 1 0 0 1 1 1 1 0 0 0
783 33 m 1 2 7
7 0 0 0 7 0 0 0 7 0 0 0 7 0
781 39 m 1.5 2.5 2
2 0 2 0 2 0 0 0 2 2 2 0 0 0
850 22 f 2 1 1
1 0 0 0 1 1 1 0 5 0 0 1 0 0
493 36 f 1 0 5
0 0 0 0 7 0 0 0 0 0 0 0 0 0
967 46 f 2 4 7
7 5 0 5 7 0 0 0 4 7 4 0 0 0
856 41 m 2 2 4
7 4 0 0 7 4 0 4 0 0 0 7 0 0
546 25 m 5 5 8
8 8 0 0 0 0 0 0 0 0 0 0 0 0
222 27 f 4 4 3
2 2 2 3 7 7 0 2 2 2 3 3 3 0
688 23 m 9 3 3
3 3 3 3 3 7 0 0 3 0 0 0 0 0
849 26 m .5 .5 8
8 0 0 0 8 0 0 0 0 8 0 0 0 0
783 29 f 3 3 1
1 0 0 0 4 0 0 4 1 0 1 0 0 0
856 34 f 1.5 2 1
7 0 0 0 7 0 0 7 4 0 0 7 0 0
966 33 m 3 5 4
7 0 0 0 7 4 5 0 7 0 0 7 4 4
493 34 f 2 5 1
1 0 0 0 1 0 0 0 7 0 1 1 8 0
467 29 m 2 4 2
2 0 0 0 2 0 0 2 2 2 2 2 2 2
677 28 f 1 4 1
1 1 1 1 1 0 0 0 1 0 1 0 0 0
781 27 m 2 2 1
1 0 1 0 4 2 4 0 2 2 1 0 1 4
467 24 m 4 4 1
7 1 0 1 1 1 0 7 1 0 0 0 0 0
859 26 m 5 5 1

RADIO

1417

1418

RADIO

Appendix 3

1 1 1 1 1 1 1 1 1 1 1 1 1 1
848 27 m 7 2 5
7 5 0 5 4 5 0 0 0 7 4 4 0 4
677 25 f 1 2 8
8 0 0 0 0 5 0 0 8 0 0 0 2 0
222 26 f 3.5 0 2
2 0 0 0 2 0 0 0 0 0 0 0 0 0
833 32 m 1 2 1
1 0 0 0 1 0 0 0 5 0 1 0 0 0
781 28 m 2 .5 7
7 0 0 0 7 0 0 0 4 0 0 0 0 0
783 28 f 1 1 1
1 0 0 0 1 0 0 0 0 0 1 1 0 0
222 28 f 5 5 2
2 6 6 2 2 0 0 0 2 2 0 0 2 2
851 33 m 4 5 3
1 0 0 0 7 3 0 3 3 3 3 3 7 5
859 39 m 2 1 1
1 0 0 0 1 0 0 0 0 0 0 1 0 0
848 45 m 2 2 7
7 0 0 0 7 0 0 0 7 0 0 0 0 0
467 37 m 2 2 7
7 0 0 0 0 7 0 0 0 7 0 0 7 0
859 32 m .25 .25 1
1 0 0 0 0 0 0 0 1 0 0 0 0 0

1419

APPENDIX

4
Recommended Reading
Recommended Reading

1419

Recommended Reading
Here is the recommended reading list for this title:
3 The Little SAS Book: A Primer, Third Edition
3 Output Delivery System: The Basics

3
3
3
3
3
3
3
3

PROC TABULATE by Example

SAS Guide to Report Writing: Examples
SAS Language Reference: Concepts
SAS Language Reference: Dictionary
SAS Output Delivery System: Users Guide
SAS Programming by Example
SAS SQL Procedure Users Guide
Step-by-Step Programming with Base SAS Software

For a complete list of SAS publications, see the current SAS Publishing Catalog. To
order the most current publications or to receive a free copy of the catalog, contact a
SAS representative at
SAS Publishing Sales
SAS Campus Drive
Cary, NC 27513
Telephone: (800) 727-3228*
Fax: (919) 677-8166
E-mail: [email protected]
Web address: support.sas.com/pubs
* For other SAS Institute business, call (919) 677-8000.
Customers outside the United States should contact their local SAS ofce.

1420

Index

Index
A
ACCELERATE= option
ITEM statement (PMENU) 671
ACROSS option
DEFINE statement (REPORT) 899
across variables 854, 899
activities data set 86, 107
AFTER= option
PROC CPORT statement 288
AGE statement
DATASETS procedure 312
aging data sets 388
aging les 312
ALL class variable 1246
ALL keyword 1095
ALLOBS option
PROC COMPARE statement 231
ALLSTATS option
PROC COMPARE statement 231
ALLVARS option
PROC COMPARE statement 231
ALPHA= option
PROC MEANS statement 528
PROC TABULATE statement 1188
ALTER= option
AGE statement (DATASETS) 312
CHANGE statement (DATASETS) 322
COPY statement (DATASETS) 328
DELETE statement (DATASETS) 334
EXCHANGE statement (DATASETS) 338
MODIFY statement (DATASETS) 349
PROC DATASETS statement 309
REPAIR statement (DATASETS) 353
SELECT statement (DATASETS) 356
ALTER TABLE statement
SQL procedure 1038
alternative hypotheses
ANALYSIS option
DEFINE statement (REPORT) 899
analysis variables 548, 855, 899
SUMMARY procedure 1179
TABULATE procedure 1211, 1212
weights for 64, 912, 1212
ANSI Standard 1122
APPEND 75
APPEND procedure 75
overview 75
syntax 75

APPEND statement
DATASETS procedure 314
appending data sets 314
APPEND procedure vs. APPEND statement 319
compressed data sets 316
indexed data sets 316
integrity constraints and 318
password-protected data sets 316
restricting observations 315
SET statement vs. APPEND statement 315
system failures 319
variables with different attributes 317
with different variables 317
with generation groups 318
APPENDVER= option
APPEND statement (DATASETS) 314
arithmetic mean
arithmetic operators 1123
ASCENDING option
CHART procedure 192
CLASS statement (MEANS) 536
CLASS statement (TABULATE) 1197
ASCII option
PROC SORT statement 1007
ASCII order 1007, 1014
ASIS option
PROC CPORT statement 288
asterisk (*) notation 1059
ATTR= option
TEXT statement (PMENU) 679
ATTRIB statement
procedures and 57
audit les
creating 321
event logging 320
AUDIT statement
DATASETS procedure 320
AUDIT_ALL= option
AUDIT statement (DATASETS) 320
AUTOLABEL option
OUTPUT statement (MEANS) 545
AUTONAME option
OUTPUT statement (MEANS) 545
axes
customizing
AXIS= option
CHART procedure 193
PLOT statement (TIMEPLOT) 1294

B
bar charts 180
horizontal 180, 189, 209
maximum number of bars 187
percentage charts 201
side-by-side 206
vertical 180, 191, 203
BASE= argument
APPEND statement (DATASETS) 314
base data set 226
BASE= option
PROC COMPARE statement 232
batch mode
creating printer denitions 789
printing from 867
BATCH option
PROC DISPLAY statement 396
BETWEEN condition 1071
block charts 181, 187
for BY groups 210
BLOCK statement
CHART procedure 187
BOX option
PLOT statement (PLOT) 615
PROC REPORT statement 872
TABLE statement (TABULATE) 1204
_BREAK_ automatic variable 862
break lines 861
_BREAK_ automatic variable 862
creating 861
order of 861, 888, 911
BREAK statement
REPORT procedure 885
BREAK window
REPORT procedure 913
breaks 861
BRIEFSUMMARY option
PROC COMPARE statement 232
browsing external les 489
BTRIM function (SQL) 1072
buttons 670
BY-group information
titles containing 20
BY-group processing 20, 59
error processing for 24
formats and 30
TABULATE procedure 1215
BY groups
block charts for 210

1421

1422

Index

plotting 646
transposing
BY lines
inserting into titles 23
suppressing the default 20
BY processing
COMPARE procedure 237
BY statement 58
BY-group processing 59
CALENDAR procedure 91
CHART procedure 188
COMPARE procedure 236
example 60
formatting BY-variable values 59
MEANS procedure 535, 562
options 58
PLOT procedure 612
PRINT procedure 716
procedures supporting 60
RANK procedure 819
REPORT procedure 889
SORT procedure 1012
STANDARD procedure 1168
TABULATE procedure 1196
TIMEPLOT procedure 1291
TRANSPOSE procedure
BY variables
formatting values 59
inserting names into titles 22
inserting values into titles 21

C
calculated columns
SQL 1073
CALCULATED component 1073
CALEDATA= option
PROC CALENDAR statement 86
CALENDAR 84
calendar, dened 104
calendar data set 86, 109
multiple calendars 105, 106
CALENDAR procedure 84
activities data set 107
activity lines 113
calendar data set 109
calendar types 79, 101
concepts 101
customizing calendar appearance 113
default calendars 103
duration 94
examples 114
holiday duration 95
holidays data set 108
input data sets 106
missing values 111
multiple calendars 79, 92, 104
ODS portability 113
output, format of 112
output, quantity of 112
overview 79
project management 83
results 112
schedule calendars 102
scheduling 83, 137
summary calendars 102

syntax 84
task tables 84, 85
workdays data set 110
calendar reports 104
CALID statement
CALENDAR procedure 92
CALL DEFINE statement
REPORT procedure 890
CAPS option
PROC FSLIST statement 491
Cartesian product 1083, 1084
case-control studies 1158
CASE expression 1073
CATALOG 154
CATALOG= argument
PROC DISPLAY statement 396
catalog concatenation 168
catalog entries
copying 159, 165, 170
deleting 156, 160, 170, 176
displaying contents of 174
excluding, for copying 162
exporting 296, 298
importing 223
modifying descriptions of 163, 174
moving, from multiple catalogs 170
renaming 157, 174
routing log or output to entries 779
saving from deletion 164
switching names of 161
CATALOG= option
CONTENTS statement (CATALOG) 158
PROC PMENU statement 667
CATALOG procedure 154
catalog concatenation 168
concepts 165
ending a step 166
entry type specication 166
error handling 166
examples 170
interactive processing with RUN groups 165
overview 153
results 169
syntax 154
task tables 154, 155, 159
catalogs
concatenating 168
exporting multiple 295
format catalogs 456
listing contents of 157
PMENU entries 667, 673, 680
repairing 353
categories 1185
headings for 1227
categories of procedures 3
CC option
FSLIST command 493
PROC FSLIST statement 491
CENTER option
DEFINE statement (REPORT) 899
PROC REPORT statement 873
centiles 344
CENTILES option
CONTENTS statement (DATASETS) 324
CFREQ option
CHART procedure 193

CHANGE statement
CATALOG procedure 157
DATASETS procedure 322
character data
converting to numeric values 470
character strings
converting to lowercase 1093
converting to uppercase 1114
formats for 448
ranges for 480
returning a substring 1106
trimming 1072
character values
formats for 466
character variables
sorting orders for 1014
CHART 185
CHART procedure 185
bar charts 180, 206
block charts 181, 187, 210
concepts 197
customizing charts 191
examples 198
formatting characters 185
frequency counts 198
horizontal bar charts 189, 209
missing values 195, 197
ODS output 198
ODS table names 198
options 192
overview 179
percentage bar charts 201
pie charts 182, 189
results 197
star charts 183, 190
syntax 185
task table 191
variable characteristics 197
vertical bar charts 191, 203
charts
bar charts 180, 201, 206
block charts 181, 187, 210
customizing 191
horizontal bar charts 189, 209
missing values 195
pie charts 182, 189
star charts 183, 190
vertical bar charts 191, 203
CHARTYPE option
PROC MEANS statement 528
check boxes 668, 670
active vs. inactive 668
color of 668
CHECKBOX statement
PMENU procedure 668
CIMPORT 216
CIMPORT procedure 216
examples 222
le transport process 215
overview 215
results 221
syntax 216
task table 216
CLASS statement
MEANS procedure 536
TABULATE procedure 1197
TIMEPLOT procedure 1292

Index

class variables 536

BY statement (MEANS) with 562
CLASSDATA= option (MEANS) with 564
combinations of 546, 549, 1235
computing descriptive statistics 560
formatting in TABULATE 1214
level value headings 1201
MEANS procedure 550
missing 1224, 1225, 1226
missing values 539, 578, 1200
multilabel value formats with 567
ordering values 551
preloaded formats with 570, 1237
TABULATE procedure 1197
TIMEPLOT procedure 1292
CLASSDATA= option
PROC MEANS statement 528, 564
PROC TABULATE statement 1188
classifying formatted data 27
CLASSLEV statement
TABULATE procedure 1201
CLEARSASUSER option
PROC REGISTRY statement 832
client/server model
exporting data 410
CLM keyword
CLONE option
COPY statement (DATASETS) 328
CNTLIN= option
PROC FORMAT statement 433
CNTLOUT= option
PROC FORMAT statement 433, 434, 448
COALESCE function (SQL) 1075
coefcient of variation
collating sequence 1007
collision states 628
COLOR= option
BREAK statement (REPORT) 886
CHECKBOX statement (PMENU) 668
DEFINE statement (REPORT) 899
RBREAK statement (REPORT) 909
RBUTTON statement (PMENU) 676
TEXT statement (PMENU) 679
column aliases 1059
column attributes 1041, 1077
reports 890
column-denition component 1076
column-header option
DEFINE statement (REPORT) 900
column headings
customizing 1244
customizing text in 727
page layout 722
column-modier component 1077
column modiers 1123
column-name component 1078
COLUMN statement
REPORT procedure 893
column width 723
columns
aliases 1059
altering 1038
calculated 1073
combinations of values 1154
for each variable value 960
in reports 893
indexes on 1042, 1043, 1057

inserting values 1057

length of 1077
modiers 1123
renaming 1042
returning values 1075
selecting 1058, 1078
SQL procedure 1029
storing values of 1060
updating values 1069
COLWIDTH= option
PROC REPORT statement 873
COMMAND option
PROC REPORT statement 873
COMMIT statement (SQL) 1124
COMPARE 229
COMPARE= option
PROC COMPARE statement 232
COMPARE procedure 229
BY processing 237
comparing selected variables 240
comparing unsorted data 238
comparing variables 240
comparisons with 240
concepts 240
customizing output 227
differences report 256
duplicate ID values 238
equality criterion 242
examples 256
ID variables 238, 241, 266
information provided by 226
listing variables for matching observations 238
log and 244
macro return codes 244
ODS table names 253
output 246
output data set 254
output statistics data set 255
overview 226
position of observations 240
restricting comparisons 239
results 244
syntax 229
task tables 229, 230
variable formats 244
COMPAREREG1 option
PROC REGISTRY statement 832
COMPAREREG2 option
PROC REGISTRY statement 833
COMPARETO= option
PROC REGISTRY statement 833
comparison data set 226
COMPLETECOLS option
PROC REPORT statement 874
COMPLETEROWS option
PROC REPORT statement 874
COMPLETETYPES option
PROC MEANS statement 529
composite indexes 1044
compound names 862
compressed data sets
appending 316
compute blocks 858
contents of 859
processing 860
referencing report items in 859

1423

starting 895
COMPUTE statement
REPORT procedure 895
COMPUTE window
REPORT procedure 916
COMPUTED option
DEFINE statement (REPORT) 901
COMPUTED VAR window
REPORT procedure 916
computed variables 855, 901
storing 983
concatenating catalogs 168
concatenating data sets 386
CONDENSE option
TABLE statement (TABULATE) 1204
condence limits 553, 573
keywords and formulas
one-sided, above the mean
one-sided, below the mean
TABULATE procedure 1188
two-sided
CONNECT statement
SQL procedure 1042
CONNECTION TO component 1079
CONSTRAINT= option
COPY statement (DATASETS) 330
PROC CPORT statement 288
CONTAINS condition 1080
CONTENTS= option
PROC PRINT statement 707
PROC REPORT statement 874
PROC TABULATE statement 1189
TABLE statement (TABULATE) 1205
CONTENTS procedure 276
overview 275
syntax 276
task table 276
versus CONTENTS statement
(DATASETS) 327
CONTENTS statement
CATALOG procedure 157
DATASETS procedure 323
contingency tables 1269
continuation messages 1185
CONTOUR= option
PLOT statement (PLOT) 615
contour plots 615, 642
converting les 215, 285
COPY 278
COPY procedure 278
concepts 278
example 279
overview 277
syntax 278
transporting data sets 278
versus COPY statement (DATASETS) 333
COPY statement
CATALOG procedure 159
DATASETS procedure 327
TRANSPOSE procedure
copying data libraries
entire data library 331
copying data sets
between hosts 279
long variable names 332
copying les 327
COPY statement vs. COPY procedure 333

1424

Index

excluding les 339

member type 331
password-protected les 332
selected les 331, 356
corrected sum of squares
CORRECTENCODING= option
MODIFY statement (DATASETS) 349
correlated subqueries 1104
CORRESPONDING keyword 1095
COUNT(*) function 1108
CPERCENT option
CHART procedure 193
CPM procedure 83, 137
CPORT 286
CPORT procedure 286
concepts 294
Data Control Blocks 294
examples 295
le transport process 286
overview 285
password-protected data sets 294
results 294
syntax 286
task table 286
CREATE INDEX statement
SQL procedure 1043
CREATE TABLE statement
SQL procedure 1045
CREATE VIEW statement
SQL procedure 1049
CRITERION= option
PROC COMPARE statement 232, 242
cross joins 1087
crosstabulation tables 1269
CSS keyword
cumulative distribution function
customized output 51
for output objects 53
customizing charts 191
CV keyword

D
DANISH option
PROC SORT statement 1007
DATA= argument
PROC EXPORT statement 404
DATA COLUMNS window
REPORT procedure 917
data components
denition 40
Data Control Blocks (DCBs) 221, 294
data libraries
copying entire library 331
copying les 327
deleting les 334
exchanging lenames 338
importing 222
printing directories of 275, 323
processing all data sets in 30
renaming les 322
saving les from deletion 355
USER data library 17
DATA= option
APPEND statement (DATASETS) 314
CONTENTS statement (DATASETS) 324

PROC CALENDAR statement 86

PROC CHART statement 185
PROC COMPARE statement 232
PROC MEANS statement 529
PROC OPTLOAD statement 602
PROC PLOT statement 609
PROC PRINT statement 708
PROC PRTDEF statement 790
PROC RANK statement 816
PROC REPORT statement 874
PROC SORT statement 1008
PROC STANDARD statement 1166
PROC TABULATE statement 1189
PROC TIMEPLOT statement 1290
PROC TRANSPOSE statement
DATA SELECTION window
REPORT procedure 917
data set options 18
SQL procedure with 1114
data sets
aging 388
appending 314
appending compressed data sets 316
appending indexed data sets 316
appending password-protected data sets 316
concatenating 386
content descriptions 323
contents of 275
copying between hosts 279
creating formats from 472
describing 384
exporting 296
input data sets 19
loading system options from 601
modifying 381
naming 16
permanent 16
printing all data sets in library 767
printing formatted values for 25
processing all data sets in a library 30
renaming variables 352
repairing 353
saving system option settings in 603
standardizing variables 1163
temporary 16
transporting 278, 333
transporting password-protected 294
USER data library and 17
writing printer attributes to 806
data sets, comparing
base data set 226
comparison data set 226
comparison summary 246
variables in different data sets 261
variables in same data set 240, 264
DATA step views
SQL procedure 1029
data summaries 1108, 1246
data summarization tools 1177
DATABASE= statement
EXPORT procedure 409
IMPORT procedure 511
DATAFILE= argument
PROC IMPORT statement 502
DATAROW= statement
IMPORT procedure 507
DATASETS 308

DATASETS procedure 308

concepts 357
directory listings 363
ending 359
error handling 358
examples 376
execution of statements 357
forcing RUN-group processing 359
generation data sets 362
member types 361
ODS and 368, 389
output 305
output data sets 370
overview 304
password errors 359
passwords with 359
procedure output 364
restricting member types 360
results 363
RUN-group processing 357
syntax 308
task tables 308, 323, 348
DATATYPE= option
PICTURE statement (FORMAT) 439
date formats 468
DATECOPY option
PROC CIMPORT statement 217
PROC CPORT statement 288
PROC SORT statement 1008
COPY statement (DATASETS) 330
DATETIME option
PROC CALENDAR statement 86
DAYLENGTH= option
PROC CALENDAR statement 86
DBMS
SORT procedure with 1013
DBMS connections
ending 1054
LIBNAME statement for 1115
Pass-Through Facility for 1115
sending DBMS statements to 1055
SQL procedure 1042
storing in views 1051
DBMS= option
PROC EXPORT statement 405
PROC IMPORT statement 503
DBMS queries 1079
DBMS tables
exporting to 409
importing 511
DBPWD= statement
EXPORT procedure 409
IMPORT procedure 512
DBSASLABEL= statement
IMPORT procedure 507, 512
DCBs (Data Control Blocks) 221, 294
DDNAME= argument
PROC FSLIST statement 490
DDNAME= option
PROC DATASETS statement 311
debugging
registry debugging 833
DEBUGOFF option
PROC REGISTRY statement 833
DEBUGON option
PROC REGISTRY statement 833

Index

DECSEP= option
PICTURE statement (FORMAT) 439
DEFAULT= option
FORMAT procedure 451
RADIOBOX statement (PMENU) 675
DEFINE option
PROC OPTIONS statement 595
DEFINE statement
REPORT procedure 897
DEFINITION window
REPORT procedure 918
DELETE option
PROC PRTDEF statement 790
DELETE statement
CATALOG procedure 160
DATASETS procedure 334
SQL procedure 1052
delimited les
exporting 408, 411
importing 506, 514
DELIMITER= statement
EXPORT procedure 408
IMPORT procedure 507
denominator denitions 1269
density function
DESC option
PROC PMENU statement 667
DESCENDING option
BY statement 58
BY statement (CALENDAR) 92
BY statement (CHART) 188
BY statement (COMPARE) 237
BY statement (MEANS) 535
BY statement (PLOT) 612
BY statement (PRINT) 716
BY statement (RANK) 819
BY statement (REPORT) 889
BY statement (SORT) 1012
BY statement (STANDARD) 1168
BY statement (TABULATE) 1196
BY statement (TIMEPLOT) 1291
BY statement (TRANSPOSE)
CHART procedure 193
CLASS statement (MEANS) 536
CLASS statement (TABULATE) 1197
DEFINE statement (REPORT) 901
ID statement (COMPARE) 238
PROC RANK statement 816
DESCENDTYPES option
PROC MEANS statement 529
DESCRIBE statement
SQL procedure 1053
DESCRIPTION= argument
MODIFY statement (CATALOG) 163
descriptive statistics 558, 1177
computing with class variables 560
keywords and formulas
table of 31
destination-independent input 43
detail reports 847
detail rows 847
DETAILS option
CONTENTS statement (DATASETS) 325
PROC DATASETS statement 309
deviation from the mean
device drivers
system fonts 425

dialog boxes 669

check boxes in 668
collecting user input 685
color for 679
input elds 678
radio buttons in 676
searching multiple values 688
text for 678
DIALOG statement
PMENU procedure 669
DICTIONARY tables 1116
performance and 1118
reporting from 1136
retrieving information about 1117
uses for 1118
difference 244
report of differences 256
DIG3SEP= option
PICTURE statement (FORMAT) 439
digit selectors 442
dimension expressions 1208
elements in 1208
operators in 1209
style elements in 1210
directives 442
DIRECTORY option
CONTENTS statement (DATASETS) 325
DISCONNECT statement
SQL procedure 1054
DISCRETE option
CHART procedure 193
DISPLAY 395
DISPLAY option
DEFINE statement (REPORT) 901
DISPLAY PAGE window
REPORT procedure 923
DISPLAY procedure 395
example 396
overview 395
syntax 395
display variables 853, 901
distribution
DMOPTLOAD command 601
DMOPTSAVE command 603
DOCUMENT destination 44
denition 40
DOL option
BREAK statement (REPORT) 886
RBREAK statement (REPORT) 909
DOUBLE option
PROC PRINT statement 708
PROC SQL statement 1034
double overlining 886, 909
double underlining 886, 910
DQUOTE= option
PROC SQL statement 1034
DROP statement
SQL procedure 1054
DTC= option
MODIFY statement (DATASETS) 349
DUL option
RBREAK statement (REPORT) 886, 910
DUPOUT= option
PROC SORT statement 1008
DUR statement
CALENDAR procedure 94

1425

E
EBCDIC option
PROC SORT statement 1007
EBCDIC order 1007, 1014
EET= option
PROC CIMPORT statement 217
PROC CPORT statement 289
efciency
statistical procedures 7
elementary statistics procedures
embedded LIBNAME statements 1051
embedded SQL 1125
encoded passwords 807, 809
in SAS programs 808, 809
saving to paste buffer 811
encoding
versus encryption 808
encryption
versus encoding 808
ENDCOMP statement
REPORT procedure 906
ENTRYTYPE= option
CATALOG procedure 167
CHANGE statement (CATALOG) 157
COPY statement (CATALOG) 159
DELETE statement (CATALOG) 161
EXCHANGE statement (CATALOG) 161
EXCLUDE statement (CATALOG) 162
EXCLUDE statement (CIMPORT) 220
EXCLUDE statement (CPORT) 292
MODIFY statement (CATALOG) 164
PROC CATALOG statement 156
SAVE statement (CATALOG) 164
SELECT statement (CATALOG) 165
SELECT statement (CIMPORT) 221
SELECT statement (CPORT) 293
EQUALS option
PROC SORT statement 1008
equijoins 1083
error checking
formats and 30
error handling
CATALOG procedure 166
ERROR option
PROC COMPARE statement 232
error processing
of BY-group specications 24
ERRORSTOP option
PROC SQL statement 1035
estimates
ET= option
PROC CIMPORT statement 218
PROC CPORT statement 289
ETYPE= option
SELECT statement (CPORT) 293
event logging 320
Excel
exporting spreadsheets 407, 416
exporting subset of observations to 414
exporting to a specic spreadsheet 415
importing spreadsheet from workbook 517,
521
importing spreadsheets 505
importing subset of records from 518
loading spreadsheet into workbook 408
EXCEPT operator 1098

1426

Index

EXCHANGE statement
CATALOG procedure 161
DATASETS procedure 338
EXCLNPWGT option
PROC REPORT statement 874
PROC STANDARD statement 1167
EXCLNPWGTS option
PROC MEANS statement 529
PROC TABULATE statement 1189
EXCLUDE statement
CATALOG procedure 162
CIMPORT procedure 219
CPORT procedure 292
DATASETS procedure 339
FORMAT procedure 434
PRTEXP procedure 804
exclusion lists 52
destinations for output objects 53
EXCLUSIVE option
CLASS statement (MEANS) 536
CLASS statement (TABULATE) 1197
DEFINE statement (REPORT) 901
PROC MEANS statement 529
PROC TABULATE statement 1189
EXEC option
PROC SQL statement 1035
EXECUTE statement
SQL procedure 1055
EXISTS condition 1080
expected value
EXPLODE 403
EXPLODE procedure 403
EXPLORE window
REPORT procedure 924
EXPORT 404
EXPORT= option
PROC REGISTRY statement 834
EXPORT procedure 404
data source statements 408
DBMS specications 405
DBMS table statements 409
examples 411
overview 403
syntax 404
exporting
catalog entries 296, 298
CPORT procedure 285
excluding les or entries 292
multiple catalogs 295
printer denitions 790
registry contents 834, 840
selecting les or entries 293
exporting data 403
client/server model 410
DBMS tables 409
delimited les 408, 411
Microsoft Access 407, 410, 415
spreadsheet compatibility 407
spreadsheets 408
EXTENDSN= option
PROC CIMPORT statement 218
external les
browsing 489
comparing registry with 841
routing output or log to 776
extreme values 580, 583

F
FEEDBACK option
PROC SQL statement 1035
FILE= option
CONTENTS statement (CATALOG) 158
PROC CIMPORT statement 218
PROC CPORT statement 289
le transport process 286
FILEREF= argument
PROC FSLIST statement 490
les
aging 312
converting 215, 285
copying 277, 327
deleting 334
exchanging names 338
excluding from copying 339
modifying attributes 348
moving 331
renaming 322
renaming groups of 312
saving from deletion 355, 380
selecting for copying 356
FILL option
PROC CALENDAR statement 86
PICTURE statement (FORMAT) 440
FIN statement
CALENDAR procedure 94
FINNISH option
PROC SORT statement 1007
oating point exception (FPE) recovery 1195
FLOW option
DEFINE statement (REPORT) 901
PROC SQL statement 1035
FMTLEN option
CONTENTS statement (DATASETS) 325
FMTLIB option
PROC FORMAT statement 433, 434, 448
font les
adding 425, 426
searching directories for 421
specifying 421
TrueType 422, 427
Type 1 422
FONTFILE statement
FONTREG procedure 421
FONTPATH statement
FONTREG procedure 421
FONTREG 420
FONTREG procedure 420
concepts 423
examples 425
font naming conventions 423
overview 419
removing fonts from registry 424
SAS/GRAPH device drivers 425
supported font types 423
syntax 420
fonts
naming conventions 423
removing from registry 424
FORCE option
APPEND statement (DATASETS) 314
COPY statement (DATASETS) 330
PROC CATALOG statement 156, 176
PROC CIMPORT statement 218

PROC DATASETS statement 310

PROC SORT statement 1009
FOREIGN option
PROC PRTDEF statement 790
FORMAT 432
format catalogs 456
format-name formats 451
FORMAT= option
DEFINE statement (REPORT) 901
MEAN statement (CALENDAR) 98
PROC TABULATE statement 1189
SUM statement (CALENDAR) 100
FORMAT procedure 432
associating formats/informats with variables 455
concepts 455
examples 463
excluding entries from processing 434
input control data set 460
options 451
output control data set 458
overview 430
procedure output 461
ranges 453
results 458
selecting entries for processing 448
storing formats/informats 456
syntax 432
task tables 432, 435, 438, 448
values 453
FORMAT statement 57
DATASETS procedure 340
FORMAT_PRECEDENCE= option
TABLE statement (TABULATE) 1205
formats
See also picture formats
assigning style attribute values 866
assigning style attributes 1221
associating with variables 430, 455
BY-group processing and 30
comparing unformatted values 244
creating from data sets 472
creating groups with 986
date formats 468
denition of 430
error checking and 30
for character values 448, 466
for columns 1077
format-name formats 451
managing with DATASETS procedure 340
missing 457
multilabel 1242
multilabel value formats 567
permanent 456, 457
picture-name formats 447
preloaded 904, 1199, 1237
preloaded, with class variables 570
printing descriptions of 477
ranges for character strings 480
retrieving permanent formats 478
storing 456
temporarily associating with variables 28
temporarily dissociating from variables 29
temporary 456
FORMATS window
REPORT procedure 925

Index

formatted values 25, 59

classifying formatted data 27
grouping formatted data 27
printing 25
FORMCHAR option
PROC CHART statement 185
PROC PLOT statement 610
PROC REPORT statement 874
PROC TABULATE statement 1189
PROC CALENDAR statement 87
forms
printing reports with 867
FORMS 487
FORMS procedure 487
formulas
for statistics
FORTCC option
FSLIST command 493
PROC FSLIST statement 491
FPE recovery 1195
FRACTION option
PROC RANK statement 816
FRAME applications
associating menus with 700
FreeType fonts 419
FREQ option
CHART procedure 193
CHART procedure 193
FREQ statement 61
example 62
MEANS procedure 539
procedures supporting 62
REPORT procedure 906
STANDARD procedure 1169
TABULATE procedure 1201
frequency counts
CHART procedure 198
displaying with denominator denitions 1269
TABULATE procedure 1269
frequency of observations 61
FROM clause
SQL procedure 1063
FSEDIT applications
menu bars for 682
FSEDIT sessions
associating menu bar with 684
FSLIST 489
FSLIST command 490, 492
FSLIST procedure 489
overview 489
syntax 489
FSLIST window 494
commands 494
display commands 498
global commands 494
scrolling commands 494
searching commands 496
FULLSTATUS option
PROC REGISTRY statement 834
functional categories of procedures 3
functions
SQL procedure and 1100, 1124
FUZZ= option
FORMAT procedure 436, 452
PROC COMPARE statement 232
TABLE statement (TABULATE) 1205

FW= option
PROC MEANS statement 529

G
G100 option
CHART procedure 194
Gaussian distribution
generation data sets
DATASETS procedure and 362
generation groups
appending with 318
changing number of 352
copying 333
deleting 335
removing passwords 352
GENERATION option
PROC CPORT statement 289
GENMAX= option
MODIFY statement (DATASETS) 349
GENNUM= data set option 318
GENNUM= option
AUDIT statement (DATASETS) 320
CHANGE statement (DATASETS) 322
DELETE statement (DATASETS) 334
MODIFY statement (DATASETS) 349
PROC DATASETS statement 310
REPAIR statement (DATASETS) 353
GETDELETED= statement
IMPORT procedure 508
GETNAMES= statement
IMPORT procedure 508
Ghostview printer denition 797
global statements 18
GRAY option
ITEM statement (PMENU) 672
grayed items 672
GROUP BY clause
SQL procedure 1065
GROUP option
DEFINE statement (REPORT) 902
CHART procedure 194
PROC OPTIONS statement 595
group variables 854, 902
grouping formatted data 27
GROUPINTERNAL option
CLASS statement (MEANS) 537
CLASS statement (TABULATE) 1198
groups
creating with formats 986
GROUPS= option
PROC RANK statement 817
GSPACE= option
CHART procedure 194
GUESSING ROWS= statement
IMPORT procedure 508

H
HAVING clause
SQL procedure 1067
HAXIS= option
PLOT statement (PLOT) 616

HBAR statement
CHART procedure 189
HEADER= option
PROC CALENDAR statement 89
HEADING= option
PROC PRINT statement 708
HEADLINE option
PROC REPORT statement 876
HEADSKIP option
PROC REPORT statement 876
HELP= option
ITEM statement (PMENU) 672
PROC REPORT statement 877
HEXPAND option
PLOT statement (PLOT) 618
hidden label characters 628
hidden observations 630
HILOC option
PLOT statement (TIMEPLOT) 1296
HOLIDATA= option
PROC CALENDAR statement 89
holidays data set 89, 108
multiple calendars 105, 106
HOLIDUR statement
CALENDAR procedure 95
HOLIFIN statement
CALENDAR procedure 96
HOLISTART statement
CALENDAR procedure 96
HOLIVAR statement
CALENDAR procedure 97
horizontal bar charts 180, 189
for subset of data 209
horizontal separators 1251
HOST option
PROC OPTIONS statement 595
host-specic procedures
HPERCENT= option
PROC PLOT statement 610
HPOS= option
PLOT statement (PLOT) 618
HREF= option
PLOT statement (PLOT) 618
HREFCHAR= option
PLOT statement (PLOT) 618
HREVERSE option
PLOT statement (PLOT) 618
HSCROLL= option
PROC FSLIST statement 492
HSPACE= option
PLOT statement (PLOT) 618
HTML destination 45
HTML les
style elements 989
TABULATE procedure 1279
HTML output
sample 35
HTML reports 725
HTML version setting 50
hypotheses
keywords and formulas
testing
HZERO option
PLOT statement (PLOT) 618

1427

1428

Index

I
IC CREATE statement
DATASETS procedure 340
IC DELETE statement
DATASETS procedure 343
IC REACTIVATE statement
DATASETS procedure 343
ID option
DEFINE statement (REPORT) 902
ITEM statement (PMENU) 672
ID statement
COMPARE procedure 238
MEANS procedure 540
PRINT procedure 717
TIMEPLOT procedure 1293
TRANSPOSE procedure
ID variables 902
COMPARE procedure 238
IDLABEL statement
TRANSPOSE procedure
IDMIN option
PROC MEANS statement 530
IMPORT 502
IMPORT= option
PROC REGISTRY statement 834
IMPORT procedure 502
data source statements 506
DBMS specications 504
examples 514
overview 501
syntax 502
IMPORT statement
DBMS table statements 511
importing
catalog entries 223
CIMPORT procedure 215
data libraries 222
excluding les or entries 219
indexed data sets 224
selecting les or entries 221
to registry 834, 839
importing data 501
DBMS tables 511
delimited les 506, 514
Excel spreadsheets 505
Microsoft Access 505, 513, 519
PC les 506
spreadsheet from Excel workbook 517, 521
spreadsheets 506
subset of records from Excel 518
IN= argument
PROC PWENCODE statement 808
IN condition 1081
in-line views 1064, 1123
querying 1148
IN= option
COPY statement (CATALOG) 159
COPY statement (DATASETS) 330
INDENT= option
TABLE statement (TABULATE) 1205
indenting row headings 1251
INDEX CENTILES statement
DATASETS procedure 344
INDEX CREATE statement
DATASETS procedure 345

INDEX DELETE statement

DATASETS procedure 346
INDEX= option
COPY statement (DATASETS) 330
PROC CPORT statement 289
indexed data sets
importing 224
indexes
appending indexed data sets 316
centiles for indexed variables 344
composite indexes 1044
creating 345
deleting 346, 1054
managing 1045
on altered columns 1042
on columns 1043, 1057
simple indexes 1044
SQL procedure 1044
UNIQUE keyword 1044
INFILE= option
PROC CIMPORT statement 218
INFORMAT statement
DATASETS procedure 347
informats
associating with variables 430, 455
character data to numeric values 470
denition of 430
for columns 1077
managing with DATASETS procedure 347
missing 457
permanent 456, 457
printing descriptions of 477
raw data values 435
storing 456
temporary 456
INITIATE argument
AUDIT statement (DATASETS) 320
inner joins 1084
INOBS= option
PROC SQL statement 1036
input data sets 19
CALENDAR procedure 106
input elds 678
input les
procedure output as 782
INSERT statement
SQL procedure 1056
integrity constraints
appending data sets and 318
copying data sets and 330
creating 340
deleting 343
names for 342
password-protected les with 327
PROC SQL tables 1042, 1049
reactivating 343
SORT procedure 1015
interactive line mode
printing from 868
interquartile range
INTERSECT operator 1099
INTERVAL= option
PROC CALENDAR statement 89
INTO clause
SQL procedure 1060
INTYPE= option
PROC CPORT statement 290

INVALUE statement
FORMAT procedure 435
IS condition 1081
ITEM statement
PMENU procedure 671
ITEMHELP= option
DEFINE statement (REPORT)

902

J
joined-table component 1082
JOINREF option
PLOT statement (TIMEPLOT) 1296
joins
cross joins 1087
denition of 1083
equijoins 1083
inner joins 1084
joining a table with itself 1084
joining more than two tables 1089
joining tables 1083
joining three tables 1145
joining two tables 1131
natural joins 1088
outer joins 1086, 1123, 1138
reexive joins 1084
rows to be returned 1083
subqueries compared with 1091
table limit 1083
types of 1083
union joins 1088
JUST option
INVALUE statement (FORMAT) 436

K
KEEPLEN option
OUTPUT statement (MEANS) 545
KEY= option
PROC OPTLOAD statement 602
PROC OPTSAVE statement 604
key sequences 671
KEYLABEL statement
TABULATE procedure 1202
keyword headings
style elements for 1202
KEYWORD statement
TABULATE procedure 1202
keywords
for statistics
KILL option
PROC CATALOG statement 156, 176
PROC DATASETS statement 310
kurtosis
KURTOSIS keyword

L
LABEL option
PROC PRINT statement 708
MODIFY statement (DATASETS) 350
ODS TRACE statement 52
PROC PRINTTO statement 773

Index

PROC TRANSPOSE statement

LABEL statement 57
DATASETS procedure 348
labels
for columns 1077
hidden label characters 628
on plots 649, 654, 658
language concepts 16
data set options 18
global statements 18
system options 17
temporary and permanent data sets 16
LCLM keyword
LEFT option
DEFINE statement (REPORT) 902
LEGEND option
PROC CALENDAR statement 89
LET option
PROC TRANSPOSE statement
LEVELS option
OUTPUT statement (MEANS) 546
CHART procedure 194
LIBNAME statement
DBMS connections with 1115
embedding in views 1051
libraries
printing all data sets 767
LIBRARY= option
PROC DATASETS statement 311
PROC FORMAT statement 433
librefs
stored views and 1050
LIKE condition 1091
line-drawing characters 872
LINE statement
REPORT procedure 907
LIST option
PLOT statement (PLOT) 618
PROC PRTDEF statement 790
PROC REGISTRY statement 834
PROC REPORT statement 877
LISTALL option
PROC COMPARE statement 232
LISTBASE option
PROC COMPARE statement 233
LISTBASEOBS option
PROC COMPARE statement 233
LISTBASEVAR option
PROC COMPARE statement 233
LISTCOMP option
PROC COMPARE statement 233
LISTCOMPOBS option
PROC COMPARE statement 233
LISTCOMPVAR option
PROC COMPARE statement 233
LISTEQUALVAR option
PROC COMPARE statement 233
LISTHELP= option
PROC REGISTRY statement 835
LISTING destination 44
denition 40
Listing output
sample 33
listing reports 704, 727
LISTOBS option
PROC COMPARE statement 233

LISTREG= option
PROC REGISTRY statement 835
LISTUSER option
PROC REGISTRY statement 835
LISTVAR option
PROC COMPARE statement 233
LOAD REPORT window
REPORT procedure 925
LOCALE option
PROC CALENDAR statement 90
log
COMPARE procedure results 244
default destinations 771
destinations for 771
displaying SQL denitions 1053
listing registry contents in 835
routing to catalog entries 779
routing to external les 776
routing to printer 775, 785
writing printer attributes to 805
writing registry contents to 834
LOG option
AUDIT statement (DATASETS) 320
PROC PRINTTO statement 773
logarithmic scale for plots 639
LONG option
PROC OPTIONS statement 596
LOOPS= option
PROC SQL statement 1036
LOWER function (SQL) 1093
LPI= option
PROC CHART statement 186
LS= option
PROC REPORT statement 877

M
macro return codes
COMPARE procedure 244
macro variables
set by SQL procedure 1119
macros
adjusting plot labels 658
markers 882, 1193
MARKUP destination 45
denition 40
markup languages 40
matching observations 226
matching patterns 1091, 1158
matching variables 226
MAX keyword
MAX= option
FORMAT procedure 436, 452
MAXDEC= option
PROC MEANS statement 530
PROC TIMEPLOT statement 1290
maximum value
MAXLABLEN= option
PROC FORMAT statement 434
MAXPRINT= option
PROC COMPARE statement 233
MAXSELEN= option
PROC FORMAT statement 434
mean
MEAN keyword

1429

MEAN option
CHART procedure 194
PROC STANDARD statement 1167
MEAN statement
CALENDAR procedure 97
MEANS 526
MEANS procedure 526
class variables 550
column width for output 556
computational resources 552
computer resources 539
concepts 550
examples 558
missing values 539, 556, 578
N Obs statistic 556
output 524
output data set 557
output statistics 575, 577, 578, 580,
overview 524
results 556
statistic keywords 533, 541
statistical computations 553
syntax 526
task tables 526, 527
MEANTYPE= option
PROC CALENDAR statement 90
measures of location
measures of shape
measures of variability
median
MEDIAN keyword
MEMOSIZE= statement
IMPORT procedure 512
MEMTYPE= option
AGE statement (DATASETS) 312
CHANGE statement (DATASETS) 323
CONTENTS statement (DATASETS) 325
COPY statement (DATASETS) 330, 331
DELETE statement (DATASETS) 334
EXCHANGE statement (DATASETS) 338
EXCLUDE statement (CIMPORT) 220
EXCLUDE statement (CPORT) 292
EXCLUDE statement (DATASETS) 339
MODIFY statement (DATASETS) 350
PROC CIMPORT statement 218
PROC CPORT statement 290
PROC DATASETS statement 311
REPAIR statement (DATASETS) 354
SAVE statement (DATASETS) 355
SELECT statement (CIMPORT) 221
SELECT statement (CPORT) 293
SELECT statement (DATASETS) 356
menu bars 665
associating with FSEDIT sessions 684, 691
associating with FSEDIT window 687
dening items 673
for FSEDIT applications 682
items in 671
key sequences for 671
menu items 671
MENU statement
PMENU procedure 673
merging data
SQL procedure 1110
message characters 442
MESSAGE= option
IC CREATE statement (DATASETS) 342

1430

Index

MESSAGES window
REPORT procedure 926
METHOD= option
PROC COMPARE statement 233
PROC PWENCODE statement 808
Microsoft Access
exporting tables 415
exporting to database 407
importing tables 505, 519
security level for tables 513
security levels 410
MIDPOINTS= option
CHART procedure 194
MIGRATE 591
MIGRATE procedure 591
MIN keyword
MIN= option
FORMAT procedure 436, 452
minimum value
missing formats/informats 457
MISSING option
CHART procedure 195
CLASS statement (MEANS) 537
CLASS statement (TABULATE) 1198
DEFINE statement (REPORT) 902
PROC CALENDAR statement 90
PROC MEANS statement 530
PROC PLOT statement 611
PROC REPORT statement 878
PROC TABULATE statement 1191
missing values
CALENDAR procedure 111
charts 195, 197
for class variables 539
MEANS procedure 539, 556, 578
NMISS keyword
PLOT procedure 630, 652
RANK procedure 821
REPORT procedure 860, 977
SQL procedure 1081, 1160
STANDARD procedure 1170
TABULATE procedure 1200, 1222
TIMEPLOT procedure 1298
TRANSPOSE procedure
MISSTEXT= option
TABLE statement (TABULATE) 1205
MIXED= statement
IMPORT procedure 508
MLF option
CLASS statement (MEANS) 537
CLASS statement (TABULATE) 1198
MNEMONIC= option
ITEM statement (PMENU) 673
mode
MODE keyword
MODE= option
PROC FONTREG statement 420
MODIFY statement
CATALOG procedure 163
DATASETS procedure 348
moment statistics 553
MOVE option
COPY statement (CATALOG) 159
COPY statement (DATASETS) 330
moving les 331
MSGLEVEL= option
PROC FONTREG statement 420

MT= option
PROC CPORT statement 290
MTYPE= option
AGE statement (DATASETS) 312
EXCLUDE statement (CPORT) 292
SELECT statement (CPORT) 293
SELECT statement (DATASETS) 356
multi-threaded sorting 1013
multilabel formats 1242
MULTILABEL option
PICTURE statement (FORMAT) 440
VALUE statement (FORMAT) 449
multilabel value formats 567
multipage tables 1253
multiple-choice survey data 1260
multiple-response survey data 1255
MULTIPLIER= option
PICTURE statement (FORMAT) 441

N
N keyword
N Obs statistic 556
N option
PROC PRINT statement 709
NAME= option
PROC TRANSPOSE statement
NAMED option
PROC REPORT statement 878
naming data sets 16
NATIONAL option
PROC SORT statement 1007
natural joins 1088
NEDIT option
PROC CPORT statement 290
nested variables 1186
NEW option
COPY statement (CATALOG) 159
PROC CIMPORT statement 218
PROC PRINTTO statement 774
APPEND statement (DATASETS) 314
NMISS keyword
NOALIAS option
PROC REPORT statement 878
NOBORDER option
PROC FSLIST statement 492
NOBS keyword
NOBYLINE system option
BY statement (MEANS) with 535
BY statement (PRINT) with 716
NOCC option
FSLIST command 493
PROC FSLIST statement 491
NOCOMPRESS option
PROC CPORT statement 290
NOCONTINUED option
TABLE statement (TABULATE) 1206
NODATE option
PROC COMPARE statement 234
NODS option
CONTENTS statement (DATASETS) 326
NODUPKEY option
PROC SORT statement 1009
NODUPRECS option
PROC SORT statement 1009

NOEDIT option
COPY statement (CATALOG) 160
PICTURE statement (FORMAT) 441
PROC CIMPORT statement 218
PROC CPORT statement 290
NOEXEC option
PROC REPORT statement 878
NOHEADER option
CHART procedure 195
PROC REPORT statement 878
NOINHERIT option
OUTPUT statement (MEANS) 546
NOLEGEND option
CHART procedure 195
PROC PLOT statement 611
NOLIST option
PROC DATASETS statement 311
NOMISS option
INDEX CREATE statement
(DATASETS) 345
PROC PLOT statement 611
NOMISSBASE option
PROC COMPARE statement 234
NOMISSCOMP option
PROC COMPARE statement 234
NOMISSING option
PROC COMPARE statement 234
NONE option
RBUTTON statement (PMENU) 676
noninteractive mode
printing from 867
NONOBS option
PROC MEANS statement 530
NOOBS option
PROC PRINT statement 709
NOPRINT option
CONTENTS statement (DATASETS) 326
DEFINE statement (REPORT) 902
PROC COMPARE statement 234
PROC SUMMARY statement 1178
NOREPLACE option
PROC FORMAT statement 434
normal distribution
NORMAL= option
PROC RANK statement 817
NORWEGIAN option
PROC SORT statement 1007
NOSEPS option
PROC TABULATE procedure 1191
NOSOURCE option
COPY statement (CATALOG) 160
NOSRC option
PROC CIMPORT statement 219
PROC CPORT statement 291
NOSTATS option
CHART procedure 195
NOSUMMARY option
PROC COMPARE statement 234
NOSYMBOL option
CHART procedure 195
NOSYMNAME option
PLOT statement (TIMEPLOT) 1296
NOTE option
PROC COMPARE statement 234
NOTRAP option
PROC MEANS statement 530

Index

NOTSORTED option
BY statement 58
BY statement (CALENDAR) 92
BY statement (CHART) 188
BY statement (COMPARE) 237
BY statement (MEANS) 535
BY statement (PLOT) 612
BY statement (PRINT) 716
BY statement (RANK) 819
BY statement (REPORT) 889
BY statement (STANDARD) 1168
BY statement (TABULATE) 1196
BY statement (TIMEPLOT) 1291
BY statement (TRANSPOSE)
FORMAT procedure 452
ID statement (COMPARE) 238
FORMAT procedure 436
NOUPDATE option
PROC FONTREG statement 421
NOVALUES option
PROC COMPARE statement 234
NOWARN option
PROC DATASETS statement 311
NOZERO option
DEFINE statement (REPORT) 903
NOZEROS option
CHART procedure 195
NPLUS1 option
PROC RANK statement 817
NPP option
PLOT statement (TIMEPLOT) 1296
NSRC option
PROC CPORT statement 291
null hypothesis
NUM option
PROC FSLIST statement 492
NUMBER option
PROC SQL statement 1036
numbers
template for printing 438
numeric values
converting raw character data to 470
summing 742
numeric variables
sorting orders for 1013
summing 737
NWAY option
PROC MEANS statement 530

O
OBS= option
PROC PRINT statement 709
observations
consolidating in reports 957
frequency of 61
grouping for reports 731
hidden 630
page layout 720
SQL procedure 1029
statistics for groups of 7
total number of
transposing variables into
weighting 549
observations, comparing
comparison summary 247, 252

matching observations 226

with ID variable 266
with output data set 270
ODS destinations
categories of 43
changing default settings 51
denition 40
destination-independent input 43
exclusion lists 52
SAS formatted destinations 44
selection lists 52
system resources and 47
third-party formatted destinations 45
ODS output
CALENDAR procedure 113
CHART procedure 198
denition 41
style elements for 989, 994, 1279
TABULATE procedure 1231
ODS (Output Delivery System) 32
customized output 51
DATASETS procedure and 368, 389
how it works 41
PLOT procedure and 630
printing reports 867
processing 41
registry and 49
samples 33
summary of 54
TABULATE procedure and 1182, 1279
terminology 40
ODS table names
CHART procedure 198
COMPARE procedure 253
DATASETS procedure 369
PLOT procedure 629
TIMEPLOT procedure 1298
ODS TRACE statement
LABEL= option 52
purpose 52
OL option
BREAK statement (REPORT) 887
RBREAK statement (REPORT) 910
ON option
CHECKBOX statement (PMENU) 668
one-tailed tests
operands
values from 1099
operating environment-specic procedures 30
operators
arithmetic 1123
in dimension expressions 1209
order of evaluation 1100
set operators 1094, 1123
truncated string comparison operators 1102
values from 1099
OPTION= option
PROC OPTIONS statement 596
OPTIONS 595
OPTIONS procedure 595
display settings for a group of options 593
examples 597
output 591
overview 591
results 596
syntax 595
task table 595

1431

OPTLOAD 602
OPTLOAD procedure 602
overview 601
syntax 602
task table 602
OPTSAVE 604
OPTSAVE procedure 604
overview 603
syntax 604
task table 604
ORDER BY clause
SQL procedure 1068, 1123
ORDER option
DEFINE statement (REPORT) 903
CLASS statement (MEANS) 537
CLASS statement (TABULATE) 1198
CONTENTS statement (DATASETS) 326
DEFINE statement (REPORT) 903
PROC MEANS statement 531
PROC TABULATE statement 1191, 1230
order variables 853, 903
orthogonal expressions 1123
OUT= argument
APPEND statement (DATASETS) 314
COPY statement (CATALOG) 159
COPY statement (DATASETS) 327
PROC IMPORT statement 503
OUT= option
CONTENTS statement (CATALOG) 158
CONTENTS statement (DATASETS) 326
OUTPUT statement (MEANS) 541
PROC COMPARE statement 235, 254, 270
PROC OPTSAVE statement 604
PROC PRTEXP statement 804
PROC PWENCODE statement 808
PROC RANK statement 817
PROC REPORT statement 878
PROC SORT statement 1010
PROC STANDARD statement 1167
PROC TABULATE statement 1192
PROC TRANSPOSE statement
OUT2= option
CONTENTS statement (DATASETS) 326
OUTALL option
PROC COMPARE statement 235
OUTBASE option
PROC COMPARE statement 235
OUTCOMP option
PROC COMPARE statement 235
OUTDIF option
PROC COMPARE statement 235
OUTDUR statement
CALENDAR procedure 98
outer joins 1086, 1123, 1138
OUTER UNION set operator 1095
OUTFILE= argument
PROC EXPORT statement 405
OUTFIN statement
CALENDAR procedure 99
OUTLIB= option
PROC CPORT statement 291
OUTNOEQUAL option
PROC COMPARE statement 235
OUTOBS= option
PROC SQL statement 1036
OUTPERCENT option
PROC COMPARE statement 235

1432

Index

output data sets

comparing observations 270
summary statistics in 273
OUTPUT destination 44
denition 40
output objects
customized output for 53
denition 40
determining destinations for 52, 53
OUTPUT= option
CALID statement (CALENDAR) 93
OUTPUT statement
MEANS procedure 541
Output window
printing from 867
OUTREPT= option
PROC REPORT statement 879
OUTSTART statement
CALENDAR procedure 99
OUTSTATS= option
PROC COMPARE statement 236, 255, 273
OUTTABLE= argument
PROC EXPORT statement 405
OUTTYPE= option
PROC CPORT statement 291
OUTWARD= option
PLOT statement (PLOT) 618
OVERLAY option
PLOT statement (PLOT) 619
PLOT statement (TIMEPLOT) 1296
overlaying plots 628, 634
overlining 886, 887, 909, 910
OVERWRITE option
PROC SORT statement 1010
OVP option
FSLIST command 494
PROC FSLIST statement 492
OVPCHAR= option
PLOT statement (TIMEPLOT) 1296

P
P keywords
p-values
page dimension 1215
page dimension text 1186
page ejects 717
page layout 720
column headings 722
column width 723
customizing 761
observations 720
plots 1297
with many variables 754
page numbering 775
PAGE option
BREAK statement (REPORT) 887
DEFINE statement (REPORT) 903
PROC FORMAT statement 434
RBREAK statement (REPORT) 910
PAGEBY statement
PRINT procedure 717
panels
in reports 968
PANELS= option
PROC REPORT statement 880

parameters
partitioned data sets
multi-threaded sorting 1013
password-protected data sets
appending 316
copying les 332
transporting 294
passwords 351
assigning 351
changing 351
DATASETS procedure with 359
encoding 807, 809
integrity constraints and 327
removing 352
pattern matching 1091, 1158
PC les
importing 506
PCTLDEF= option
PROC MEANS statement 532
PROC REPORT statement 882
PROC TABULATE statement 1193
PCTN statistic 1217
denominator for 1217
PCTSUM statistic 1217
denominator for 1218
PDF les
style elements 989
TABULATE procedure 1279
PDF output
sample 37
PDF reports 729
peakedness
penalties 626
changing 627, 661
index values for 626
PENALTIES= option
PLOT statement (PLOT) 619
percent coefcient of variation
percent difference 244
PERCENT option
CHART procedure 195
PROC RANK statement 818
percentage bar charts 201
percentages
displaying with denominator denitions 1269
in reports 975
TABULATE procedure 1216, 1266, 1269
percentiles
keywords and formulas
permanent data sets 16
permanent formats/informats 456
accessing 457
retrieving 478
picture formats 438
creating 464
digit selectors 442
directives 442
lling 483
message characters 442
steps for building 443
picture-name formats 447
PICTURE statement
FORMAT procedure 438
pie charts 182, 189
PIE statement
CHART procedure 189

PLACEMENT= option
PLOT statement (PLOT) 619
PLOT 609
PLOT procedure 609
combinations of variables 615
computational resources 628
concepts 624
examples 631
generating data with program statements 625
hidden observations 630
labeling plot points 625
missing values 630, 652
ODS table names 629
overview 606
portability of ODS output 630
printed output 629
results 629
RUN groups 624
scale of axes 629
syntax 609
task tables 609, 613
variable lists in plot requests 615
PLOT statement
PLOT procedure 613
TIMEPLOT procedure 1293
plots
collision states 628
contour plots 615, 642
customizing axes
customizing plotting symbols
data on logarithmic scale 639
data values on axis 640
hidden label characters 628
horizontal axis 632
labels 649, 654, 658
multiple observations, on one line
multiple plots per page 636
overlaying 628, 634
page layout 1297
penalties 626
plotting a single variable
plotting BY groups 646
pointer symbols 625
reference lines 628, 632
specifying in TIMEPLOT 1293
superimposing
plotting symbols 631
customizing
variables for
PMENU 667
PMENU catalog entries
naming 673
steps for building and using 680
storing 667
PMENU command 665
PMENU procedure 667
concepts 679
ending 680
examples 682
execution of 679
initiating 679
overview 665
PMENU catalog entries 680
syntax 667
task tables 667, 671
templates for 681
pointer symbols 625

Index

populations
PORT= statement
EXPORT procedure 410
IMPORT procedure 508, 512
POS= option
PLOT statement (TIMEPLOT) 1296
PostScript les 751
PostScript output
previewing 797
sample 35
power of statistical tests
PREFIX= option
PICTURE statement (FORMAT) 441
PROC TRANSPOSE statement
preloaded formats 904, 1199
class variables with 570, 1237
PRELOADFMT option
CLASS statement (MEANS) 538
CLASS statement (TABULATE) 1199
DEFINE statement (REPORT) 904
PRINT 706
PRINT option
PROC MEANS statement 531
PROC SQL statement 1036
PROC STANDARD statement 1167
PROC PRINTTO statement 774
PRINT procedure 706
examples 723
HTML reports 725
listing reports 724, 727
overview 703
page layout 720, 754, 761
PDF reports 729
PostScript les 751
procedure output 720
results 720
RTF reports 734
style denitions with 49
style elements 711
syntax 706
task tables 706, 707
XML les 740
PRINTALL option
PROC COMPARE statement 236
PRINTALLTYPES option
PROC MEANS statement 532
printer attributes
extracting from registry 803
writing to data sets 806
writing to log 805
printer denitions 789
adding 799
available to all users 798
creating 805
deleting 790, 799, 801
exporting 790
for Ghostview printer 797
in SASHELP library 790
modifying 799, 805
multiple 796
replicating 805
PRINTER destination 46
denition 40
printers
list of 790
routing log or output to 775, 785

PRINTIDVARS option
PROC MEANS statement 532
printing
See also printing reports
all data sets in library 767
data set contents 275
formatted values 25
grouping observations 731
informat/format descriptions 477
page ejects 717
page layout 720, 754, 761
selecting variables for 719, 723
template for printing numbers 438
printing reports 867
batch mode 867
from Output window 867
from REPORT window 867
interactive line mode 868
noninteractive mode 867
PRINTTO procedure 868
with forms 867
with ODS 867
PRINTMISS option
TABLE statement (TABULATE) 1206
PRINTTO 772
PRINTTO procedure 772
concepts 775
examples 776
overview 771
printing reports 868
syntax 772
task table 772
probability function
probability values
PROBT keyword
PROC CALENDAR statement 85
PROC CATALOG statement 155
options 156
PROC CHART statement 185
PROC CIMPORT statement 216
PROC COMPARE statement 230
PROC CONTENTS statement 276
PROC CPORT statement 286
PROC DATASETS statement 308
restricting member types 360
PROC DISPLAY statement 396
PROC EXPORT statement 404
PROC FONTREG statement 420
PROC FORMAT statement 432
PROC FSLIST statement 490
PROC IMPORT statement 502
PROC MEANS statement 527
PROC OPTIONS statement 595
PROC OPTLOAD statement 602
PROC OPTSAVE statement 604
PROC PLOT statement 609
PROC PMENU statement 667
PROC PRINT statement 707
PROC PRINTTO statement 772
PROC PRTDEF statement 790
PROC PRTEXP statement 804
PROC PWENCODE statement 807
PROC RANK statement 816
PROC REGISTRY statement 832
PROC REPORT statement 870
PROC SORT statement 1005
PROC SQL statement 1034

1433

PROC SQL tables 1029

adding rows 1056
aliases 1064, 1083
altering columns 1038
altering integrity constraints 1038
changing column attributes 1041
combining 1134
counting rows 1108
creating 1045, 1125
creating, from query expressions 1049
creating, from query results 1127
deleting 1054
deleting rows 1052
indexes on columns 1042
initial values of columns 1041
inserting data 1125
inserting values 1057
integrity constraints 1042, 1049
joining 1082, 1131, 1152
joining a table with itself 1082, 1084
joining more than two tables 1089
joining three tables 1145
ordering rows 1068
recursive table references 1049
renaming columns 1042
retrieving data from 1093
selecting columns 1058
selecting rows 1058
source tables 1063
table denitions 1053
table expressions 1094, 1113
updating 1069, 1070, 1129
without rows 1048
PROC SQL views
adding rows 1056
creating, from query expressions 1049
creating, from query results 1143
deleting 1054
deleting rows 1052
embedding LIBNAME statements in 1051
inserting rows 1057
librefs and stored views 1050
selecting columns 1058
selecting rows 1058
sorting data retrieved by 1050
source views 1063
SQL procedure 1029
storing DBMS connection information 1051
updating 1051, 1121
updating column values 1069
updating tables through 1070
view denitions 1053, 1123
PROC STANDARD statement 1166
PROC SUMMARY statement 1178
PROC TABULATE statement 1187
PROC TIMEPLOT statement 1290
PROC TRANSPOSE statement
procedure concepts 19
formatted values 25
input data sets 19
operating environment-specic procedures 30
processing all data sets in a library 30
RUN-group processing 19
shortcut notations for variable names 24
statistics, computational requirements 32
statistics, descriptions of 31
titles containing BY-group information 20

1434

Index

procedure output
as input le 782
default destinations 771
destinations for 771
page numbering 775
routing to catalog entries 779
routing to external les 776
routing to printer 775, 785
procedures
descriptions of 10
ending 63
functional categories 3
host-specic
raw data for examples
report-writing procedures 3, 4
statistical procedures 3, 6
style denitions with 48
utility procedures 4, 7
PROFILE= option
PROC REPORT statement 880
PROFILE window
REPORT procedure 926
project management 83
PROMPT option
PROC REPORT statement 881
PROC SQL statement 1037
PROMPTER window
REPORT procedure 927
PROTO 789
PROTO procedure 789
PRTDEF 789
PRTDEF procedure 789
examples 796
input data set 791
optional variables 793
overview 789
required variables 792
syntax 789
task table 790
valid variables 791
PRTEXP 803
PRTEXP procedure 803
concepts 805
examples 805
overview 803
syntax 803
PS= option
PROC REPORT statement 881
PSPACE= option
PROC REPORT statement 881
pull-down menus 665
activating 665
associating FRAME applications with 700
dening 674
for DATA step window applications 694
grayed items 672
items in 671
key sequences for 671
separator lines 678
submenus 678
PUT statement
compared with LINE statement (REPORT) 908
PW= option
MODIFY statement (DATASETS) 350
PROC DATASETS statement 311

PWD= statement
EXPORT procedure 409
IMPORT procedure 512
PWENCODE 807
PWENCODE procedure 807
concepts 808
encoding vs. encryption 808
examples 809
overview 807
syntax 807

Q
Q keywords
QMARKERS= option
PROC MEANS statement 532
PROC REPORT statement 882
PROC TABULATE statement 1193
QMETHOD= option
PROC MEANS statement 532
PROC REPORT statement 882
PROC TABULATE statement 1193
QNTLDEF= option
PROC MEANS statement 532
PROC REPORT statement 882
PROC TABULATE statement 1193
QRANGE keyword
quantiles 882, 1193
efciency issues 7
MEANS procedure 555
queries
creating tables from results 1127
creating views from results 1143
DBMS queries 1079
in-line view queries 1148
query-expression component 1093
query expressions 1094
creating PROC SQL tables from 1049
creating PROC SQL views from 1049
subqueries 1102
validating syntax 1070
QUIT statement 63
procedures supporting 63

R
radio boxes 670, 675
radio buttons 670, 676
color of 676
default 675
RADIOBOX statement
PMENU procedure 675
range
RANGE keyword
RANGE= statement
IMPORT procedure 508
ranges
for character strings 480
FORMAT procedure 453
RANK 815
RANK procedure 815
computer resources 820
concepts 820
examples 822

input variables 820

missing values 821
output data set 821
overview 813
ranking data 814
results 821
statistical applications 820
syntax 815
task tables 815, 816
variables for rank values 819
ranking data 814
ranks
groups based on 826
of multiple variables 822
values within BY groups 823
RANKS statement
RANK procedure 819
raw data
character data to numeric values 470
informats for 435
procedures examples
RBREAK statement
REPORT procedure 908
RBUTTON statement
PMENU procedure 676
READ= option
MODIFY statement (DATASETS) 350
PROC DATASETS statement 311
REF= option
CHART procedure 195
PLOT statement (TIMEPLOT) 1296
REFCHAR= option
PLOT statement (TIMEPLOT) 1297
reference lines 628, 632
reexive joins 1084
REFRESH option
INDEX CENTILES statement
(DATASETS) 344
registry 831
changing default HTML version setting 50
changing ODS destination default settings 51
clearing SASUSER 832
comparing le contents with 833, 841
comparing registries 832, 833, 842
debugging 833
exporting contents of 834
extracting printer attributes from 803
importing to 834, 839
keys, subkeys, and values 834, 835
listing 840
listing contents in log 835
loading system options from 601
ODS and 49
removing fonts from 424
sample entries 837
SASHELP specication 836
saving system option settings in 603
system fonts in 419
uppercasing key names 835
writing contents to log 834
writing SASHELP to log 835
writing SASUSER to log 835
registry les
creating 836
key names 836
sample registry entries 837
structure of 836

Index

values for keys 836

REGISTRY procedure 832
creating registry les 836
examples 839
overview 831
syntax 832
task table 832
remerging data
SQL procedure 1110
RENAME statement
DATASETS procedure 352
renaming les 322
REPAIR statement
DATASETS procedure 353
REPLACE option
PROC EXPORT statement 408
PROC IMPORT statement 506
PROC PRTDEF statement 790
PROC STANDARD statement 1167
report denitions
specifying 882
storing and reusing 868, 966
report items 897
report layout 852
across variables 854, 899
analysis variables 855, 899, 912
computed variables 855, 901
display variables 853, 901
group variables 854, 902
order variables 853
planning 852
statistics 857
variables, position and usage 855
variables usage 853
REPORT= option
PROC REPORT statement 882
REPORT procedure 869
See also REPORT procedure windows
break lines 861
compound names 862
compute blocks 858
concepts 852
ending program statements 906
examples 846
formatting characters 874
layout of reports 852
missing values 860, 977
output data set 980
overview 847
printing reports 867
report-building 936
report denitions 868
report types 847
sample reports 847
statistics 857
style denitions with 49
style elements 863, 989, 994
summary lines 937
syntax 869
task tables 869, 870, 885, 897,
REPORT procedure windows 913
BREAK 913
COMPUTE 916
COMPUTED VAR 916
DATA COLUMNS 917
DATA SELECTION 917
DEFINITION 918

DISPLAY PAGE 923

EXPLORE 924
FORMATS 925
LOAD REPORT 925
MESSAGES 926
PROFILE 926
PROMPTER 927
REPORT 928
ROPTIONS 928
SAVE DATA SET 933
SAVE DEFINITION 933
SOURCE 934
STATISTICS 934
WHERE 935
WHERE ALSO 936
report variables 936
REPORT window
printing from 867
REPORT procedure 928
report-writing procedures 3, 4
reports 847
See also report layout
building 936
code for 877
colors for 899, 909
column attributes 890
column for each variable value 960
columns 893
computed variables 983
consolidating observations 957
customized 704
customized summaries 907, 971
default summaries 885, 908
detail reports 847
from DICTONARY tables 1136
grouping observations 731
groups 986
header arrangement 893
help for 877
ID variables 902
limiting sums in 748
listing reports 704, 727
multiple-choice survey data 1260
multiple-response survey data 1255
order variables 903
ordering rows in 951
panels 968
PDF 729
percentages in 975
printing 867
RTF 734
samples of 847
selecting variables for 719, 948
shrinking 866
statistics in 954, 964
stub-and-banner reports 1269
summary reports 847
suppressing 878
RESET statement
SQL procedure 1058
RESUME option
AUDIT statement (DATASETS) 321
REVERSE option
PLOT statement (TIMEPLOT) 1297
PROC SORT statement 1011
RIGHT option
DEFINE statement (REPORT) 904

ROLLBACK statement (SQL) 1124

ROPTIONS window
REPORT procedure 928
ROUND option
PICTURE statement (FORMAT) 441
PROC PRINT statement 709
row headings
customizing 1244
eliminating 1249
indenting 1251
ROW= option
TABLE statement (TABULATE) 1206
row spacing 866
rows
adding to tables or views 1056
consolidating observations 957
counting 1108
deleting 1052
inserting 1057
joins and 1083
ordering 1068
ordering in reports 951
returned by subqueries 1080
selecting 1058, 1071
SQL procedure 1029
ROWS= option
PROC PRINT statement 710
RTF destination 46
denition 40
RTF les
style elements 989
TABULATE procedure 1279
RTF output
sample 36
RTF reports 734
RTSPACE= option
TABLE statement (TABULATE) 1206
RUN-group processing 19
CATALOG procedure 165
DATASETS procedure 357, 359
RUN groups
PLOT procedure 624

S
S= option
PLOT statement (PLOT) 622
samples
sampling distribution
SAS/ACCESS views
SQL procedure 1029
updating 1121
SAS/AF applications
executing 395, 396
SAS data views
DICTIONARY tables 1116
SQL procedure 1029
SAS Explorer window
list of available styles 48
SAS formatted destinations 43, 44
SAS/GRAPH device drivers
system fonts 425
SASHELP views 1116
retrieving information about 1117
SASUSER library
Ghostview printer denition in 797

1435

1436

Index

SAVAGE option
PROC RANK statement 818
SAVE DATA SET window
REPORT procedure 933
SAVE DEFINITION window
REPORT procedure 933
SAVE statement
CATALOG procedure 164
DATASETS procedure 355
SCANMEMO= statement
IMPORT procedure 512
SCANTEXT= statement
IMPORT procedure 509
SCANTIME= statement
IMPORT procedure 509, 512
schedule calendars 79, 102
advanced 80
simple 79
scheduling 83
searching for patterns 1091, 1158
SELECT clause
SQL procedure 1058
SELECT statement
CATALOG procedure 165
CIMPORT procedure 221
CPORT procedure 293
DATASETS procedure 356
FORMAT procedure 448
PRTEXP procedure 804
SQL procedure 1058
selection lists 52
destinations for output objects 53
SELECTION statement
PMENU procedure 677
separator lines 678
SEPARATOR statement
PMENU procedure 678
SERVER= statement
EXPORT procedure 410
IMPORT procedure 509, 513
SERVICE= statement
EXPORT procedure 410
IMPORT procedure 509, 513
set membership 1081
set operators 1094, 1123
SET statement
appending data 315
SHEET= statement
EXPORT procedure 408
IMPORT procedure 509
SHORT option
CONTENTS statement (DATASETS) 326
PROC OPTIONS statement 596
SHOWALL option
PROC REPORT statement 882
shrinking reports 866
signicance
simple indexes 1044
simple random sample
skewness
SKEWNESS keyword
SKIP option
BREAK statement (REPORT) 887
RBREAK statement (REPORT) 910
SLIST= option
PLOT statement (PLOT) 623
SORT 1005

sort order
ASCII 1007, 1014
EBCDIC 1007, 1014
for character variables 1014
for numeric variables 1013
SORT procedure 1005
character variable sorting orders 1014
collating-sequence options 1007
concepts 1013
DBMS data source 1013
examples 1016
integrity constraints 1015
multi-threaded sorting 1013
numeric variable sorting orders 1013
output 1016
output data set 1016
overview 1003
results 1016
sorting data sets 1004
stored sort information 1015
syntax 1005
task tables 1005, 1016
SORTEDBY= option
MODIFY statement (DATASETS) 350
sorting
by multiple variable values 1017
collating sequence 1007
data retrieved by views 1050
data sets 1004
in descending order 1019
maintaining relative order of observations 1020
multi-threaded 1013
retaining rst observation of BY groups 1023
stored sort information 1015
SORTMSG option
PROC SQL statement 1037
SORTSEQ= option
PROC SORT statement 1007
PROC SQL statement 1037
SORTSIZE= option
PROC SORT statement 1011
SOUNDS-LIKE operator 1150
SOURCE window
REPORT procedure 934
SPACE= option
CHART procedure 195
SPACING= option
DEFINE statement (REPORT) 904
PROC REPORT statement 883
SPLIT= option
PLOT statement (PLOT) 623
PROC PRINT statement 710
PROC REPORT statement 883
PROC TIMEPLOT statement 1290
spread of values
spreadsheets
exporting 407, 416
exporting subset of observations to 414
exporting to specic spreadsheet 415
importing 505, 506
importing from Excel workbook 517, 521
importing subset of records from 518
SQL 1031
SQL, embedded 1125
SQL components 1071
BETWEEN condition 1071

BTRIM function 1072

CALCULATED 1073
CASE expression 1073
COALESCE function 1075
column-denition 1076
column-modier 1077
column-name 1078
CONNECTION TO 1079
CONTAINS condition 1080
EXISTS condition 1080
IN condition 1081
IS condition 1081
joined-table 1082
LIKE condition 1091
LOWER function 1093
query-expression 1093
sql-expression 1099
SUBSTRING function 1106
summary-function 1107
table-expression 1113
UPPER function 1114
sql-expression component 1099
SQL procedure 1031
See also SQL components
ANSI Standard and 1122
coding conventions 1030
collating sequences 1123
column modiers 1123
concepts 1114
data set options with 1114
data types and dates 1076
DBMS connections 1115
DICTIONARY tables 1116
examples 1125
functions supported by 1124
indexes 1044
list of available styles 48
macro variables set by 1119
missing values 1081, 1160
naming conventions 1124
orthogonal expressions 1123
overview 1029
PROC SQL tables 1029
reserved words 1122
resetting options 1058
statistical functions 1124
syntax 1031
task tables 1033, 1034
terminology 1029
three-valued logic 1124
user privileges 1124
views 1029
SQL Procedure Pass-Through Facility
DBMS connections 1115
return codes 1115
SQLOBS macro variable 1119
SQLOOPS macro variable 1120
SQLRC macro variable 1120
SQLXMSG macro variable 1120
SQLXRC macro variable 1120
STANDARD 1165
standard deviation
standard error of the mean
STANDARD procedure 1165
examples 1171
missing values 1170
output data set 1170

Index

overview 1163
results 1170
standardizing data 1163
statistical computations 1171
syntax 1165
task tables 1165, 1166
standardizing data 1163
order of variables 1169
specifying variables 1169
weights for analysis variables 1170
star charts 183, 190
STAR statement
CHART procedure 190
START statement
CALENDAR procedure 100
STARTAT= option
PROC REGISTRY statement 835
STATE= option
ITEM statement (PMENU) 673
statements with same function in multiple procedures 57
ATTRIB 57
BY 58
FORMAT 57
FREQ 61
LABEL 57
QUIT 63
WEIGHT 64
WHERE 69
STATES option
PLOT statement (PLOT) 623
statistic, dened
statistic option
DEFINE statement (REPORT) 904
statistical analysis
transposing data for
statistical procedures 3, 6
efciency issues 7
quantiles 7
statistical summaries 1107
statistically signicant
statistics
based on number of arguments 1109
computational requirements for 32
descriptive statistics 1177
for groups of observations 7
formulas for
in reports 954
keywords for
measures of location
measures of shape
measures of variability
normal distribution
percentiles
populations
REPORT procedure 857
samples
sampling distribution
summarization procedures
table of descriptive statistics 31
TABULATE procedure 1213
testing hypotheses
weighted statistics 64
weights
statistics procedures
STATISTICS window
REPORT procedure 934

STATS option
PROC COMPARE statement 236
STD keyword
STD= option
PROC STANDARD statement 1167
STDDEV keyword
STDERR keyword
STDMEAN keyword
STIMER option
PROC SQL statement 1037
string comparison operators
truncated 1102
stub-and-banner reports 1269
Students t distribution
Students t statistic
two-tailed p-value
Students t test 554
STYLE= attribute
CALL DEFINE statement (REPORT) 892
style attributes 46
applying to table cells 1221
assigning with formats 1221
denition 48
style denitions
denition of 47
procedures with 48
SAS-supplied 48
style elements
class variable level value headings 1201
denition 47
for keyword headings 1202
for ODS output 989, 994
in dimension expressions 1210
PRINT procedure 711
REPORT procedure 863, 989, 994
TABULATE procedure 1194, 1220, 1279
STYLE= option
BREAK statement (REPORT) 887
CLASS statement (TABULATE) 1200
CLASSLEV statement (TABULATE) 1201
COMPUTE statement (REPORT) 896
DEFINE statement (REPORT) 904
ID statement (PRINT) 717
KEYWORD statement (TABULATE) 1202
PROC PRINT statement 711
PROC REPORT statement 883
PROC TABULATE statement 1194
RBREAK statement (REPORT) 910
REPORT procedure 863, 864
SUM statement (PRINT) 718
TABLE statement (TABULATE) 1207
TABULATE procedure 1220
VAR statement (PRINT) 720
VAR statement (TABULATE) 1211
SUBGROUP= option
CHART procedure 196
SUBMENU statement
PMENU procedure 678
submenus 678
subqueries 1102
compared with joins 1091
correlated 1104
efciency and 1105
returning rows 1080
subsetting data
SQL procedure 1065, 1067
WHERE statement 69

SUBSTITUTE= option
CHECKBOX statement (PMENU) 668
RBUTTON statement (PMENU) 676
SUBSTRING function (SQL) 1106
subtables 1186
SUM keyword
sum of squares
corrected
uncorrected
sum of the weights
SUM option
CHART procedure 196
SUM statement
CALENDAR procedure 100
PRINT procedure 718, 737, 742
SUMBY statement
PRINT procedure 719
summarization procedures
data requirements
SUMMARIZE option
BREAK statement (REPORT) 887
RBREAK statement (REPORT) 910
summarizing data
SQL procedure 1108
SUMMARY 1178
summary calendars 79, 102
multiple activities per day 107
simple 81
summary-function component 1107
summary lines 847
construction of 937
SUMMARY procedure 1178
overview 1177
syntax 1178
summary reports 847
summary statistics
COMPARE procedure 250, 273
SUMSIZE= option
PROC MEANS statement 533
SUMVAR= option
CHART procedure 196
SUMWGT keyword
superimposing plots
SUPPRESS option
BREAK statement (REPORT) 888
survey data
multiple-choice 1260
multiple-response 1255
SUSPEND option
AUDIT statement (DATASETS) 321
SWEDISH option
PROC SORT statement 1007
SYMBOL= option
CHART procedure 196
symbol variables
TIMEPLOT procedure 1292
SYSINFO macro variable 244
system fonts 419
SAS/GRAPH device drivers for 425
system options
display setting for single option 598
display settings for a group 593
list of current settings 591
loading from registry or data sets 601
OPTIONS procedure 591
procedures and 17
saving current settings 603

1437

1438

Index

short form listing 597

T
T keyword
table aliases 1064, 1083
TABLE= argument
PROC IMPORT statement 503
table attributes
denition 47
table denitions 1053
denition of 40, 47
table elements
denition 47
table-expression component 1113
table expressions 1094
TABLE statement
TABULATE procedure 1203
tables
See also PROC SQL tables
applying style attributes to cells 1221
cells with missing values 1228
class variable combinations 1235
crosstabulation 1269
customizing headings 1244
describing for printing 1203
formatting values in 1215
multipage 1253
subtables 1186
two-dimensional 1232
TABULATE 1187
TABULATE procedure 1187
BY-group processing 1215
complex tables 1181
concepts 1213
dimension expressions 1208
examples 1232
formatting characters 1189
formatting class variables 1214
formatting values in tables 1215
headings 1227, 1229, 1230
missing values 1200, 1222
ODS and 1182
overview 1180
page dimension 1215
percentages 1216
portability of ODS output 1231
results 1222
simple tables 1180
statistics 1213
style denitions with 49
style elements 1194, 1220
syntax 1187
task tables 1187, 1203
terminology 1183
tagsets 45
list of 41
TAGSORT option
PROC SORT statement 1011
TAPE option
PROC CIMPORT statement 219
PROC CPORT statement 291
TEMPLATE procedure
list of available styles 48
templates
for printing numbers 438

PMENU procedure 681

temporary data sets 16
temporary formats/informats 456
temporary variables 936
TERMINATE option
AUDIT statement (DATASETS) 321
text elds 670, 678
TEXT statement
PMENU procedure 678
TEXTSIZE= statement
IMPORT procedure 508
third-party formatted destinations 45
denition 43
formatting control and 46
threads
multi-threaded sorting 1013
THREADS option
PROC MEANS statement 534
PROC REPORT statement 883
PROC SQL statement 1037
PROC TABULATE statement 1195
SORT procedure 1012
three-valued logic 1124
TIES= option
PROC RANK statement 818
TIMEPLOT 1290
TIMEPLOT procedure 1290
data considerations 1297
examples
missing values 1298
ODS table names 1298
overview 1287
page layout 1297
procedure output 1287
results 1297
symbol variables 1292
syntax 1290
task tables 1290, 1293
titles
BY-group information in 20
TRANSLATE= option
PROC CPORT statement 291
translation tables
applying to transport les 297
for exporting catalogs 294
transport les 215
applying translation tables to 297
COPY procedure 278
CPORT procedure 285
transporting data sets 333
COPY procedure 278
password-protected 294
TRANSPOSE
TRANSPOSE option
PROC COMPARE statement 236, 252
TRANSPOSE procedure
complex transposition
copying variables without transposing
duplicate ID values
examples
formatted ID values
labeling transposed variables
listing variables to transpose
missing values
output data set
overview
results

simple transposition
syntax
task table
transposing BY groups
variable names, from numeric values
transposed variables
attributes of
labeling
naming
TRANTAB statement
CPORT procedure 294
TRAP option
PROC TABULATE procedure 1195
TrueType font les
replacing from a directory 427
searching directories for 422
TRUETYPE statement
FONTREG procedure 422
truncated string comparison operators 1102
two-dimensional tables 1232
two-tailed tests
Type 1 font les 422
Type I error rate
Type II error rate
TYPE= option
CHART procedure 196
MODIFY statement (DATASETS) 351
TYPE1 statement
FONTREG procedure 422
TYPES statement
MEANS procedure 546

U
UCLM keyword
UID= statement
EXPORT procedure 409
IMPORT procedure 513
UL option
BREAK statement (REPORT) 888
RBREAK statement (REPORT) 911
uncorrected sum of squares
underlining 886, 888, 910, 911
UNDO_POLICY= option
PROC SQL statement 1037
unformatted values
comparing 244
UNIFORM option
PROC PLOT statement 611
PROC TIMEPLOT statement 1290
UNINSTALL= option
PROC REGISTRY statement 835
union joins 1088
UNION operator 1097
UNIQUE keyword 1044
UNIQUE option
CREATE INDEX statement
(DATASETS) 345
UNIT= argument
PROC FSLIST statement 491
UNIT= option
PROC PRINTTO statement 775
universe
unsorted data
comparing 238

Index

UPCASE option
INVALUE statement (FORMAT) 436
PROC REGISTRY statement 835
UPDATE statement
SQL procedure 1069
UPDATECENTILES= option
CREATE INDEX statement
(DATASETS) 346
INDEX CENTILES statement
(DATASETS) 344
UPPER function (SQL) 1114
USEDATE= statement
IMPORT procedure 510, 513
USER data library 17
user input
collecting in dialog boxes 685
USER literal 1100
USER_VAR option
AUDIT statement (DATASETS) 321
USESASHELP option
PROC FONTREG statement 421
PROC PRTDEF statement 790
PROC PRTEXP statement 804
PROC REGISTRY statement 836
USS keyword
utility procedures 4, 7

V
VALIDATE statement
SQL procedure 1070
VALUE option
PROC OPTIONS statement 596
value-range-sets 453
VALUE statement
FORMAT procedure 448
VAR keyword
VAR statement
CALENDAR procedure 101
COMPARE procedure 239
MEANS procedure 548
PRINT procedure 719
RANK procedure 820
STANDARD procedure 1169
SUMMARY procedure 1179
TABULATE procedure 1211
TRANSPOSE procedure
VARDEF= option
PROC MEANS statement 534
PROC REPORT statement 883
PROC STANDARD statement 1167
PROC TABULATE statement 1195
variability
variable formats
COMPARE procedure 244
variable names
shortcut notations for 24
variables
across variables 854, 899
analysis variables 855, 899
associating formats/informats with 430, 455
attributes of 348
CHART procedure 197
class variables 1197
computed variables 855, 901, 983
copying without transposing

display variables 853, 901

group variables 854, 902
ID variables 902
in reports 853
labels 348
nested 1186
order of 719
order variables 853, 903
position and usage in reports 855
renaming 352
report variables 936
selecting for printing 719, 723
selecting for reports 948
SQL procedure 1029
standardizing 1163
temporarily associating formats with 28
temporarily dissociating formats from 29
temporary 936
transposing into observations
variables, comparing
by position 240
comparison summary 246
different data sets 261
different variable names 240
listing for matching 238
matching variables 226
multiple times 263
same data set 240, 264
selected variables 240
value comparison results 249
values comparison summary 248
variance
VARNUM option
CONTENTS statement (DATASETS) 326
VAXIS= option
PLOT statement (PLOT) 623
VBAR statement
CHART procedure 191
VERSION= statement
EXPORT procedure 410
IMPORT procedure 510, 513
vertical bar charts 180, 191
subdividing bars 203
VEXPAND option
PLOT statement (PLOT) 623
view denitions 1053
views
in-line 1064, 1123, 1148
SAS/ACCESS views 1121
SAS data views 1116
SASHELP views 1116, 1117
SQL procedure 1029
VPERCENT= option
PROC PLOT statement 611
VPOS= option
PLOT statement (PLOT) 623
VREF= option
PLOT statement (PLOT) 624
VREFCHAR= option
PLOT statement (PLOT) 624
VREVERSE option
PLOT statement (PLOT) 624
VSPACE= option
PLOT statement (PLOT) 624
VTOH= option
PROC PLOT statement 611

VZERO option
PLOT statement (PLOT) 624

W
WARNING option
PROC COMPARE statement 236
WAYS option
OUTPUT statement (MEANS) 546
WAYS statement
MEANS procedure 549
WBUILD macro 693
WEEKDAYS option
PROC CALENDAR statement 90
WEIGHT= option
DEFINE statement (REPORT) 905
VAR statement (MEANS) 548
VAR statement (TABULATE) 1212
WEIGHT statement 64
calculating weighted statistics 64
example 65
MEANS procedure 549
procedures supporting 64
REPORT procedure 912
STANDARD procedure 1170
TABULATE procedure 1212
weight values 874, 1189
weighted statistics 64
weighting observations 549
weights
analysis variables 64
WGDB= statement
EXPORT procedure 409
IMPORT procedure 513
WHERE ALSO window
REPORT procedure 936
WHERE clause
SQL procedure 1065
WHERE statement 69
example 69
procedures supporting 69
WHERE window
REPORT procedure 935
WIDTH= option
CHART procedure 197
DEFINE statement (REPORT) 905
PROC PRINT statement 714
window applications
menus for 694
windows
associating with menus 697
WINDOWS option
PROC REPORT statement 884
WITH statement
COMPARE procedure 240
WORKDATA= option
PROC CALENDAR statement 91
workdays data set 91, 110
default workshifts instead of 110
missing values 111
workshifts 110
WRAP option
PROC REPORT statement 884
WRITE= option
MODIFY statement (DATASETS) 351

1439

1440

Index

X
XML les

XML output
sample 38
740

Your Turn
If you have comments or suggestions about Base SAS 9.1 Procedures Guide, please
send them to us on a photocopy of this page, or send us electronic mail.
For comments about this book, please return the photocopy to
SAS Publishing
SAS Campus Drive
Cary, NC 27513
email: [email protected]
Send suggestions about the software, please return the photocopy to
SAS Institute Inc.
Technical Support Division
SAS Campus Drive
Cary, NC 27513
email: [email protected]

Base SAS 9.1 Procedures Guide

Volume 3
CORR, FREQ, and UNIVARIATE Procedures

Contents
Chapter 1. The CORR Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 2. The FREQ Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Chapter 3. The UNIVARIATE Procedure . . . . . . . . . . . . . . . . . . . . . . . . . 191
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
Syntax Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387

Chapter 1

The CORR Procedure

Chapter Contents
OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

GETTING STARTED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

SYNTAX . . . . . . . . .
PROC CORR Statement
BY Statement . . . . .
FREQ Statement . . . .
PARTIAL Statement . .
VAR Statement . . . .
WEIGHT Statement . .
WITH Statement . . . .

.
.
.
.
.
.
.
.

6
7
12
13
13
13
13
14

DETAILS . . . . . . . . . . . . . . . . .
Pearson Product-Moment Correlation .
Spearman Rank-Order Correlation . .
Kendalls Tau-b Correlation Coefcient
Hoeffding Dependence Coefcient . .
Partial Correlation . . . . . . . . . . .
Fishers z Transformation . . . . . . .
Cronbachs Coefcient Alpha . . . . .
Missing Values . . . . . . . . . . . . .
Output Tables . . . . . . . . . . . . .
Output Data Sets . . . . . . . . . . . .
Determining Computer Resources . . .
ODS Table Names . . . . . . . . . . .
ODS Graphics (Experimental) . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.

14
14
16
16
18
18
21
24
26
26
27
28
30
31

EXAMPLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Example 1.1. Computing Four Measures of Association . . . . . . .
Example 1.2. Computing Correlations between Two Sets of Variables
Example 1.3. Analysis Using Fishers z Transformation . . . . . . .
Example 1.4. Applications of Fishers z Transformation . . . . . . .
Example 1.5. Computing Cronbachs Coefcient Alpha . . . . . . .
Example 1.6. Saving Correlations in an Output Data Set . . . . . . .
Example 1.7. Creating Scatter Plots . . . . . . . . . . . . . . . . . .
Example 1.8. Computing Partial Correlations . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.

34
34
38
42
44
48
52
53
58

Chapter 1. The CORR Procedure

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Chapter 1

The CORR Procedure

Overview
The CORR procedure computes Pearson correlation coefcients, three nonparametric
measures of association, and the probabilities associated with these statistics. The
correlation statistics include
Pearson product-moment correlation
Spearman rank-order correlation
Kendalls tau-b coefcient
Hoeffdings measure of dependence, D
Pearson, Spearman, and Kendall partial correlation
Pearson product-moment correlation is a parametric measure of a linear relationship
between two variables. For nonparametric measures of association, Spearman rankorder correlation uses the ranks of the data values and Kendalls tau-b uses the number of concordances and discordances in paired observations. Hoeffdings measure of
dependence is another nonparametric measure of association that detects more general departures from independence. A partial correlation provides a measure of the
correlation between two variables after controlling the effects of other variables.
With only one set of analysis variables specied, the default correlation analysis includes descriptive statistics for each analysis variable and Pearson correlation statistics for these variables. You can also compute Cronbachs coefcient alpha for estimating reliability.
With two sets of analysis variables specied, the default correlation analysis includes
descriptive statistics for each analysis variable and Pearson correlation statistics between these two sets of variables.
You can save the correlation statistics in a SAS data set for use with other statistical
and reporting procedures.
For a Pearson or Spearman correlation, the Fishers z transformation can be used to
derive its condence limits and a p-value under a specied null hypothesis H0 : =
0 . Either a one-sided or a two-sided alternative is used for these statistics.
Experimental ODS graphics are now available with the CORR procedure, including
scatter plots and a scatter plot matrix of the analysis variables. For more information,
see the ODS Graphics section on page 31.

Chapter 1. The CORR Procedure

Getting Started
The following statements create the data set Fitness, which has been altered to contain some missing values:
*----------------- Data on Physical Fitness -----------------*
| These measurements were made on men involved in a physical |
| fitness course at N.C. State University.
|
| The variables are Age (years), Weight (kg),
|
| Runtime (time to run 1.5 miles in minutes), and
|
| Oxygen (oxygen intake, ml per kg body weight per minute)
|
| Certain values were changed to missing for the analysis.
|
*------------------------------------------------------------*;
data Fitness;
input Age Weight Oxygen RunTime @@;
datalines;
44 89.47 44.609 11.37
40 75.07 45.313 10.07
44 85.84 54.297 8.65
42 68.15 59.571 8.17
38 89.02 49.874
.
47 77.45 44.811 11.63
40 75.98 45.681 11.95
43 81.19 49.091 10.85
44 81.42 39.442 13.08
38 81.87 60.055 8.63
44 73.03 50.541 10.13
45 87.66 37.388 14.03
45 66.45 44.754 11.12
47 79.15 47.273 10.60
54 83.12 51.855 10.33
49 81.42 49.156 8.95
51 69.63 40.836 10.95
51 77.91 46.672 10.00
48 91.63 46.774 10.25
49 73.37
.
10.08
57 73.37 39.407 12.63
54 79.38 46.080 11.17
52 76.32 45.441 9.63
50 70.87 54.625 8.92
51 67.25 45.118 11.08
54 91.63 39.203 12.88
51 73.71 45.790 10.47
57 59.08 50.545 9.93
49 76.32
.
.
48 61.24 47.920 11.50
52 82.78 47.467 10.50
;

The following statements invoke the CORR procedure and request a correlation analysis:
ods html;
ods graphics on;
proc corr data=Fitness plots;
run;
ods graphics off;
ods html close;

This graphical display is requested by specifying the experimental ODS GRAPHICS

statement and the experimental PLOTS option. For general information about ODS
graphics, refer to Chapter 15, Statistical Graphics Using ODS (SAS/STAT Users
Guide). For specic information about the graphics available in the CORR procedure,
see the section ODS Graphics on page 31.

Getting Started

The CORR Procedure

Variables:

Age

Weight

Oxygen

RunTime

Simple Statistics
Variable

Mean

Std Dev

Sum

Minimum

Maximum

31
31
29
29

Age
Weight
Oxygen
RunTime

47.67742
77.44452
47.22721
10.67414

5.21144
8.32857
5.47718
1.39194

1478
2401
1370
309.55000

38.00000
59.08000
37.38800
8.17000

57.00000
91.63000
60.05500
14.03000

Figure 1.1. Univariate Statistics

By default, all numeric variables not listed in other statements are used in the analysis. Observations with nonmissing values for each variable are used to derive the
univariate statistics for that variable.
Pearson Correlation Coefficients
Prob > |r| under H0: Rho=0
Number of Observations
Age
Age

Weight

Oxygen

RunTime

1.00000

-0.23354
0.2061
31

-0.31474
0.0963
29

0.14478
0.4536
29

-0.23354
0.2061
31

1.00000

-0.15358
0.4264
29

0.20072
0.2965
29

-0.31474
0.0963
29

-0.15358
0.4264
29

1.00000

-0.86843
<.0001
28

0.14478
0.4536
29

0.20072
0.2965
29

-0.86843
<.0001
28

31
Weight

Oxygen

RunTime

1.00000
29

Figure 1.2. Pearson Correlation Coefcients

By default, Pearson correlation statistics are computed from observations with nonmissing values for each pair of analysis variables. With missing values in the analysis,
the Pearson Correlation Coefcients table shown in Figure 1.2 displays the correlation, the p-value under the null hypothesis of zero correlation, and the number of
nonmissing observations for each pair of variables.
The table displays a correlation of 0.86843 between Runtime and Oxygen, which
is signicant with a p-value less than 0.0001. That is, there exists an inverse linear relationship between these two variables. As Runtime (time to run 1.5 miles
in minutes) increases, Oxygen (oxygen intake, ml per kg body weight per minute)
decreases.

Chapter 1. The CORR Procedure

The experimental PLOTS option displays a symmetric matrix plot for the analysis
variables. This inverse linear relationship between these two variables, Oxygen and
Runtime, is also shown in Figure 1.3.

Figure 1.3. Symmetric Matrix Plot (Experimental)

Syntax
The following statements are available in PROC CORR.

PROC CORR < options > ;

BY variables ;
FREQ variable ;
PARTIAL variables ;
VAR variables ;
WEIGHT variable ;
WITH variables ;
The BY statement species groups in which separate correlation analyses are performed.
The FREQ statement species the variable that represents the frequency of occurrence for other values in the observation.

PROC CORR Statement

The PARTIAL statement identies controlling variables to compute Pearson,
Spearman, or Kendall partial-correlation coefcients.
The VAR statement lists the numeric variables to be analyzed and their order in the
correlation matrix. If you omit the VAR statement, all numeric variables not listed in
other statements are used.
The WEIGHT statement identies the variable whose values weight each observation
to compute Pearson product-moment correlation.
The WITH statement lists the numeric variables with which correlations are to be
computed.
The PROC CORR statement is the only required statement for the CORR procedure.
The rest of this section provides detailed syntax information for each of these statements, beginning with the PROC CORR statement. The remaining statements are in
alphabetical order.

PROC CORR Statement

PROC CORR < options > ;
The following table summarizes the options available in the PROC CORR statement.
Table 1.1. Summary of PROC CORR Options

Tasks
Specify data sets
Input data set
Output data set with Hoeffdings D statistics
Output data set with Kendall correlation statistics
Output data set with Pearson correlation statistics
Output data set with Spearman correlation statistics
Control statistical analysis
Exclude observations with nonpositive weight values
from the analysis
Exclude observations with missing analysis values
from the analysis
Request Hoeffdings measure of dependence, D
Request Kendalls tau-b
Request Pearson product-moment correlation
Request Spearman rank-order correlation
Request Pearson correlation statistics using Fishers
z transformation
Request Spearman rank-order correlation statistics
using Fishers z transformation
Control Pearson correlation statistics
Compute Cronbachs coefcient alpha
Compute covariances
Compute corrected sums of squares and crossproducts

Options
DATA=
OUTH=
OUTK=
OUTP=
OUTS=
EXCLNPWGT
NOMISS
HOEFFDING
KENDALL
PEARSON
SPEARMAN
FISHER PEARSON
FISHER SPEARMAN

ALPHA
COV
CSSCP

Chapter 1. The CORR Procedure

Table 1.1. (continued)

Tasks

Options
Compute correlation statistics based on Fishers
z transformation
Exclude missing values
Specify singularity criterion
Compute sums of squares and crossproducts
Specify the divisor for variance calculations

Control printed output

Display a specied number of ordered correlation coefcients
Suppress Pearson correlations
Suppress all printed output
Suppress p-values
Suppress descriptive statistics
Display ordered correlation coefcients

FISHER
NOMISS
SINGULAR=
SSCP
VARDEF=
BEST=
NOCORR
NOPRINT
NOPROB
NOSIMPLE
RANK

The following options (listed in alphabetical order) can be used in the PROC CORR
statement:
ALPHA

calculates and prints Cronbachs coefcient alpha. PROC CORR computes separate coefcients using raw and standardized values (scaling the variables to a unit
variance of 1). For each VAR statement variable, PROC CORR computes the correlation between the variable and the total of the remaining variables. It also computes
Cronbachs coefcient alpha using only the remaining variables.
If a WITH statement is specied, the ALPHA option is invalid. When you specify
the ALPHA option, the Pearson correlations will also be displayed. If you specify the
OUTP= option, the output data set also contains observations with Cronbachs coefcient alpha. If you use the PARTIAL statement, PROC CORR calculates Cronbachs
coefcient alpha for partialled variables. See the section Partial Correlation on
page 18.
BEST=n

prints the n highest correlation coefcients for each variable, n 1. Correlations

are ordered from highest to lowest in absolute value. Otherwise, PROC CORR prints
correlations in a rectangular table using the variable names as row and column labels.
If you specify the HOEFFDING option, PROC CORR displays the D statistics in
order from highest to lowest.
COV

displays the variance and covariance matrix. When you specify the COV option, the
Pearson correlations will also be displayed. If you specify the OUTP= option, the
output data set also contains the covariance matrix with the corresponding TYPE
variable value COV. If you use the PARTIAL statement, PROC CORR computes a
partial covariance matrix.

PROC CORR Statement

CSSCP

displays a table of the corrected sums of squares and crossproducts. When you specify the CSSCP option, the Pearson correlations will also be displayed. If you specify
the OUTP= option, the output data set also contains a CSSCP matrix with the corresponding TYPE variable value CSSCP. If you use a PARTIAL statement, PROC
CORR prints both an unpartial and a partial CSSCP matrix, and the output data set
contains a partial CSSCP matrix.
DATA=SAS-data-set

names the SAS data set to be analyzed by PROC CORR. By default, the procedure
uses the most recently created SAS data set.
EXCLNPWGT

excludes observations with nonpositive weight values from the analysis. By default,
PROC CORR treats observations with negative weights like those with zero weights
and counts them in the total number of observations.
FISHER < ( sher-options ) >

requests condence limits and p-values under a specied null hypothesis, H0 : =

0 , for correlation coefcients using Fishers z transformation. These correlations
include the Pearson correlations and Spearman correlations.
The following sher-options are available:
ALPHA=

species the level of the condence limits for the correlation, 100(1 )%.
The value of the ALPHA= option must be between 0 and 1, and the default is
ALPHA=0.05.
BIASADJ= YES | NO

species whether or not the bias adjustment is used in constructing condence

limits. The BIASADJ=YES option also produces a new correlation estimate
using the bias adjustment. By default, BIASADJ=YES.
RHO0= 0

species the value 0 in the null hypothesis H0 : = 0 , where 1 < 0 < 1.

By default, RHO0=0.
TYPE= LOWER | UPPER | TWOSIDED

species the type of condence limits. The TYPE=LOWER option requests a lower condence limit from the lower alternative H1 : < 0 , the
TYPE=UPPER option requests an upper condence limit from the upper alternative H1 : > 0 , and the default TYPE=TWOSIDED option requests twosided condence limits from the two-sided alternative H1 : = 0 .
HOEFFDING

requests a table of Hoeffdings D statistics. This D statistic is 30 times larger than the
usual denition and scales the range between 0.5 and 1 so that large positive values
indicate dependence. The HOEFFDING option is invalid if a WEIGHT or PARTIAL
statement is used.

Chapter 1. The CORR Procedure

KENDALL

requests a table of Kendalls tau-b coefcients based on the number of concordant

and discordant pairs of observations. Kendalls tau-b ranges from 1 to 1.
The KENDALL option is invalid if a WEIGHT statement is used. If you use a
PARTIAL statement, probability values for Kendalls partial tau-b are not available.
NOCORR

suppresses displaying of Pearson correlations. If you specify the OUTP= option, the
data set type remains CORR. To change the data set type to COV, CSSCP, or SSCP,
use the TYPE= data set option.
NOMISS

excludes observations with missing values from the analysis. Otherwise, PROC
CORR computes correlation statistics using all of the nonmissing pairs of variables.
Using the NOMISS option is computationally more efcient.
NOPRINT

suppresses all displayed output. Use NOPRINT if you want to create an output data
set only.
NOPROB

suppresses displaying the probabilities associated with each correlation coefcient.

NOSIMPLE

suppresses printing simple descriptive statistics for each variable. However, if you
request an output data set, the output data set still contains simple descriptive statistics
for the variables.
OUTH=output-data-set

creates an output data set containing Hoeffdings D statistics. The contents of the
output data set are similar to the OUTP= data set. When you specify the OUTH=
option, the Hoeffdings D statistics will be displayed, and the Pearson correlations
will be displayed only if the PEARSON, ALPHA, COV, CSSCP, SSCP, or OUT=
option is also specied.
OUTK=output-data-set

creates an output data set containing Kendall correlation statistics. The contents of
the output data set are similar to those of the OUTP= data set. When you specify the
OUTK= option, the Kendall correlation statistics will be displayed, and the Pearson
correlations will be displayed only if the PEARSON, ALPHA, COV, CSSCP, SSCP,
or OUT= option is also specied.
OUTP=output-data-set
OUT=output-data-set

creates an output data set containing Pearson correlation statistics. This data set also
includes means, standard deviations, and the number of observations. The value of
the TYPE variable is CORR. When you specify the OUTP= option, the Pearson
correlations will also be displayed. If you specify the ALPHA option, the output data
set also contains six observations with Cronbachs coefcient alpha.

PROC CORR Statement

OUTS=SAS-data-set

creates an output data set containing Spearman correlation coefcients. The contents of the output data set are similar to the OUTP= data set. When you specify
the OUTS= option, the Spearman correlation coefcients will be displayed, and the
Pearson correlations will be displayed only if the PEARSON, ALPHA, COV, CSSCP,
SSCP, or OUT= option is also specied.
PEARSON

requests a table of Pearson product-moment correlations. If you do not specify

the HOEFFDING, KENDALL, SPEARMAN, OUTH=, OUTK=, or OUTS= option, the CORR procedure produces Pearson product-moment correlations by default.
Otherwise, you must specify the PEARSON, ALPHA, COV, CSSCP, SSCP, or OUT=
option for Pearson correlations. The correlations range from 1 to 1.
RANK

displays the ordered correlation coefcients for each variable. Correlations are ordered from highest to lowest in absolute value. If you specify the HOEFFDING
option, the D statistics are displayed in order from highest to lowest.
SINGULAR=p

species the criterion for determining the singularity of a variable if you use a
PARTIAL statement. A variable is considered singular if its corresponding diagonal element after Cholesky decomposition has a value less than p times the original
unpartialled value of that variable. The default value is 1E8. The range of is
between 0 and 1.
SPEARMAN

requests a table of Spearman correlation coefcients based on the ranks of the variables. The correlations range from 1 to 1. If you specify a WEIGHT statement, the
SPEARMAN option is invalid.
SSCP

displays a table the sums of squares and crossproducts. When you specify the SSCP
option, the Pearson correlations will also be displayed. If you specify the OUTP=
option, the output data set contains a SSCP matrix and the corresponding TYPE
variable value is SSCP. If you use a PARTIAL statement, the unpartial SSCP matrix
is displayed, and the output data set does not contain an SSCP matrix.
VARDEF=d

species the variance divisor in the calculation of variances and covariances. The
following table shows the possible values for the value d and associated divisors,
where k is the number of PARTIAL statement variables. The default is VARDEF=DF.
Table 1.2. Possible Values for VARDEF=

Value
DF
N
WDF
WEIGHT|WGT

Divisor
degrees of freedom
number of observations
sum of weights minus one
sum of weights

Formula
nk1
n
(wi ) k 1
wi

Chapter 1. The CORR Procedure

The variance is computed as
1
d

(xi x)2

where x is the sample mean.

If a WEIGHT statement is used, the variance is computed as

1
d

wi (xi xw )2

where wi is the weight for the ith observation and xw is the weighted mean.

If you use the WEIGHT statement and VARDEF=DF, the variance is an estimate
of s2 , where the variance of the ith observation is V (xi ) = s2 /wi . This yields an
estimate of the variance of an observation with unit weight.
If you use the WEIGHT statement and VARDEF=WGT, the computed variance is
asymptotically an estimate of s2 /w, where w is the average weight (for large n). This

yields an asymptotic estimate of the variance of an observation with average weight.

BY Statement
BY variables ;
You can specify a BY statement with PROC CORR to obtain separate analyses on
observations in groups dened by the BY variables. If a BY statement appears, the
procedure expects the input data set to be sorted in order of the BY variables.
If your input data set is not sorted in ascending order, use one of the following alternatives:
Sort the data using the SORT procedure with a similar BY statement.
Specify the BY statement option NOTSORTED or DESCENDING in the BY
statement for the CORR procedure. The NOTSORTED option does not mean
that the data are unsorted but rather that the data are arranged in groups (according to values of the BY variables) and that these groups are not necessarily
in alphabetical or increasing numeric order.
Create an index on the BY variables using the DATASETS procedure.
For more information on the BY statement, refer to the discussion in SAS Language
Reference: Concepts. For more information on the DATASETS procedure, refer to
the discussion in the SAS Procedures Guide.

WEIGHT Statement

FREQ Statement
FREQ variable ;
The FREQ statement lists a numeric variable whose value represents the frequency
of the observation. If you use the FREQ statement, the procedure assumes that each
observation represents n observations, where n is the value of the FREQ variable. If
n is not an integer, SAS truncates it. If n is less than 1 or is missing, the observation
is excluded from the analysis. The sum of the frequency variable represents the total
number of observations.
The effects of the FREQ and WEIGHT statements are similar except when calculating degrees of freedom.

PARTIAL Statement
PARTIAL variables ;
The PARTIAL statement lists variables to use in the calculation of partial correlation statistics. Only the Pearson partial correlation, Spearman partial rank-order
correlation, and Kendalls partial tau-b can be computed. It is not valid with the
HOEFFDING option. When you use the PARTIAL statement, observations with
missing values are excluded.
With a PARTIAL statement, PROC CORR also displays the partial variance and standard deviation for each analysis variable if the PEARSON option is specied.

VAR Statement
VAR variables ;
The VAR statement lists variables for which to compute correlation coefcients. If
the VAR statement is not specied, PROC CORR computes correlations for all numeric variables not listed in other statements.

WEIGHT Statement
WEIGHT variable ;
The WEIGHT statement lists weights to use in the calculation of Pearson weighted
product-moment correlation. The HOEFFDING, KENDALL, and SPEARMAN options are not valid with the WEIGHT statement.
The observations with missing weights are excluded from the analysis. By default,
for observations with nonpositive weights, weights are set to zero and the observations are included in the analysis. You can use the EXCLNPWGT option to exclude
observations with negative or zero weights from the analysis.
Note that most SAS/STAT procedures, such as PROC GLM, exclude negative and
zero weights by default. If you use the WEIGHT statement, consider which value of
the VARDEF= option is appropriate. See the discussion of the VARDEF= option for
more information.

Chapter 1. The CORR Procedure

WITH Statement
WITH variables;
The WITH statement lists variables with which correlations of the VAR statement
variables are to be computed. The WITH statement requests correlations of the form
r(Xi , Yj ), where X1 , . . . , Xm are analysis variables specied in the VAR statement,
and Y1 , . . . , Yn are variables specied in the WITH statement. The correlation matrix
has a rectangular structure of the form

r(Y1 , X1 ) r(Y1 , Xm )

.
.
..
.
.

.
.
.
r(Yn , X1 ) r(Yn , Xm )

For example, the statements

proc corr;
var x1 x2;
with y1 y2 y3;
run;

produce correlations for the following combinations:

r(Y 1, X1) r(Y 1, X2)

r(Y 2, X1) r(Y 2, X2)
r(Y 3, X1) r(Y 3, X2)

Details
Pearson Product-Moment Correlation
The Pearson product-moment correlation is a parametric measure of association for
two variables. It measures both the strength and direction of a linear relationship. If
one variable X is an exact linear function of another variable Y, a positive relationship
exists if the correlation is 1 and a negative relationship exists if the correlation is 1.
If there is no linear predictability between the two variables, the correlation is 0. If
the two variables are normal with a correlation 0, the two variables are independent.
However, correlation does not imply causality because, in some cases, an underlying
causal relationship may not exist.
The following scatter plot matrix displays the relationship between two numeric random variables under various situations.

Pearson Product-Moment Correlation

Figure 1.4. Correlations between Two Variables

The scatter plot matrix shows a positive correlation between variables Y1 and X1, a
negative correlation between Y1 and X2, and no clear correlation between Y2 and
X1. The plot also shows no clear linear correlation between Y2 and X2, even though
Y2 is dependent on X2.
The formula for the population Pearson product-moment correlation, denoted xy , is
xy =

Cov(x, y)
V (x)V (y)

E( (x E(x))(y E(y)) )
E(x E(x))2 E(y E(y))2

The sample correlation, such as a Pearson product-moment correlation or weighted

product-moment correlation, estimates the population correlation. The formula for
the sample Pearson product-moment correlation is
rxy =

x)(yi y ) )

2
i (xi x)
i (yi y )
i ( (xi

where x is the sample mean of x and y is the sample mean of y. The formula for a

weighted Pearson product-moment correlation is

rxy =

wi (xi xw )(yi yw )

2
i wi (xi xw )
i wi (yi yw )
i

Chapter 1. The CORR Procedure

where wi is the weight, xw is the weighted mean of x, and yw is the weighted mean

of y.

Probability Values
Probability values for the Pearson correlation are computed by treating
1/2

t = (n 2)

r2
1 r2

1/2

as coming from a t distribution with (n2) degrees of freedom, where r is the sample
correlation.

Spearman Rank-Order Correlation

Spearman rank-order correlation is a nonparametric measure of association based on
the ranks of the data values. The formula is
=

i ( (Ri
i (Ri

R)(Si S) )

R)2

(Si S)2

where Ri is the rank of xi , Si is the rank of yi , R is the mean of the Ri values, and S
is the mean of the Si values.
PROC CORR computes the Spearman correlation by ranking the data and using the
ranks in the Pearson product-moment correlation formula. In case of ties, the averaged ranks are used.

Probability Values
Probability values for the Spearman correlation are computed by treating
1/2

t = (n 2)

r2
1 r2

1/2

as coming from a t distribution with (n2) degrees of freedom, where r is the sample
Spearman correlation.

Kendalls Tau-b Correlation Coefcient

Kendalls tau-b is a nonparametric measure of association based on the number of
concordances and discordances in paired observations. Concordance occurs when
paired observations vary together, and discordance occurs when paired observations
vary differently. The formula for Kendalls tau-b is
=

i<j

(sgn(xi xj )sgn(yi yj ))
(T0 T1 )(T0 T2 )

Kendalls Tau-b Correlation Coefcient

where T0 = n(n 1)/2, T1 = k tk (tk 1)/2, and T2 = l ul (ul 1)/2. The
tk is the number of tied x values in the kth group of tied x values, ul is the number
of tied y values in the lth group of tied y values, n is the number of observations, and
sgn(z) is dened as

if z > 0
1
0
if z = 0
sgn(z) =

1 if z < 0
PROC CORR computes Kendalls tau-b by ranking the data and using a method similar to Knight (1966). The data are double sorted by ranking observations according
to values of the rst variable and reranking the observations according to values of
the second variable. PROC CORR computes Kendalls tau-b from the number of interchanges of the rst variable and corrects for tied pairs (pairs of observations with
equal values of X or equal values of Y).

Probability Values
Probability values for Kendalls tau-b are computed by treating
s
V (s)
as coming from a standard normal distribution where
(sgn(xi xj )sgn(yi yj ))

s=
i<j

and V (s), the variance of s, is computed as

V (s) =

v0 vt vu
v1
v2
+
+
18
2n(n 1) 9n(n 1)(n 2)

where
v0 = n(n 1)(2n + 5)
vt =

tk (tk 1)(2tk + 5)

vu =

ul (ul 1)(2ul + 5)

v1 = (

tk (tk 1)) (

ui (ul 1))

v2 = (

ti (tk 1)(tk 2)) (

ul (ul 1)(ul 2))

The sums are over tied groups of values where ti is the number of tied x values and ui
is the number of tied y values (Noether 1967). The sampling distribution of Kendalls
partial tau-b is unknown; therefore, the probability values are not available.

Chapter 1. The CORR Procedure

Hoeffding Dependence Coefcient

Hoeffdings measure of dependence, D, is a nonparametric measure of association
that detects more general departures from independence. The statistic approximates a
weighted sum over observations of chi-square statistics for two-by-two classication
tables (Hoeffding 1948). Each set of (x, y) values are cut points for the classication.
The formula for Hoeffdings D is
D = 30

(n 2)(n 3)D1 + D2 2(n 2)D3

n(n 1)(n 2)(n 3)(n 4)

where D1 = i (Qi 1)(Qi 2), D2 = i (Ri 1)(Ri 2)(Si 1)(Si 2), and
D3 = i (Ri 2)(Si 2)(Qi 1). Ri is the rank of xi , Si is the rank of yi , and
Qi (also called the bivariate rank) is 1 plus the number of points with both x and y
values less than the ith point.
A point that is tied on only the x value or y value contributes 1/2 to Qi if the other
value is less than the corresponding value for the ith point.
A point that is tied on both x and y contributes 1/4 to Qi . PROC CORR obtains
the Qi values by rst ranking the data. The data are then double sorted by ranking
observations according to values of the rst variable and reranking the observations
according to values of the second variable. Hoeffdings D statistic is computed using
the number of interchanges of the rst variable. When no ties occur among data
set observations, the D statistic values are between 0.5 and 1, with 1 indicating
complete dependence. However, when ties occur, the D statistic may result in a
smaller value. That is, for a pair of variables with identical values, the Hoeffdings
D statistic may be less than 1. With a large number of ties in a small data set, the D
statistic may be less than 0.5. For more information about Hoeffdings D, refer to
Hollander and Wolfe (1973, p. 228).

Probability Values
The probability values for Hoeffdings D statistic are computed using the asymptotic
distribution computed by Blum, Kiefer, and Rosenblatt (1961). The formula is
4
(n 1) 4
D+
60
72
which comes from the asymptotic distribution. If the sample size is less than 10, refer
to the tables for the distribution of D in Hollander and Wolfe (1973).

Partial Correlation
A partial correlation measures the strength of a relationship between two variables,
while controlling the effect of other variables. The Pearson partial correlation between two variables, after controlling for variables in the PARTIAL statement, is
equivalent to the Pearson correlation between the residuals of the two variables after
regression on the controlling variables.

Partial Correlation
Let y = (y1 , y2 , . . . , yv ) be the set of variables to correlate and z = (z1 , z2 , . . . , zp )
be the set of controlling variables. The population Pearson partial correlation between
the ith and the jth variables of y given z is the correlation between errors (yi E(yi ))
and (yj E(yj )), where
E(yi ) = i + z i

and E(yj ) = j + z j

are the regression models for variables yi and yj given the set of controlling variables
z, respectively.
For a given sample of observations, a sample Pearson partial correlation between yi
and yj given z is derived from the residuals yi yi and yj yj , where

yi = i + z i

and yj = j + z j

are tted values from regression models for variables yi and yj given z.
The partial corrected sums of squares and crossproducts (CSSCP) of y given z are
the corrected sums of squares and crossproducts of the residuals y y . Using these

partial corrected sums of squares and crossproducts, you can calculate the partial
partial covariances and partial correlations.
PROC CORR derives the partial corrected sums of squares and crossproducts matrix by applying the Cholesky decomposition algorithm to the CSSCP matrix. For
Pearson partial correlations, let S be the partitioned CSSCP matrix between two sets
of variables, z and y:
S =

Szz Szy
Szy Syy

PROC CORR calculates Syy.z , the partial CSSCP matrix of y after controlling for z,
by applying the Cholesky decomposition algorithm sequentially on the rows associated with z, the variables being partialled out.
After applying the Cholesky decomposition algorithm to each row associated with
variables z, PROC CORR checks all higher numbered diagonal elements associated
with z for singularity. A variable is considered singular if the value of the corresponding diagonal element is less than times the original unpartialled corrected
sum of squares of that variable. You can specify the singularity criterion using
the SINGULAR= option. For Pearson partial correlations, a controlling variable z is
considered singular if the R2 for predicting this variable from the variables that are
already partialled out exceeds 1 . When this happens, PROC CORR excludes the
variable from the analysis. Similarly, a variable is considered singular if the R2 for
predicting this variable from the controlling variables exceeds 1 . When this happens, its associated diagonal element and all higher numbered elements in this row or
column are set to zero.
After the Cholesky decomposition algorithm is performed on all rows associated with
z, the resulting matrix has the form

Chapter 1. The CORR Procedure

T =

Tzz Tzy
0 Syy.z

where Tzz is an upper triangular matrix with Tzz Tzz = Szz , Tzz Tzy = Szy , and
Syy.z = Syy Tzy Tzy .
If Szz is positive denite, then Tzy = Tzz 1 Szy and the partial CSSCP matrix Syy.z
is identical to the matrix derived from the formula
1
Syy.z = Syy Szy Szz Szy

The partial variance-covariance matrix is calculated with the variance divisor

(VARDEF= option). PROC CORR then uses the standard Pearson correlation formula on the partial variance-covariance matrix to calculate the Pearson partial correlation matrix.
When a correlation matrix is positive denite, the resulting partial correlation between variables x and y after adjusting for a single variable z is identical to that
obtained from the rst-order partial correlation formula
rxy.z =

rxy rxz ryz

2
2
(1 rxz )(1 ryz )

where rxy , rxz , and ryz are the appropriate correlations.

The formula for higher-order partial correlations is a straightforward extension of the
preceding rst-order formula. For example, when the correlation matrix is positive
denite, the partial correlation between x and y controlling for both z1 and z2 is
identical to the second-order partial correlation formula
rxy.z1 z2 =

rxy.z1 rxz2 .z1 ryz2 .z1

2
2
(1 rxz2 .z1 )(1 ryz2 .z1 )

where rxy.z1 , rxz2 .z1 , and ryz2 .z1 are rst-order partial correlations among variables
x, y, and z2 given z1 .
To derive the corresponding Spearman partial rank-order correlations and Kendall
partial tau-b correlations, PROC CORR applies the Cholesky decomposition algorithm to the Spearman rank-order correlation matrix and Kendalls tau-b correlation
matrix and uses the correlation formula. That is, the Spearman partial correlation is
equivalent to the Pearson correlation between the residuals of the linear regression
of the ranks of the two variables on the ranks of the partialled variables. Thus, if a
PARTIAL statement is specied with the CORR=SPEARMAN option, the residuals
of the ranks of the two variables are displayed in the plot. The partial tau-b correlations range from 1 to 1. However, the sampling distribution of this partial tau-b is
unknown; therefore, the probability values are not available.

Fishers z Transformation
Probability Values
Probability values for the Pearson and Spearman partial correlations are computed by
treating
(n k 2)1/2 r
(1 r2 )1/2
as coming from a t distribution with (n k 2) degrees of freedom, where r is the
partial correlation and k is the number of variables being partialled out.

Fishers z Transformation
For a sample correlation r using a sample from a bivariate normal distribution with
correlation = 0, the statistic
tr = (n 2)1/2

r2
1 r2

1/2

has a Student-t distribution with (n 2) degrees of freedom.

With the monotone transformation of the correlation r (Fisher 1921)
zr = tanh1 (r) =

1
log
2

1+r
1r

the statistic z has an approximate normal distribution with mean and variance
E(zr ) = +

V (zr ) =

2(n 1)

1
n3

where = tanh1 ().

Chapter 1. The CORR Procedure

Assuming further that the two samples are from populations with identical correlation, a combined correlation estimate can be computed. The weighted average of the
corresponding z values is

(n1 3)z1 + (n2 3)z2

n1 + n2 6

where the weights are inversely proportional to their variances.

Thus, a combined correlation estimate is r = tanh() and V () = 1/(n1 + n2 6).

z
z
See Example 1.4 for further illustrations of these applications.
Note that this approach can be extended to include more than two samples.

Cronbachs Coefcient Alpha

Analyzing latent constructs such as job satisfaction, motor ability, sensory recognition, or customer satisfaction requires instruments to accurately measure the constructs. Interrelated items may be summed to obtain an overall score for each participant. Cronbachs coefcient alpha estimates the reliability of this type of scale
by determining the internal consistency of the test or the average correlation of items
within the test (Cronbach 1951).
When a value is recorded, the observed value contains some degree of measurement
error. Two sets of measurements on the same variable for the same individual may
not have identical values. However, repeated measurements for a series of individuals
will show some consistency. Reliability measures internal consistency from one set
of measurements to another. The observed value Y is divided into two components,
a true value T and a measurement error E. The measurement error is assumed to be
independent of the true value, that is,
Y =T +E

Cov(T, E) = 0

The reliability coefcient of a measurement test is dened as the squared correlation

between the observed value Y and the true value T , that is,
r2 (Y, T ) =

Cov(Y, T )2
V (T )2
V (T )
=
=
V (Y )V (T )
V (Y )V (T )
V (Y )

which is the proportion of the observed variance due to true differences among individuals in the sample. If Y is the sum of several observed variables measuring the
same feature, you can estimate V (T ). Cronbachs coefcient alpha, based on a lower
bound for V (T ), is an estimate of the reliability coefcient.
Suppose p variables are used with Yj = Tj + Ej for j = 1, 2, . . . , p, where Yj is the
observed value, Tj is the true value, and Ej is the measurement error. The measurement errors (Ej ) are independent of the true values (Tj ) and are also independent of

Cronbachs Coefcient Alpha

each other. Let Y0 =
true score. Because

Yj be the total observed score and T0 =

V (Tj )

(p 1)
j

Tj be the total

Cov(Ti , Tj )
i=j

a lower bound for V (T0 ) is given by

p
p1

Cov(Ti , Tj )
i=j

With Cov(Yi , Yj ) = Cov(Ti , Tj ) for i = j, a lower bound for the reliability coefcient, V (T0 )/V (Y0 ), is then given by the Cronbachs coefcient alpha:

p
p1

i=j

Cov(Yi , Yj )
V (Y0 )

p
p1

V (Yj )

V (Y0 )

If the variances of the items vary widely, you can standardize the items to a standard
deviation of 1 before computing the coefcient alpha. If the variables are dichotomous (0,1), the coefcient alpha is equivalent to the Kuder-Richardson 20 (KR-20)
reliability measure.
When the correlation between each pair of variables is 1, the coefcient alpha has a
maximum value of 1. With negative correlations between some variables, the coefcient alpha can have a value less than zero. The larger the overall alpha coefcient, the
more likely that items contribute to a reliable scale. Nunnally and Bernstein (1994)
suggests 0.70 as an acceptable reliability coefcient; smaller reliability coefcients
are seen as inadequate. However, this varies by discipline.
To determine how each item reects the reliability of the scale, you calculate a coefcient alpha after deleting each variable independently from the scale. The Cronbachs
coefcient alpha from all variables except the kth variable is given by
k =

p1
p2

i=k

V (Yi )

i=k

Yi )

If the reliability coefcient increases after an item is deleted from the scale, you can
assume that the item is not correlated highly with other items in the scale. Conversely,
if the reliability coefcient decreases, you can assume that the item is highly correlated with other items in the scale. Refer to SAS Communications, Fourth Quarter
1994, for more information on how to interpret Cronbachs coefcient alpha.
Listwise deletion of observations with missing values is necessary to correctly calculate Cronbachs coefcient alpha. PROC CORR does not automatically use listwise
deletion if you specify the ALPHA option. Therefore, you should use the NOMISS
option if the data set contains missing values. Otherwise, PROC CORR prints a
warning message indicating the need to use the NOMISS option with the ALPHA
option.

Chapter 1. The CORR Procedure

Missing Values
PROC CORR excludes observations with missing values in the WEIGHT and FREQ
variables. By default, PROC CORR uses pairwise deletion when observations contain missing values. PROC CORR includes all nonmissing pairs of values for each
pair of variables in the statistical computations. Therefore, the correlation statistics
may be based on different numbers of observations.
If you specify the NOMISS option, PROC CORR uses listwise deletion when a value
of the VAR or WITH statement variable is missing. PROC CORR excludes all observations with missing values from the analysis. Therefore, the number of observations
for each pair of variables is identical.
The PARTIAL statement always excludes the observations with missing values by
automatically invoking the NOMISS option. With the NOMISS option, the data are
processed more efciently because fewer resources are needed. Also, the resulting
correlation matrix is nonnegative denite.
In contrast, if the data set contains missing values for the analysis variables and the
NOMISS option is not specied, the resulting correlation matrix may not be nonnegative denite. This leads to several statistical difculties if you use the correlations
as input to regression or other statistical procedures.

Output Tables
By default, PROC CORR prints a report that includes descriptive statistics and correlation statistics for each variable. The descriptive statistics include the number of
observations with nonmissing values, the mean, the standard deviation, the minimum,
and the maximum.
If a nonparametric measure of association is requested, the descriptive statistics include the median. Otherwise, the sample sum is included. If a Pearson partial correlation is requested, the descriptive statistics also include the partial variance and
partial standard deviation.
If variable labels are available, PROC CORR labels the variables. If you specify the
CSSCP, SSCP, or COV option, the appropriate sum-of-squares and crossproducts and
covariance matrix appears at the top of the correlation report. If the data set contains
missing values, PROC CORR prints additional statistics for each pair of variables.
These statistics, calculated from the observations with nonmissing row and column
variable values, may include
SSCP(W,V), uncorrected sum-of-squares and crossproducts
USS(W), uncorrected sum-of-squares for the row variable
USS(V), uncorrected sum-of-squares for the column variable
CSSCP(W,V), corrected sum-of-squares and crossproducts
CSS(W), corrected sum-of-squares for the row variable
CSS(V), corrected sum-of-squares for the column variable

Output Data Sets

COV(W,V), covariance
VAR(W), variance for the row variable
VAR(V), variance for the column variable
DF(W,V), divisor for calculating covariance and variances
For each pair of variables, PROC CORR prints the correlation coefcients, the number of observations used to calculate the coefcient, and the p-value.
If you specify the ALPHA option, PROC CORR prints Cronbachs coefcient alpha,
the correlation between the variable and the total of the remaining variables, and
Cronbachs coefcient alpha using the remaining variables for the raw variables and
the standardized variables.

Output Data Sets

If you specify the OUTP=, OUTS=, OUTK=, or OUTH= option, PROC CORR creates an output data set containing statistics for Pearson correlation, Spearman correlation, Kendalls tau-b, or Hoeffdings D, respectively. By default, the output data set
is a special data set type (TYPE=CORR) that many SAS/STAT procedures recognize,
including PROC REG and PROC FACTOR. When you specify the NOCORR option
and the COV, CSSCP, or SSCP option, use the TYPE= data set option to change the
data set type to COV, CSSCP, or SSCP.
The output data set includes the following variables:
BY variables, which identify the BY group when using a BY statement
TYPE variable, which identies the type of observation
NAME variable, which identies the variable that corresponds to a given
row of the correlation matrix
INTERCEPT variable, which identies variable sums when specifying the
SSCP option
VAR variables, which identify the variables listed in the VAR statement
You can use a combination of the TYPE and NAME variables to identify the
contents of an observation. The NAME variable indicates which row of the correlation matrix the observation corresponds to. The values of the TYPE variable are
SSCP, uncorrected sums of squares and crossproducts
CSSCP, corrected sums of squares and crossproducts
COV, covariances
MEAN, mean of each variable
STD, standard deviation of each variable
N, number of nonmissing observations for each variable
SUMWGT, sum of the weights for each variable when using a WEIGHT statement
CORR, correlation statistics for each variable.

Chapter 1. The CORR Procedure

If you specify the SSCP option, the OUTP= data set includes an additional observation that contains intercept values. If you specify the ALPHA option, the OUTP=
data set also includes observations with the following TYPE values:
RAWALPHA, Cronbachs coefcient alpha for raw variables
STDALPHA, Cronbachs coefcient alpha for standardized variables
RAWALDEL, Cronbachs coefcient alpha for raw variables after deleting one
variable
STDALDEL, Cronbachs coefcient alpha for standardized variables after
deleting one variable
RAWCTDEL, the correlation between a raw variable and the total of the remaining raw variables
STDCTDEL, the correlation between a standardized variable and the total of
the remaining standardized variables
If you use a PARTIAL statement, the statistics are calculated after the variables are
partialled. If PROC CORR computes Pearson correlation statistics, MEAN equals
zero and STD equals the partial standard deviation associated with the partial variance
for the OUTP=, OUTK=, and OUTS= data sets. Otherwise, PROC CORR assigns
missing values to MEAN and STD.

Determining Computer Resources

The only factor limiting the number of variables that you can analyze is the amount
of available memory. The computer resources that PROC CORR requires depend
on which statements and options you specify. To determine the computer resources,
dene the following variables as follows:
L
C
V
W
P

=
=
=
=
=

number of observations in the data set

number of correlation types (C = 1, 2, 3, 4)
number of VAR statement variables
number of WITH statement variables
number of PARTIAL statement variables

Furthermore, dene the following variables:

= V +W +P
V W
K =
V (V + 1)/2
K
L =
T (T + 1)/2

when W > 0
when W = 0
when P = 0
when P > 0

Determining Computer Resources

For small N and large K, the CPU time varies as K for all types of correlations. For
large N , the CPU time depends on the type of correlation:
K N
T N log N
K N log N

with PEARSON (default)

with SPEARMAN
with HOEFFDING or KENDALL

You can reduce CPU time by specifying NOMISS. With NOMISS, processing is
much faster when most observations do not contain missing values. The options and
statements you use in the procedure require different amounts of storage to process
the data. For Pearson correlations, the amount of temporary storage needed (in bytes)
is
M

= 40T + 16L + 56K + 56T

The NOMISS option decreases the amount of temporary storage by 56K bytes, the
FISHER option increases the storage by 24K bytes, the PARTIAL statement increases the storage by 12T bytes, and the ALPHA option increases the storage by
32V + 16 bytes.
The following example uses a PARTIAL statement, which excludes missing values.
proc corr;
var x1 x2;
with y1 y2 y3;
partial z1;
run;

Therefore, using 40T + 16L + 56T + 12T , the minimum temporary storage equals
984 bytes (T = 2 + 3 + 1 and L = T (T + 1)/2).
Using the SPEARMAN, KENDALL, or HOEFFDING option requires additional
temporary storage for each observation. For the most time-efcient processing, the
amount of temporary storage (in bytes) is
M

= 40T + 8K + 8L C + 12T N + 28N + QS + QP + QK

where
QS

QK =

0
68T
56K
0
32N
0

with NOSIMPLE
otherwise
with PEARSON and without NOMISS
otherwise
with KENDALL or HOEFFDING
otherwise

Chapter 1. The CORR Procedure

The following example requests Kendalls tau-b coefcients:
proc corr kendall;
var x1 x2 x3;
run;

Therefore, the minimum temporary storage in bytes is

= 40 3 + 8 6 + 8 6 1 + 12 3N + 28N + 3 68 + 32N
= 420 + 96N

where N is the number of observations.

If M bytes are not available, PROC CORR must process the data multiple times to
compute all the statistics. This reduces the minimum temporary storage you need by
12(T 2)N bytes. When this occurs, PROC CORR prints a note suggesting a larger
memory region.

ODS Table Names

PROC CORR assigns a name to each table it creates. You can use these names to
reference the table when using the Output Delivery System (ODS) to select tables
and create output data sets.
Table 1.3. ODS Tables Produced with the PROC CORR Statement

ODS Name
Cov
CronbachAlpha
CronbachAlphaDel
Csscp
FisherPearsonCorr
FisherSpearmanCorr
HoeffdingCorr
KendallCorr
PearsonCorr
SimpleStats
SpearmanCorr
Sscp
VarInformation

Description
Covariances
Coefcient alpha
Coefcient alpha with deleted variable
Corrected sums of squares and crossproducts
Pearson correlation statistics using Fishers
z Transformation
Spearman correlation statistics using Fishers
z Transformation
Hoeffdings D statistics
Kendalls tau-b coefcients
Pearson correlations
Simple descriptive statistics
Spearman correlations
Sums of squares and crossproducts
Variable information

Option
COV
ALPHA
ALPHA
CSSCP
FISHER
FISHER SPEARMAN
HOEFFDING
KENDALL
PEARSON
SPEARMAN
SSCP

Table 1.4. ODS Tables Produced with the PARTIAL Statement

ODS Name
FisherPearsonPartialCorr
FisherSpearmanPartialCorr

Description
Pearson Partial Correlation Statistics
Using Fishers z Transformation
Spearman Partial Correlation Statistics
Using Fishers z Transformation

Option
FISHER
FISHER SPEARMAN

ODS Graphics (Experimental)

Table 1.4. (continued)

ODS Name
PartialCsscp
PartialCov
PartialKendallCorr
PartialPearsonCorr
PartialSpearmanCorr

Description
Partial corrected sums of squares and
crossproduct
Partial covariances
Partial Kendall tau-b coefcients
Partial Pearson correlations
Partial Spearman correlations

Option
CSSCP
COV
KENDALL
SPEARMAN

ODS Graphics (Experimental)

This section describes the use of ODS for creating graphics with the CORR procedure. These graphics are experimental in this release, meaning that both the graphical
results and the syntax for specifying them are subject to change in a future release.
To request these graphs, you must specify the ODS GRAPHICS statement in addition
to the following options in the PROC CORR statement. For more information on the
ODS GRAPHICS statement, refer to Chapter 15, Statistical Graphics Using ODS
(SAS/STAT Users Guide).
PLOTS
PLOTS = MATRIX < ( matrix-options ) >
PLOTS = SCATTER < ( scatter-options ) >
PLOTS = ( MATRIX < ( matrix-options ) > SCATTER < ( scatter-options ) > )

requests a scatter plot matrix for all variables, scatter plots for all pairs of variables, or
both. If only the option keyword PLOTS is specied, the PLOTS=MATRIX option
is used. When you specify the PLOTS option, the Pearson correlations will also be
displayed.
You can specify the following with the PLOTS= option:
MATRIX < ( matrix-options ) >

requests a scatter plot matrix for all variables. That is, the procedure displays a symmetric matrix plot with variables in the VAR list if a WITH statement is not specied.
Otherwise, the procedure displays a rectangular matrix plot with the WITH variables
appear down the side and the VAR variables appear across the top.
The available matrix-options are:
NMAXVAR=n

species the maximum number of variables in the VAR list to be displayed

in the matrix plot, where n 0. If you specify NMAXVAR=0, then the total
number of variables in the VAR list is used and no restriction occurs. By default,
NMAXVAR=5.
NMAXWITH=n

species the maximum number of variables in the WITH list to be displayed in

the matrix plot, where n 0. If you specify NMAXWITH=0, then the total
number of variables in the WITH list is used and no restriction occurs. By
default, NMAXWITH=5.

Chapter 1. The CORR Procedure

SCATTER < ( scatter-options ) >

requests a scatter plot for each pair of variables. That is, the procedure displays a
scatter plot for each pair of distinct variables from the VAR list if a WITH statement
is not specied. Otherwise, the procedure displays a scatter plot for each pair of
variables, one from the WITH list and the other from the VAR list.
The available scatter-options are:
ALPHA=numbers

species the values for the condence or prediction ellipses to be displayed

in the scatter plots, where 0 < < 1. For each value specied, a (1 )
condence or prediction ellipse is created. By default, = 0.05.
ELLIPSE=PREDICTION | MEAN | NONE

requests prediction ellipses for new observations (ELLIPSE=PREDICTION),

condence ellipses for the mean (ELLIPSE=MEAN), or no ellipses
(ELLIPSE=NONE) to be created in the scatter plots.
By default,
ELLIPSE=PREDICTION.
NOINSET

suppresses the default inset of summary information for the scatter plot. The
inset table is displayed next to the scatter plot and contains statistics such as
number of observations (NObs), correlation, and p-value (Prob >|r|).
NOLEGEND

suppresses the default legend for overlaid prediction or condence ellipses. The
legend table is displayed next to the scatter plot and identies each ellipse displayed in the plot.
NMAXVAR=n

species the maximum number of variables in the VAR list to be displayed in

the plots, where n 0. If you specify NMAXVAR=0, then the total number of variables in the VAR list is used and no restriction occurs. By default,
NMAXVAR=5.
NMAXWITH=n

species the maximum number of variables in the WITH list to be displayed in

the plots, where n 0. If you specify NMAXWITH=0, then the total number
of variables in the WITH list is used and no restriction occurs. By default,
NMAXWITH=5.
When the relationship between two variables is nonlinear or when outliers are present,
the correlation coefcient may incorrectly estimate the strength of the relationship.
Plotting the data enables you to verify the linear relationship and to identify the potential outliers.
The partial correlation between two variables, after controlling for variables in the
PARTIAL statement, is the correlation between the residuals of the linear regression
of the two variables on the partialled variables. Thus, if a PARTIAL statement is
also specied, the residuals of the analysis variables are displayed in the scatter plot
matrix and scatter plots.

ODS Graphics (Experimental)

Condence and Prediction Ellipses
The CORR procedure optionally provides two types of ellipses for each pair of variables in a scatter plot. One is a condence ellipse for the population mean, and the
other is a prediction ellipse for a new observation. Both assume a bivariate normal
distribution.

Let Z and S be the sample mean and sample covariance matrix of a random sample
of size n from a bivariate normal distribution with mean and covariance matrix .

The variable Z is distributed as a bivariate normal variate with mean zero and
covariance (1/n), and it is independent of S. Using Hotellings T 2 statistic, which
is dened as

T 2 = n(Z ) S1 (Z )
a 100(1 )% condence ellipse for is computed from the equation
n
2

(Z ) S1 (Z ) =
F2,n2 (1 )
n1
n2
where F2,n2 (1 ) is the (1 ) critical value of an F distribution with degrees
of freedom 2 and n 2.
A prediction ellipse is a region for predicting a new observation in the population. It
also approximates a region containing a specied percentage of the population.
Denote a new observation as the bivariate random variable Znew . The variable

Znew Z = (Znew ) (Z )
is distributed as a bivariate normal variate with mean zero (the zero vector) and covariance (1 + 1/n), and it is independent of S. A 100(1 )% prediction ellipse
is then given by the equation
n
2(n + 1)

(Z ) S1 (Z ) =
F2,n2 (1 )
n1
n2
The family of ellipses generated by different critical values of the F distribution has
a common center (the sample mean) and common major and minor axis directions.
The shape of an ellipse depends on the aspect ratio of the plot. The ellipse indicates
the correlation between the two variables if the variables are standardized (by dividing the variables by their respective standard deviations). In this situation, the ratio
between the major and minor axis lengths is
1 + |r|
1 |r|
In particular, if r = 0, the ratio is 1, which corresponds to a circular condence
contour and indicates that the variables are uncorrelated. A larger value of the ratio
indicates a larger positive or negative correlation between the variables.

Chapter 1. The CORR Procedure

ODS Graph Names

The CORR procedure assigns a name to each graph it creates using ODS. You can
use these names to reference the graphs when using ODS. The names are listed in
Table 1.5.
To request these graphs you must specify the ODS GRAPHICS statement in addition
to the options and statements indicated in Table 1.5. For more information on the
ODS GRAPHICS statement, refer to Chapter 15, Statistical Graphics Using ODS
(SAS/STAT Users Guide).
Table 1.5. ODS Graphics Produced by PROC CORR

ODS Graph Name

ScatterPlot
RecMatrixPlot
SymMatrixPlot

Plot Description
Scatter plot
Rectangular scatter plot matrix
Symmetric scatter plot matrix

Option
PLOTS=SCATTER
PLOTS=MATRIX
PLOTS=MATRIX

Statement
WITH
(omit WITH)

Examples
Example 1.1. Computing Four Measures of Association
This example produces a correlation analysis with descriptive statistics and four measures of association: the Pearson product-moment correlation, the Spearman rankorder correlation, Kendalls tau-b coefcients, and Hoeffdings measure of dependence, D.
The Fitness data set created in the Getting Started section beginning on page 4
contains measurements from a study of physical tness of 31 participants. The following statements request all four measures of association for the variables Weight,
Oxygen, and Runtime.
ods html;
ods graphics on;
title Measures of Association for a Physical Fitness Study;
proc corr data=Fitness pearson spearman kendall hoeffding
plots;
var Weight Oxygen RunTime;
run;
ods graphics off;
ods html close;

Note that Pearson correlations are computed by default only if all three nonparametric correlations (SPEARMAN, KENDALL, and HOEFFDING) are not specied.
Otherwise, you need to specify the PEARSON option explicitly to compute Pearson
correlations.
By default, observations with nonmissing values for each variable are used to derive the univariate statistics for that variable. When nonparametric measures of association are specied, the procedure displays the median instead of the sum as an
additional descriptive measure.

Example 1.1. Computing Four Measures of Association

Output 1.1.1. Simple Statistics

Measures of Association for a Physical Fitness Study
The CORR Procedure
3

Variables:

Weight

Oxygen

RunTime

Simple Statistics
Variable
Weight
Oxygen
RunTime

Mean

Std Dev

Median

Minimum

Maximum

31
29
29

77.44452
47.22721
10.67414

8.32857
5.47718
1.39194

77.45000
46.67200
10.50000

59.08000
37.38800
8.17000

91.63000
60.05500
14.03000

Output 1.1.2. Pearson Correlation Coefcients

Measures of Association for a Physical Fitness Study
Pearson Correlation Coefficients
Prob > |r| under H0: Rho=0
Number of Observations
Weight
Weight

Oxygen

RunTime

1.00000

-0.15358
0.4264
29

0.20072
0.2965
29

-0.15358
0.4264
29

1.00000

-0.86843
<.0001
28

0.20072
0.2965
29

-0.86843
<.0001
28

31
Oxygen

RunTime

1.00000
29

The Pearson correlation is a parametric measure of association for two continuous

random variables. When there is missing data, the number of observations used to
calculate the correlation can vary.
In Output 1.1.2, the Pearson correlation between Runtime and Oxygen is 0.86843,
which is signicant with a p-value less than 0.0001. This indicates a strong negative
linear relationship between these two variables. As Runtime increases, Oxygen
decreases linearly.
The Spearman rank-order correlation is a nonparametric measure of association based
on the ranks of the data values. The Spearman Correlation Coefcients table
shown in Output 1.1.3 displays results similar to those of the Pearson Correlation
Coefcients table.

Chapter 1. The CORR Procedure

Output 1.1.3. Spearman Correlation Coefcients

Measures of Association for a Physical Fitness Study
Spearman Correlation Coefficients
Prob > |r| under H0: Rho=0
Number of Observations
Weight
Weight

Oxygen

RunTime

1.00000

-0.06824
0.7250
29

0.13749
0.4769
29

-0.06824
0.7250
29

1.00000

-0.80131
<.0001
28

0.13749
0.4769
29

-0.80131
<.0001
28

31
Oxygen

RunTime

1.00000
29

Output 1.1.4. Kendalls Tau-b Correlation Coefcients

Measures of Association for a Physical Fitness Study
Kendall Tau b Correlation Coefficients
Prob > |r| under H0: Rho=0
Number of Observations
Weight
Weight

Oxygen

RunTime

1.00000

-0.00988
0.9402
29

0.06675
0.6123
29

-0.00988
0.9402
29

1.00000

-0.62434
<.0001
28

0.06675
0.6123
29

-0.62434
<.0001
28

31
Oxygen

RunTime

1.00000
29

Kendalls tau-b is a nonparametric measure of association based on the number

of concordances and discordances in paired observations. The Kendall Tau-b
Correlation Coefcients table shown in Output 1.1.4 displays results similar to those
of the Pearson Correlation Coefcients table in Output 1.1.2.
Hoeffdings measure of dependence, D, is a nonparametric measure of association
that detects more general departures from independence. Without ties in the variables,
the values of the D statistic can vary between -0.5 and 1, with 1 indicating complete
dependence. Otherwise, the D statistic can result in a smaller value. Since ties occur
in the variable Weight, the D statistic for the Weight variable is less than 1, as shown
in the Hoeffding Dependence Coefcients table in Output 1.1.5.

Example 1.1. Computing Four Measures of Association

Output 1.1.5. Hoeffdings Dependence Coefcients

Measures of Association for a Physical Fitness Study
Hoeffding Dependence Coefficients
Prob > D under H0: D=0
Number of Observations
Weight

Oxygen

RunTime

Weight

0.97690
<.0001
31

-0.00497
0.5101
29

-0.02355
1.0000
29

Oxygen

-0.00497
0.5101
29

1.00000

0.23449
<.0001
28

-0.02355
1.0000
29

0.23449
<.0001
28

RunTime

1.00000
29

Output 1.1.6. Symmetric Scatter Plot Matrix (Experimental)

The experimental PLOTS option requests a symmetric scatter plot for the analysis
variables listed in the VAR statement. The strong negative linear relationship between
Oxygen and Runtime is evident in Output 1.1.6.

Chapter 1. The CORR Procedure

This display is requested by specifying both the ODS GRAPHICS statement and
the PLOTS option. For general information about ODS graphics, refer to Chapter 15,
Statistical Graphics Using ODS (SAS/STAT Users Guide). For specic information
about the graphics available in the CORR procedure, see the section ODS Graphics
on page 31.

Example 1.2. Computing Correlations between Two Sets of

Variables
The following statements create a data set which contains measurements for four iris
parts from Fishers iris data (1936): sepal length, sepal width, petal length, and petal
width. Each observation represents one specimen.
*------------------- Data on Iris Setosa --------------------*
| The data set contains 50 iris specimens from the species
|
| Iris Setosa with the following four measurements:
|
| SepalLength (sepal length)
|
| SepalWidth (sepal width)
|
| PetalLength (petal length)
|
| PetalWidth (petal width)
|
| Certain values were changed to missing for the analysis.
|
*------------------------------------------------------------*;
data Setosa;
input SepalLength SepalWidth PetalLength PetalWidth @@;
label sepallength=Sepal Length in mm.
sepalwidth=Sepal Width in mm.
petallength=Petal Length in mm.
petalwidth=Petal Width in mm.;
datalines;
50 33 14 02 46 34 14 03 46 36 . 02
51 33 17 05 55 35 13 02 48 31 16 02
52 34 14 02 49 36 14 01 44 32 13 02
50 35 16 06 44 30 13 02 47 32 16 02
48 30 14 03 51 38 16 02 48 34 19 02
50 30 16 02 50 32 12 02 43 30 11 .
58 40 12 02 51 38 19 04 49 30 14 02
51 35 14 02 50 34 16 04 46 32 14 02
57 44 15 04 50 36 14 02 54 34 15 04
52 41 15 .
55 42 14 02 49 31 15 02
54 39 17 04 50 34 15 02 44 29 14 02
47 32 13 02 46 31 15 02 51 34 15 02
50 35 13 03 49 31 15 01 54 37 15 02
54 39 13 04 51 35 14 03 48 34 16 02
48 30 14 01 45 23 13 03 57 38 17 03
51 38 15 03 54 34 17 02 51 37 15 04
52 35 15 02 53 37 15 02
;

The following statements request a correlation analysis between two sets of variables,
the sepal measurements and the petal measurements.

Example 1.2. Computing Correlations between Two Sets of Variables

ods html;
ods graphics on;
title Fisher (1936) Iris Setosa Data;
proc corr data=Setosa sscp cov plots;
var sepallength sepalwidth;
with petallength petalwidth;
run;
ods graphics off;
ods html close;

The CORR procedure displays univariate statistics for variables in the VAR and
WITH statements.
Output 1.2.1. Simple Statistics
Fisher (1936) Iris Setosa Data
The CORR Procedure
2 With Variables:
2
Variables:

PetalLength PetalWidth
SepalLength SepalWidth

Simple Statistics
Variable
PetalLength
PetalWidth
SepalLength
SepalWidth

Mean

Std Dev

Sum

49
48
50
50

14.71429
2.52083
50.06000
34.28000

1.62019
1.03121
3.52490
3.79064

721.00000
121.00000
2503
1714

Simple Statistics
Variable
PetalLength
PetalWidth
SepalLength
SepalWidth

Minimum

Maximum

Label

11.00000
1.00000
43.00000
23.00000

19.00000
6.00000
58.00000
44.00000

Petal
Petal
Sepal
Sepal

Length in mm.
Width in mm.
Length in mm.
Width in mm.

When the WITH statement is specied together with the VAR statement, the
CORR procedure produces rectangular matrices for statistics such as covariances
and correlations. The matrix rows correspond to the WITH variables (PetalLength
and PetalWidth) while the matrix columns correspond to the VAR variables
(SepalLength and SepalWidth). The CORR procedure uses the WITH variable
labels to label the matrix rows.
The SSCP option requests a table of the uncorrected sum-of-squares and crossproducts matrix, and the COV option requests a table of the covariance matrix. The SSCP
and COV options also produce a table of the Pearson correlations.

Chapter 1. The CORR Procedure

The sum-of-squares and crossproducts statistics for each pair of variables are computed by using observations with nonmissing row and column variable values. The
Sums of Squares and Crossproducts table shown in Output 1.2.2 displays the
crossproduct, sum of squares for the row variable, and sum of squares for the column variable for each pair of variables.
Output 1.2.2. Sum-of-squares and Crossproducts
Fisher (1936) Iris Setosa Data
Sums of Squares and Crossproducts
SSCP / Row Var SS / Col Var SS
SepalLength

SepalWidth

PetalLength
Petal Length in mm.

36214.00000
10735.00000
123793.0000

24756.00000
10735.00000
58164.0000

PetalWidth
Petal Width in mm.

6113.00000
355.00000
121356.0000

4191.00000
355.00000
56879.0000

The variances are computed by using observations with nonmissing row and column
variable values. The Variances and Covariances table shown in Output 1.2.3 displays the covariance, variance for the row variable, variance for the column variable,
and the associated degrees of freedom for each pair of variables.
Output 1.2.3. Variances and Covariances
Fisher (1936) Iris Setosa Data
Variances and Covariances
Covariance / Row Var Variance / Col Var Variance / DF
SepalLength

SepalWidth

PetalLength
Petal Length in mm.

1.270833333
2.625000000
12.33333333
48

1.363095238
2.625000000
14.60544218
48

PetalWidth
Petal Width in mm.

0.911347518
1.063386525
11.80141844
47

1.048315603
1.063386525
13.62721631
47

Example 1.2. Computing Correlations between Two Sets of Variables

Output 1.2.4. Pearson Correlation Coefcients

Fisher (1936) Iris Setosa Data
Pearson Correlation Coefficients
Prob > |r| under H0: Rho=0
Number of Observations
Sepal
Length

Sepal
Width

PetalLength
Petal Length in mm.

0.22335
0.1229
49

0.22014
0.1285
49

PetalWidth
Petal Width in mm.

0.25726
0.0775
48

0.27539
0.0582
48

When there are missing values in the analysis variables, the Pearson Correlation
Coefcients table shown in Output 1.2.4 displays the correlation, the p-value under
the null hypothesis of zero correlation, and the number of observations for each pair
of variables. Only the correlation between PetalWidth and SepalLength and the
correlation between PetalWidth and SepalWidth are slightly positive.
The experimental PLOTS option displays a rectangular scatter plot matrix for the two
sets of variables. The VAR variables SepalLength and SepalWidth are listed across
the top of the matrix, and the WITH variables PetalLength and PetalWidth are listed
down the side of the matrix. As measured in Output 1.2.4, the plot for PetalWidth
and SepalLength and the plot for PetalWidth and SepalWidth show slight positive
correlations.

Chapter 1. The CORR Procedure

Output 1.2.5. Rectangular Matrix Plot (Experimental)

Example 1.3. Analysis Using Fishers z Transformation

The following statements request Pearson correlation statistics using Fishers z transformation for the data set Fitness.
proc corr data=Fitness nosimple fisher;
var weight oxygen runtime;
run;

The NOSIMPLE option suppresses the table of descriptive statistics. The Pearson
Correlation Coefcients table is displayed by default.

Example 1.3. Analysis Using Fishers z Transformation

Output 1.3.1. Sample Correlations

Fisher (1936) Iris Setosa Data
The CORR Procedure
Pearson Correlation Coefficients
Prob > |r| under H0: Rho=0
Number of Observations
Weight

Oxygen

RunTime

1.00000

-0.15358
0.4264
29

0.20072
0.2965
29

-0.15358
0.4264
29

1.00000

-0.86843
<.0001
28

0.20072
0.2965
29

-0.86843
<.0001
28

Weight

31
Oxygen

RunTime

1.00000
29

The FISHER option requests correlation statistics using Fishers z transformation,

which are shown in Output 1.3.2.
Output 1.3.2. Correlation Statistics Using Fishers z Transformation
Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

Weight
Weight
Oxygen

Oxygen
RunTime
RunTime

Sample
Correlation

Fishers z

Bias
Adjustment

Correlation
Estimate

29
29
28

-0.15358
0.20072
-0.86843

-0.15480
0.20348
-1.32665

-0.00274
0.00358
-0.01608

-0.15090
0.19727
-0.86442

Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

95% Confidence Limits

Weight
Weight
Oxygen

Oxygen
RunTime
RunTime

-0.490289
-0.182422
-0.935728

0.228229
0.525765
-0.725221

p Value for
H0:Rho=0
0.4299
0.2995
<.0001

See the section Fishers z Transformation on page 21 for details on Fishers z transformation.
The following statements request one-sided hypothesis tests and condence limits for
the correlation using Fishers z transformation.
proc corr data=Fitness nosimple nocorr fisher (type=lower);
var weight oxygen runtime;
run;

Chapter 1. The CORR Procedure

The NOSIMPLE option suppresses the Simple Statistics table, and the NOCORR
option suppresses the Pearson Correlation Coefcients table.
Output 1.3.3. One-sided Correlation Analysis Using Fishers z Transformation
The CORR Procedure
Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

Weight
Weight
Oxygen

Oxygen
RunTime
RunTime

Sample
Correlation

Fishers z

Bias
Adjustment

Correlation
Estimate

29
29
28

-0.15358
0.20072
-0.86843

-0.15480
0.20348
-1.32665

-0.00274
0.00358
-0.01608

-0.15090
0.19727
-0.86442

Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

Weight
Weight
Oxygen

Oxygen
RunTime
RunTime

Lower 95% CL

p Value for
H0:Rho<=0

-0.441943
-0.122077
-0.927408

0.7850
0.1497
1.0000

The TYPE=LOWER option requests a lower condence limit and a p-value for the
test of the one-sided hypothesis H0 : 0 against the alternative hypothesis H1 : >
0. Here Fishers z, the bias adjustment, and the estimate of the correlation are the
same as for the two-sided alternative. However, because TYPE=LOWER is specied,
only a lower condence limit is computed for each correlation, and one-sided pvalues are computed.

Example 1.4. Applications of Fishers z Transformation

This example illustrates some applications of Fishers z transformation. For details,
see the section Fishers z Transformation on page 21.
The following statements simulate independent samples of variables X and Y from
a bivariate normal distribution. The rst batch of 150 observations is sampled using
a known correlation of 0.3, the second batch of 150 observations is sampled using a
known correlation of 0.25, and the third batch of 100 observations is sampled using
a known correlation of 0.3.
data Sim (drop=i);
do i=1 to 400;
X = rannor(135791);
Batch = 1 + (i>150)
if Batch = 1 then Y
if Batch = 2 then Y
if Batch = 3 then Y
output;
end;
run;

+
=
=
=

(i>300);
0.3*X + 0.9*rannor(246791);
0.25*X + sqrt(.8375)*rannor(246791);
0.3*X + 0.9*rannor(246791);

Example 1.4. Applications of Fishers z Transformation

This data set will be used to illustrate the following applications of Fishers z transformation:
Testing whether a population correlation is equal to a given value
Testing for equality of two population correlations
Combining correlation estimates from different samples
See the section Fishers z Transformation on page 21.

Testing Whether a Population Correlation Is Equal to a Given Value 0

You can use the following statements to test the null hypothesis H0 : = 0.5 against
a two-sided alternative H1 : = 0.5.
ods select FisherPearsonCorr;
title Analysis for Batch 1;
proc corr data=Sim (where=(Batch=1)) fisher(rho0=.5);
var X Y;
run;

The test is requested with the option FISHER(RHO0=0.5). The results, which are
based on Fishers transformation, are shown in Output 1.4.1.
Output 1.4.1. Fishers Test for H0 : = 0
Analysis for Batch 1
The CORR Procedure
Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

Sample
Correlation

Fishers z

Bias
Adjustment

Correlation
Estimate

150

0.22081

0.22451

0.0007410

0.22011

Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

95% Confidence Limits

0.062034

0.367409

------H0:Rho=Rho0----Rho0
p Value
0.50000

<.0001

The null hypothesis is rejected since the p-value is less than 0.0001.

Testing for Equality of Two Population Correlations

You can use the following statements to test for equality of two population correlations, 1 and 2 . Here, the null hypothesis H0 : 1 = 2 is tested against the alternative
H 1 : 1 = 2 .

Chapter 1. The CORR Procedure

ods select FisherPearsonCorr;
ods output FisherPearsonCorr=SimCorr;
title Testing Equality of Population Correlations;
proc corr data=Sim (where=(Batch=1 or Batch=2)) fisher;
var X Y;
by Batch;
run;

The ODS SELECT statement restricts the output from PROC CORR to the
FisherPearsonCorr table, which is shown in Output 1.4.2; see the section ODS
Table Names on page 30. The output data set SimCorr contains Fishers z statistics
for both batches.
Output 1.4.2. Fishers Correlation Statistics
Testing Equality of Population Correlations
----------------------------------- Batch=1 -----------------------------------The CORR Procedure
Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

Sample
Correlation

Fishers z

Bias
Adjustment

Correlation
Estimate

150

0.22081

0.22451

0.0007410

0.22011

Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

p Value for
H0:Rho=0

95% Confidence Limits

0.062034

0.367409

0.0065

Testing Equality of Population Correlations

----------------------------------- Batch=2 -----------------------------------The CORR Procedure
Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

Sample
Correlation

Fishers z

Bias
Adjustment

Correlation
Estimate

150

0.33694

0.35064

0.00113

0.33594

Pearson Correlation Statistics (Fishers z Transformation)

Variable

With
Variable

95% Confidence Limits

0.185676

0.470853

p Value for
H0:Rho=0
<.0001

Example 1.4. Applications of Fishers z Transformation

The p-value for testing H0 is derived by treating the difference z1 z2 as a normal
random variable with mean zero and variance 1/(n1 3) + 1/(n2 3), where z1 and
z2 are Fishers z transformation of the sample correlations r1 and r2 , respectively,
and where n1 and n2 are the corresponding sample sizes.
The following statements compute the p-value shown in Output 1.4.3.
data SimTest (drop=Batch);
merge SimCorr (where=(Batch=1) keep=Nobs ZVal Batch
rename=(Nobs=n1 ZVal=z1))
SimCorr (where=(Batch=2) keep=Nobs ZVal Batch
rename=(Nobs=n2 ZVal=z2));
variance = 1/(n1-3) + 1/(n2-3);
z = (z1 - z2) / sqrt( variance );
pval = probnorm(z);
if (pval > 0.5) then pval = 1 - pval;
pval = 2*pval;
run;
proc print data=SimTest noobs;
run;

Output 1.4.3. Test of Equality of Observed Correlations

variance

150

0.22451

150

0.35064

0.013605

z
-1.08135

pval
0.27954

In Output 1.4.3, the p-value of 0.2795 does not provide evidence to reject the null
hypothesis that 1 = 2 . The sample sizes n1 = 150 and n2 = 150 are not large
enough to detect the difference 1 2 = 0.05 at a signicance level of = 0.05.

Combining Correlation Estimates from Different Samples

Assume that sample correlations r1 and r2 are computed from two independent samples of n1 and n2 observations, respectively. A combined correlation estimate is
given by r = tanh(), where z is the weighted average of the z-transformations of

r1 and r2 :
z=

(n1 3)z1 + (n2 3)z2

n1 + n2 6

Chapter 1. The CORR Procedure

The following statements compute a combined estimate of using Batch 1 and
Batch 3:
ods output FisherPearsonCorr=SimCorr2;
proc corr data=Sim (where=(Batch=1 or Batch=3)) fisher noprint;
var X Y;
by Batch;
run;
data SimComb (drop=Batch);
merge SimCorr2 (where=(Batch=1) keep=Nobs ZVal Batch
rename=(Nobs=n1 ZVal=z1))
SimCorr2 (where=(Batch=3) keep=Nobs ZVal Batch
rename=(Nobs=n2 ZVal=z2));
z = ((n1-3)*z1 + (n2-3)*z2) / (n1+n2-6);
corr = tanh(z);
var = 1/(n1+n2-6);
lcl = corr - probit(0.975)*sqrt(var);
ucl = corr + probit(0.975)*sqrt(var);
run;
proc print data=SimComb noobs;
var n1 z1 n2 z2 corr lcl ucl;
run;

Output 1.4.4 displays the combined estimate of .

Output 1.4.4. Combined Correlation Estimate
n1

150

0.22451

100

0.23929

corr
0.22640

lcl

ucl

0.10092

0.35187

Thus, a correlation estimate from the combined samples is r = 0.23. The 95%
condence interval displayed in Output 1.4.4 is (0.10, 0.35) using the variance of the
combined estimate. Note that this interval contains the population correlation 0.3.
See the section Applications of Fishers z Transformation on page 23.

Example 1.5. Computing Cronbachs Coefcient Alpha

The following statements create the data set Fish1 from the Fish data set used in
Chapter 67, The STEPDISC Procedure. The cubic root of the weight (Weight3) is
computed as a one-dimensional measure of the size of a sh.
*------------------- Fish Measurement Data ----------------------*
| The data set contains 35 fish from the species Bream caught in |
| Finlands lake Laengelmavesi with the following measurements: |
| Weight
(in grams)
|
| Length3 (length from the nose to the end of its tail, in cm) |
| HtPct
(max height, as percentage of Length3)
|
| WidthPct (max width, as percentage of Length3)
|
*----------------------------------------------------------------*;

Example 1.5. Computing Cronbachs Coefcient Alpha

data Fish1 (drop=HtPct WidthPct);
title Fish Measurement Data;
input Weight Length3 HtPct WidthPct @@;
Weight3= Weight**(1/3);
Height=HtPct*Length3/100;
Width=WidthPct*Length3/100;
datalines;
242.0 30.0 38.4 13.4
290.0 31.2 40.0 13.8
340.0 31.1 39.8 15.1
363.0 33.5 38.0 13.3
430.0 34.0 36.6 15.1
450.0 34.7 39.2 14.2
500.0 34.5 41.1 15.3
390.0 35.0 36.2 13.4
450.0 35.1 39.9 13.8
500.0 36.2 39.3 13.7
475.0 36.2 39.4 14.1
500.0 36.2 39.7 13.3
500.0 36.4 37.8 12.0
. 37.3 37.3 13.6
600.0 37.2 40.2 13.9
600.0 37.2 41.5 15.0
700.0 38.3 38.8 13.8
700.0 38.5 38.8 13.5
610.0 38.6 40.5 13.3
650.0 38.7 37.4 14.8
575.0 39.5 38.3 14.1
685.0 39.2 40.8 13.7
620.0 39.7 39.1 13.3
680.0 40.6 38.1 15.1
700.0 40.5 40.1 13.8
725.0 40.9 40.0 14.8
720.0 40.6 40.3 15.0
714.0 41.5 39.8 14.1
850.0 41.6 40.6 14.9
1000.0 42.6 44.5 15.5
920.0 44.1 40.9 14.3
955.0 44.0 41.1 14.3
925.0 45.3 41.4 14.9
975.0 45.9 40.6 14.7
950.0 46.5 37.9 13.7
;

The following statements request a correlation analysis and compute Cronbachs coefcient alpha for the variables Weight3, Length3, Height, and Width.
ods html;
ods graphics on;
title Fish Measurement Data;
proc corr data=fish1 nomiss alpha plots;
var Weight3 Length3 Height Width;
run;
ods graphics off;
ods html close;

The NOMISS option excludes observations with missing values, and the PLOTS option requests a symmetric scatter plot matrix for the analysis variables.
By default, the CORR procedure displays descriptive statistics for each variable, as
shown in Output 1.5.1.

Chapter 1. The CORR Procedure

Output 1.5.1. Simple Statistics

Fish Measurement Data
The CORR Procedure
4

Variables:

Weight3

Length3

Height

Width

Simple Statistics
Variable

Mean

Std Dev

Sum

Minimum

Maximum

34
34
34
34

Weight3
Length3
Height
Width

8.44751
38.38529
15.22057
5.43805

0.97574
4.21628
1.98159
0.72967

287.21524
1305
517.49950
184.89370

6.23168
30.00000
11.52000
4.02000

10.00000
46.50000
18.95700
6.74970

Since the NOMISS option is specied, the same set of 34 observations is used to
compute the correlation for each pair of variables. The correlations are shown in
Output 1.5.2.
Output 1.5.2. Pearson Correlation Coefcients
Fish Measurement Data
Pearson Correlation Coefficients, N = 34
Prob > |r| under H0: Rho=0
Weight3

Length3

Height

Width

Weight3

1.00000

0.96523
<.0001

0.96261
<.0001

0.92789
<.0001

Length3

0.96523
<.0001

1.00000

0.95492
<.0001

0.92171
<.0001

Height

0.96261
<.0001

0.95492
<.0001

1.00000

0.92632
<.0001

Width

0.92789
<.0001

0.92171
<.0001

0.92632
<.0001

1.00000

Since the data set contains only one species of sh, all the variables are highly correlated. This is evidenced in the scatter plot matrix for the analysis variables, which is
shown in Output 1.7.3, created in Example 1.7.
Positive correlation is needed for the alpha coefcient because variables measure a
common entity.
With the ALPHA option, the CORR procedure computes Cronbachs coefcient alpha, which is a lower bound for the reliability coefcient for the raw variables and
the standardized variables.

Example 1.6. Computing Cronbachs Coefcient Alpha

Output 1.5.3. Cronbachs Coefcient Alpha

Fish Measurement Data
Cronbach Coefficient Alpha
Variables
Alpha
---------------------------Raw
0.822134
Standardized
0.985145

Because the variances of some variables vary widely, you should use the standardized
score to estimate reliability. The overall standardized Cronbachs coefcient alpha of
0.985145 provides an acceptable lower bound for the reliability coefcient. This
is much greater than the suggested value of 0.70 given by Nunnally and Bernstein
(1994).
Output 1.5.4. Cronbachs Coefcient Alpha with Deleted Variables
Fish Measurement Data
Cronbach Coefficient Alpha with Deleted Variable
Raw Variables

Standardized Variables

Deleted
Correlation
Correlation
Variable
with Total
Alpha
with Total
Alpha
-----------------------------------------------------------------------Weight3
0.975379
0.783365
0.973464
0.977103
Length3
0.967602
0.881987
0.967177
0.978783
Height
0.964715
0.655098
0.968079
0.978542
Width
0.934635
0.824069
0.937599
0.986626

The standardized alpha coefcient provides information on how each variable reects the reliability of the scale with standardized variables. If the standardized
alpha decreases after removing a variable from the construct, then this variable is
strongly correlated with other variables in the scale. On the other hand, if the standardized alpha increases after removing a variable from the construct, then removing this variable from the scale makes the construct more reliable. The Cronbach
Coefcient Alpha with Deleted Variables table in Output 1.5.4 does not show signicant increase or decrease for the standardized alpha coefcients. See the section Cronbachs Coefcient Alpha on page 24 for more information regarding constructs and Cronbachs alpha.

Chapter 1. The CORR Procedure

Example 1.6. Saving Correlations in an Output Data Set

The following statements compute Pearson correlations and covariances.
title Correlations for a Fitness and Exercise Study;
proc corr data=Fitness nomiss outp=CorrOutp;
var weight oxygen runtime;
run;

The NOMISS option excludes observations with missing values of the VAR statement variables from the analysis. The NOSIMPLE option suppresses the display
of descriptive statistics, and the OUTP= option creates an output data set named
CorrOutp that contains the Pearson correlation statistics. Since the NOMISS option
is specied, the same set of 28 observations is used to compute the correlation for
each pair of variables.
Output 1.6.1. Pearson Correlation Coefcients
Correlations for a Fitness and Exercise Study
The CORR Procedure
Pearson Correlation Coefficients, N = 28
Prob > |r| under H0: Rho=0
Weight

Oxygen

RunTime

Weight

1.00000

-0.18419
0.3481

0.19505
0.3199

Oxygen

-0.18419
0.3481

1.00000

-0.86843
<.0001

0.19505
0.3199

-0.86843
<.0001

1.00000

RunTime

The following statements display the output data set, which is shown in Output 1.6.2.
title Output Data Set from PROC CORR;
proc print data=CorrOutp noobs;
run;

Output 1.6.2. OUTP= Data Set with Pearson Correlations

Output Data Set from PROC CORR
_TYPE_
MEAN
STD
N
CORR
CORR
CORR

_NAME_

Weight
Oxygen
RunTime

Weight

Oxygen

RunTime

77.2168
8.4495
28.0000
1.0000
-0.1842
0.1950

47.1327
5.5535
28.0000
-0.1842
1.0000
-0.8684

10.6954
1.4127
28.0000
0.1950
-0.8684
1.0000

Example 1.7. Creating Scatter Plots

The output data set has the default type CORR and can be used as an input data set
for regression or other statistical procedures. For example, the following statements
request a regression analysis using CorrOutp, without reading the original data in the
REG procedure:
title Input Type CORR Data Set from PROC REG;
proc reg data=CorrOutp;
model runtime= weight oxygen;
run;

The preceding statements generate the same results as the following statements:
proc reg data=Fitness nomiss;
model runtime= weight oxygen;
run;

Example 1.7. Creating Scatter Plots

The following statements request a correlation analysis and a scatter plot matrix for
the variables in the data set Fish1, which was created in Example 1.5. This data
set contains 35 observations, one of which contains a missing value for the variable
Weight3.
ods html;
ods graphics on;
title Fish Measurement Data;
proc corr data=fish1 nomiss plots=matrix;
var Height Width Length3 Weight3;
run;
ods graphics off;
ods html close;

By default, the CORR procedure displays descriptive statistics for the VAR statement
variables, which are shown in Output 1.7.1.
Output 1.7.1. Simple Statistics
Fish Measurement Data
The CORR Procedure
4

Variables:

Height

Width

Length3

Weight3

Simple Statistics
Variable
Height
Width
Length3
Weight3

Mean

Std Dev

Sum

Minimum

Maximum

34
34
34
34

15.22057
5.43805
38.38529
8.44751

1.98159
0.72967
4.21628
0.97574

517.49950
184.89370
1305
287.21524

11.52000
4.02000
30.00000
6.23168

18.95700
6.74970
46.50000
10.00000

Chapter 1. The CORR Procedure

Since the NOMISS option is specied, the same set of 34 observations is used to
compute the correlation for each pair of variables. The correlations are shown in
Output 1.7.2.
Output 1.7.2. Pearson Correlation Coefcients
Fish Measurement Data
Pearson Correlation Coefficients, N = 34
Prob > |r| under H0: Rho=0
Height

Width

Length3

Weight3

Height

1.00000

0.92632
<.0001

0.95492
<.0001

0.96261
<.0001

Width

0.92632
<.0001

1.00000

0.92171
<.0001

0.92789
<.0001

Length3

0.95492
<.0001

0.92171
<.0001

1.00000

0.96523
<.0001

Weight3

0.96261
<.0001

0.92789
<.0001

0.96523
<.0001

1.00000

The variables are highly correlated. For example, the correlation between Height and
Width is 0.92632.
The experimental PLOTS=MATRIX option requests a scatter plot matrix for the VAR
statement variables, which is shown in Output 1.7.3.
In order to create this display, you must specify the experimental ODS GRAPHICS
statement in addition to the PLOTS=MATRIX option. For general information about
ODS graphics, refer to Chapter 15, Statistical Graphics Using ODS (SAS/STAT
Users Guide). For specic information about ODS graphics available in the CORR
procedure, see the section ODS Graphics on page 31.

Example 1.7. Creating Scatter Plots

Output 1.7.3. Scatter Plot Matrix (Experimental)

To explore the correlation between Height and Width, the following statements request a scatter plot with prediction ellipses for the two variables, which is shown in
Output 1.7.4. A prediction ellipse is a region for predicting a new observation from
the population, assuming bivariate normality. It also approximates a region containing a specied percentage of the population.
ods html;
ods graphics on;
proc corr data=fish1 nomiss noprint
plots=scatter(nmaxvar=2 alpha=.20 .30);
var Height Width Length3 Weight3;
run;
ods graphics off;
ods html close;

Chapter 1. The CORR Procedure

The NOMISS option is specied with the original VAR statement to ensure that
the same set of 34 observations is used for this analysis. The experimental
PLOTS=SCATTER(NMAXVAR=2) option requests a scatter plot for the rst two
variables in the VAR list. The ALPHA= suboption requests 80% and 70% prediction
ellipses.
Output 1.7.4. Scatter Plot with Prediction Ellipses (Experimental)

The prediction ellipse is centered at the means (, y ). For further details, see the
x
section Condence and Prediction Ellipses on page 33.
Note that the following statements can also be used to create a scatter plot for Height
and Width:
ods html;
ods graphics on;
proc corr data=fish1 noprint
plots=scatter(alpha=.20 .30);
var Height Width;
run;
ods graphics off;
ods html close;

Example 1.7. Creating Scatter Plots

Output 1.7.5. Scatter Plot with Prediction Ellipses (Experimental)

Output 1.7.5 includes the point (13.9, 5.1), which was excluded from Output 1.7.4
because the observation had a missing value for Weight3. The prediction ellipses in
Output 1.7.5 also reect the inclusion of this observation.
The following statements request a scatter plot with condence ellipses for the mean,
which is shown in Output 1.7.6:
ods html;
ods graphics on;
title Fish Measurement Data;
proc corr data=fish1 nomiss noprint
plots=scatter(ellipse=mean nmaxvar=2 alpha=.05 .01);
var Height Width Length3 Weight3;
run;
ods graphics off;
ods html close;

Chapter 1. The CORR Procedure

The experimental PLOTS=SCATTER option requests scatter plots for all the variables in the VAR statement, and the NMAXVAR=2 suboption restricts the number of
plots created to the rst two variables in the VAR statement. The ELLIPSE=MEAN
and ALPHA= suboptions request 95% and 99% condence ellipses for the mean.
Output 1.7.6. Scatter Plot with Condence Ellipses (Experimental)

The condence ellipse for the mean is centered at the means (, y ). For further
x
details, see the section Condence and Prediction Ellipses on page 33.

Example 1.8. Computing Partial Correlations

A partial correlation measures the strength of the linear relationship between two
variables, while adjusting for the effect of other variables.
The following statements request a partial correlation analysis of variables Height
and Width while adjusting for the variables Length3 and Weight. The latter variables, which are said to be partialled out of the analysis, are specied with the
PARTIAL statement.

Example 1.8. Computing Partial Correlations

ods html;
ods graphics on;
title Fish Measurement Data;
proc corr data=fish1 plots=scatter(alpha=.20 .30);
var Height Width;
partial Length3 Weight3;
run;
ods graphics off;
ods html close;

By default, the CORR procedure displays descriptive statistics for all the variables
and the partial variance and partial standard deviation for the VAR statement variables, as shown in Output 1.8.1.
Output 1.8.1. Descriptive Statistics
Fish Measurement Data
The CORR Procedure
2 Partial Variables:
2
Variables:

Length3
Height

Weight3
Width

Simple Statistics
Variable
Length3
Weight3
Height
Width

Mean

Std Dev

Sum

Minimum

Maximum

34
34
34
34

38.38529
8.44751
15.22057
5.43805

4.21628
0.97574
1.98159
0.72967

1305
287.21524
517.49950
184.89370

30.00000
6.23168
11.52000
4.02000

46.50000
10.00000
18.95700
6.74970

Simple Statistics

Variable

Partial
Variance

Partial
Std Dev

Length3
Weight3
Height
Width

0.26607
0.07315

0.51582
0.27047

When a PARTIAL statement is specied, observations with missing values are excluded from the analysis. The partial correlations for the VAR statement variables
are shown in Output 1.8.2.
The partial correlation between the variables Height and Width is 0.25692, which
is much less than the unpartialled correlation, 0.92632. The p-value for the partial
correlation is 0.1558.
The PLOTS=SCATTER option requests a scatter plot of the residuals for the variables
Height and Width after controlling for the effect of variables Length3 and Weight.

Chapter 1. The CORR Procedure

The ALPHA= suboption requests 80% and 70% prediction ellipses. The scatter plot
is shown in Output 1.8.3.
Output 1.8.2. Pearson Partial Correlation Coefcients
Fish Measurement Data
Pearson Partial Correlation Coefficients, N = 34
Prob > |r| under H0: Partial Rho=0
Height

Width

Height

1.00000

0.25692
0.1558

Width

0.25692
0.1558

1.00000

Output 1.8.3. Partial Residual Scatter Plot (Experimental)

In Output 1.8.3, a standard deviation of Height has roughly the same length on the
X-axis as a standard deviation of Width on the Y-axis. The major axis length is not
signicantly larger than the minor axis length, indicating a weak partial correlation
between Height and Width.

References

References
Anderson, T.W. (1984), An Introduction to Multivariate Statistical Analysis, Second
Edition, New York: John Wiley & Sons.
Blum, J.R., Kiefer, J., and Rosenblatt, M. (1961), Distribution Free Tests
of Independence Based on the Sample Distribution Function, Annals of
Mathematical Statistics , 32, 485498.
Conover, W.J. (1998), Practical Nonparametric Statistics, Third Edition, New York:
John Wiley & Sons, Inc.
Cronbach, L.J. (1951), Coefcient Alpha and the Internal Structure of Tests,
Psychometrika, 16, 297334.
Fisher, R.A. (1915), Frequency Distribution of the Values of the Correlation
Coefcient in Samples from an Indenitely Large Population, Biometrika, 10,
507521.
Fisher, R.A. (1921), On the Probable Error of a Coefcient of Correlation
Deduced from a Small Sample, Metron, 1, 332.
Fisher, R.A. (1936), The Use of Multiple Measurements in Taxonomic Problems,
Annals of Eugenics, 7, 179188.
Fisher, R.A. (1970), Statistical Methods for Research Workers, Fourteenth Edition,
Davien, CT: Hafner Publishing Company.
Hoeffding, W. (1948), A Non-Parametric Test of Independence, Annals of
Mathematical Statistics , 19, 546557.
Hollander, M. and Wolfe, D. (1999), Nonparametric Statistical Methods, Second
Edition, New York: John Wiley & Sons, Inc.
Keeping, E.S. (1962), Introduction to Statistical Inference, New York: D. Van
Nostrand Cimpany, Inc.
Knight, W.E. (1966), A Computer Method for Calculating Kendalls Tau with
Ungrouped Data, Journal of the American Statistical Association , 61, 436439.
Moore, D.S. (2000), Statistics: Concepts and Controversies, Fifth Edition, New
York: W.H. Freeman & Company.
Mudholkar, G.S. (1983), Fishers z-Transformation, Encyclopedia of Statistical
Sciences, 3, 130135.
Noether, G.E. (1967), Elements of Nonparametric Statistics, New York: John Wiley
& Sons, Inc.
Novick, M.R. (1967), Coefcient Alpha and the Reliability of Composite
Measurements, Psychometrika , 32, 113.
Nunnally, J.C. and Bernstein, I.H. (1994), Psychometric Theory, Third Edition, New
York: McGraw-Hill Companies.
Ott, R.L. and Longnecker, M.T. (2000), An Introduction to Statistical Methods and
Data Analysis, Fifth Edition, Belmont, CA: Wadsworth Publishing Company.

Chapter 1. The CORR Procedure

SAS Institute Inc., Measuring the Internal Consistency of a Test, Using PROC
CORR to Compute Cronbachs Coefcient Alpha, SAS Communications , 20:4,
TT2TT5.
Spector, P.E. (1992), Summated Rating Scale Construction: An Introduction,
Newbury Park: Sage.

Chapter 2

The FREQ Procedure

Chapter Contents
OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
GETTING STARTED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Frequency Tables and Statistics . . . . . . . . . . . . . . . . . . . . . . . . 66
Agreement Study Example . . . . . . . . . . . . . . . . . . . . . . . . . . 72
SYNTAX . . . . . . . . .
PROC FREQ Statement
BY Statement . . . . .
EXACT Statement . . .
OUTPUT Statement . .
TABLES Statement . .
TEST Statement . . . .
WEIGHT Statement . .

.
.
.
.
.
.
.
.

74
75
77
77
80
84
96
97

DETAILS . . . . . . . . . . . . . . . . . . . . . . .
Inputting Frequency Counts . . . . . . . . . . . .
Grouping with Formats . . . . . . . . . . . . . .
Missing Values . . . . . . . . . . . . . . . . . . .
Statistical Computations . . . . . . . . . . . . . .
Denitions and Notation . . . . . . . . . . . .
Chi-Square Tests and Statistics . . . . . . . .
Measures of Association . . . . . . . . . . . .
Binomial Proportion . . . . . . . . . . . . . .
Risks and Risk Differences . . . . . . . . . .
Odds Ratio and Relative Risks for 2 x 2 Tables
Cochran-Armitage Test for Trend . . . . . . .
Jonckheere-Terpstra Test . . . . . . . . . . . .
Tests and Measures of Agreement . . . . . . .
Cochran-Mantel-Haenszel Statistics . . . . . .
Exact Statistics . . . . . . . . . . . . . . . . .
Computational Resources . . . . . . . . . . . . .
Output Data Sets . . . . . . . . . . . . . . . . . .
Displayed Output . . . . . . . . . . . . . . . . .
ODS Table Names . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

98
98
99
100
102
102
103
108
118
120
122
124
125
127
134
142
147
148
151
158

EXAMPLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

Chapter 2. The FREQ Procedure

Example 2.1. Creating an Output Data Set with Table Cell Frequencies . . .
Example 2.2. Computing Chi-Square Tests for One-Way Frequency Tables .
Example 2.3. Computing Binomial Proportions for One-Way Frequency
Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Example 2.4. Analyzing a 2x2 Contingency Table . . . . . . . . . . . . . .
Example 2.5. Creating an Output Data Set Containing Chi-Square Statistics
Example 2.6. Computing Cochran-Mantel-Haenszel
Statistics for a Stratied Table . . . . . . . . . . . . . . . . . . . .
Example 2.7. Computing the Cochran-Armitage Trend Test . . . . . . . . .
Example 2.8. Computing Friedmans Chi-Square Statistic . . . . . . . . . .
Example 2.9. Testing Marginal Homogeneity with
Cochrans Q . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

161
164
166
169
172
174
177
180
182

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

Chapter 2

The FREQ Procedure

Overview
The FREQ procedure produces one-way to n-way frequency and crosstabulation
(contingency) tables. For two-way tables, PROC FREQ computes tests and measures
of association. For n-way tables, PROC FREQ does stratied analysis, computing
statistics within, as well as across, strata. Frequencies and statistics can also be output to SAS data sets.
For one-way frequency tables, PROC FREQ can compute statistics to test for equal
proportions, specied proportions, or the binomial proportion. For contingency tables, PROC FREQ can compute various statistics to examine the relationships between two classication variables adjusting for any stratication variables. PROC
FREQ automatically displays the output in a report and can also save the output in a
SAS data set.
For some pairs of variables, you may want to examine the existence or the strength
of any association between the variables. To determine if an association exists, chisquare tests are computed. To estimate the strength of an association, PROC FREQ
computes measures of association that tend to be close to zero when there is no association and close to the maximum (or minimum) value when there is perfect association. The statistics for contingency tables include
chi-square tests and measures
measures of association
risks (binomial proportions) and risk differences for 2 2 tables
odds ratios and relative risks for 2 2 tables
tests for trend
tests and measures of agreement
Cochran-Mantel-Haenszel statistics
PROC FREQ computes asymptotic standard errors, condence intervals, and tests for
measures of association and measures of agreement. Exact p-values and condence
intervals are available for various test statistics and measures. PROC FREQ also
performs stratied analyses that compute statistics within, as well as across, strata for
n-way tables. The statistics include Cochran-Mantel-Haenszel statistics and measures
of agreement.
In choosing measures of association to use in analyzing a two-way table, you should
consider the study design (which indicates whether the row and column variables are
dependent or independent), the measurement scale of the variables (nominal, ordinal,

Chapter 2. The FREQ Procedure

or interval), the type of association that each measure is designed to detect, and any
assumptions required for valid interpretation of a measure. You should exercise care
in selecting measures that are appropriate for your data.
Similar comments apply to the choice and interpretation of the test statistics. For
example, the Mantel-Haenszel chi-square statistic requires an ordinal scale for both
variables and is designed to detect a linear association. The Pearson chi-square, on
the other hand, is appropriate for all variables and can detect any kind of association,
but it is less powerful for detecting a linear association because its power is dispersed
over a greater number of degrees of freedom (except for 2 2 tables).
Several SAS procedures produce frequency counts; only PROC FREQ computes chisquare tests for one-way to n-way tables and measures of association and agreement
for contingency tables. Other procedures to consider for counting are TABULATE,
CHART, and UNIVARIATE. When you want to t models to categorical data, use a
procedure such as CATMOD, GENMOD, LOGISTIC, PHREG, or PROBIT.
For more information on selecting the appropriate statistical analyses, refer to Agresti
(1996) or Stokes, Davis, and Koch (1995).

Getting Started
Frequency Tables and Statistics
The FREQ procedure provides easy access to statistics for testing for association in a
crosstabulation table.
In this example, high school students applied for courses in a summer enrichment
program: these courses included journalism, art history, statistics, graphic arts, and
computer programming. The students accepted were randomly assigned to classes
with and without internships in local companies. The following table contains counts
of the students who enrolled in the summer program by gender and whether they were
assigned an internship slot.
Table 2.1. Summer Enrichment Data

Gender
boys
boys
girls
girls

Internship
yes
no
yes
no

Enrollment
Yes No Total
35 29
64
14 27
41
32 10
42
53 23
76

The SAS data set SummerSchool is created by inputting the summer enrichment
data as cell count data, or providing the frequency count for each combination
of variable values. The following DATA step statements create the SAS data set
SummerSchool.

Frequency Tables and Statistics

data SummerSchool;
input Gender $ Internship
datalines;
boys yes yes 35
boys yes
boys
no yes 14
boys
no
girls yes yes 32
girls yes
girls no yes 53
girls no
;

$ Enrollment $ Count @@;

no
no
no
no

29
27
10
23

The variable Gender takes the values boys or girls, the variable Internship takes
the values yes and no, and the variable Enrollment takes the values yes and
no. The variable Count contains the number of students corresponding to each
combination of data values. The double at sign (@@) indicates that more than one
observation is included on a single data line. In this DATA step, two observations are
included on each line.
Researchers are interested in whether there is an association between internship status
and summer program enrollment. The Pearson chi-square statistic is an appropriate
statistic to assess the association in the corresponding 2 2 table. The following
PROC FREQ statements specify this analysis.
You specify the table for which you want to compute statistics with the TABLES
statement. You specify the statistics you want to compute with options after a slash
(/) in the TABLES statement.
proc freq data=SummerSchool order=data;
weight count;
tables Internship*Enrollment / chisq;
run;

The ORDER= option controls the order in which variable values are displayed in the
rows and columns of the table. By default, the values are arranged according to the
alphanumeric order of their unformatted values. If you specify ORDER=DATA, the
data are displayed in the same order as they occur in the input data set. Here, since
yes appears before no in the data, yes appears rst in any table. Other options for
controlling order include ORDER=FORMATTED, which orders according to the formatted values, and ORDER=FREQUENCY, which orders by descending frequency
count.
In the TABLES statement, Internship*Enrollment species a table where the rows
are internship status and the columns are program enrollment. The CHISQ option
requests chi-square statistics for assessing association between these two variables.
Since the input data are in cell count form, the WEIGHT statement is required. The
WEIGHT statement names the variable Count, which provides the frequency of each
combination of data values.

Chapter 2. The FREQ Procedure

Figure 2.1 presents the crosstabulation of Internship and Enrollment. In each cell,
the values printed under the cell count are the table percentage, row percentage, and
column percentage, respectively. For example, in the rst cell, 63.21 percent of those
offered courses with internships accepted them and 36.79 percent did not.
The FREQ Procedure
Table of Internship by Enrollment
Internship

Enrollment

Frequency|
Percent |
Row Pct |
Col Pct |yes
|no
| Total
---------+--------+--------+
yes
|
67 |
39 |
106
| 30.04 | 17.49 | 47.53
| 63.21 | 36.79 |
| 50.00 | 43.82 |
---------+--------+--------+
no
|
67 |
50 |
117
| 30.04 | 22.42 | 52.47
| 57.26 | 42.74 |
| 50.00 | 56.18 |
---------+--------+--------+
Total
134
89
223
60.09
39.91
100.00

Figure 2.1. Crosstabulation Table

Statistics for Table of Internship by Enrollment

Statistic
DF
Value
Prob
-----------------------------------------------------Chi-Square
1
0.8189
0.3655
Likelihood Ratio Chi-Square
1
0.8202
0.3651
Continuity Adj. Chi-Square
1
0.5899
0.4425
Mantel-Haenszel Chi-Square
1
0.8153
0.3666
Phi Coefficient
0.0606
Contingency Coefficient
0.0605
Cramers V
0.0606

Fishers Exact Test

---------------------------------Cell (1,1) Frequency (F)
67
Left-sided Pr <= F
0.8513
Right-sided Pr >= F
0.2213
Table Probability (P)
Two-sided Pr <= P

0.0726
0.4122

Sample Size = 223

Figure 2.2. Statistics Produced with the CHISQ Option

Frequency Tables and Statistics

Figure 2.2 displays the statistics produced by the CHISQ option. The Pearson chisquare statistic is labeled Chi-Square and has a value of 0.8189 with 1 degree of
freedom. The associated p-value is 0.3655, which means that there is no signicant
evidence of an association between internship status and program enrollment. The
other chi-square statistics have similar values and are asymptotically equivalent. The
other statistics (Phi Coefcient, Contingency Coefcient, and Cramers V ) are measures of association derived from the Pearson chi-square. For Fishers exact test,
the two-sided p-value is 0.4122, which also shows no association between internship
status and program enrollment.
The analysis, so far, has ignored gender. However, it may be of interest to ask whether
program enrollment is associated with internship status after adjusting for gender.
You can address this question by doing an analysis of a set of tables, in this case, by
analyzing the set consisting of one for boys and one for girls. The Cochran-MantelHaenszel statistic is appropriate for this situation: it addresses whether rows and
columns are associated after controlling for the stratication variable. In this case,
you would be stratifying by gender.
The FREQ statements for this analysis are very similar to those for the rst analysis,
except that there is a third variable, Gender, in the TABLES statement. When you
cross more than two variables, the two rightmost variables construct the rows and
columns of the table, respectively, and the leftmost variables determine the stratication.
proc freq data=SummerSchool;
weight count;
tables Gender*Internship*Enrollment / chisq cmh;
run;

This execution of PROC FREQ rst produces two individual crosstabulation tables
of Internship*Enrollment, one for boys and one for girls. Chi-square statistics are
produced for each individual table. Figure 2.3 shows the results for boys. Note that
the chi-square statistic for boys is signicant at the = 0.05 level of signicance.
Boys offered a course with an internship are more likely to enroll than boys who are
not.
If you look at the individual table for girls, displayed in Figure 2.4, you see that
there is no evidence of association for girls between internship offers and program
enrollment.

Chapter 2. The FREQ Procedure

The FREQ Procedure

Table 1 of Internship by Enrollment
Controlling for Gender=boys
Internship

Enrollment

Frequency|
Percent |
Row Pct |
Col Pct |no
|yes
| Total
---------+--------+--------+
no
|
27 |
14 |
41
| 25.71 | 13.33 | 39.05
| 65.85 | 34.15 |
| 48.21 | 28.57 |
---------+--------+--------+
yes
|
29 |
35 |
64
| 27.62 | 33.33 | 60.95
| 45.31 | 54.69 |
| 51.79 | 71.43 |
---------+--------+--------+
Total
56
49
105
53.33
46.67
100.00

Statistics for Table 1 of Internship by Enrollment

Controlling for Gender=boys
Statistic
DF
Value
Prob
-----------------------------------------------------Chi-Square
1
4.2366
0.0396
Likelihood Ratio Chi-Square
1
4.2903
0.0383
Continuity Adj. Chi-Square
1
3.4515
0.0632
Mantel-Haenszel Chi-Square
1
4.1963
0.0405
Phi Coefficient
0.2009
Contingency Coefficient
0.1969
Cramers V
0.2009

Fishers Exact Test

---------------------------------Cell (1,1) Frequency (F)
27
Left-sided Pr <= F
0.9885
Right-sided Pr >= F
0.0311
Table Probability (P)
Two-sided Pr <= P

0.0196
0.0467

Sample Size = 105

Figure 2.3. Crosstabulation Table and Statistics for Boys

Frequency Tables and Statistics

Table 2 of Internship by Enrollment

Controlling for Gender=girls
Internship

Enrollment

Frequency|
Percent |
Row Pct |
Col Pct |no
|yes
| Total
---------+--------+--------+
no
|
23 |
53 |
76
| 19.49 | 44.92 | 64.41
| 30.26 | 69.74 |
| 69.70 | 62.35 |
---------+--------+--------+
yes
|
10 |
32 |
42
|
8.47 | 27.12 | 35.59
| 23.81 | 76.19 |
| 30.30 | 37.65 |
---------+--------+--------+
Total
33
85
118
27.97
72.03
100.00

Statistics for Table 2 of Internship by Enrollment

Controlling for Gender=girls
Statistic
DF
Value
Prob
-----------------------------------------------------Chi-Square
1
0.5593
0.4546
Likelihood Ratio Chi-Square
1
0.5681
0.4510
Continuity Adj. Chi-Square
1
0.2848
0.5936
Mantel-Haenszel Chi-Square
1
0.5545
0.4565
Phi Coefficient
0.0688
Contingency Coefficient
0.0687
Cramers V
0.0688

Fishers Exact Test

---------------------------------Cell (1,1) Frequency (F)
23
Left-sided Pr <= F
0.8317
Right-sided Pr >= F
0.2994
Table Probability (P)
Two-sided Pr <= P

0.1311
0.5245

Sample Size = 118

Figure 2.4. Crosstabulation Table and Statistics for Girls

Chapter 2. The FREQ Procedure

These individual table results demonstrate the occasional problems with combining
information into one table and not accounting for information in other variables such
as Gender. Figure 2.5 contains the CMH results. There are three summary (CMH)
statistics; which one you use depends on whether your rows and/or columns have
an order in r c tables. However, in the case of 2 2 tables, ordering does not
matter and all three statistics take the same value. The CMH statistic follows the chisquare distribution under the hypothesis of no association, and here, it takes the value
4.0186 with 1 degree of freedom. The associated p-value is 0.0450, which indicates
a signicant association at the = 0.05 level.
Thus, when you adjust for the effect of gender in these data, there is an association
between internship and program enrollment. But, if you ignore gender, no association
is found. Note that the CMH option also produces other statistics, including estimates
and condence limits for relative risk and odds ratios for 22 tables and the BreslowDay Test. These results are not displayed here.
Summary Statistics for Internship by Enrollment
Controlling for Gender
Cochran-Mantel-Haenszel Statistics (Based on Table Scores)
Statistic
Alternative Hypothesis
DF
Value
Prob
--------------------------------------------------------------1
Nonzero Correlation
1
4.0186
0.0450
2
Row Mean Scores Differ
1
4.0186
0.0450
3
General Association
1
4.0186
0.0450
Total Sample Size = 223

Figure 2.5. Test for the Hypothesis of No Association

Agreement Study Example

Medical researchers are interested in evaluating the efcacy of a new treatment for a
skin condition. Dermatologists from participating clinics were trained to conduct the
study and to evaluate the condition. After the training, two dermatologists examined
patients with the skin condition from a pilot study and rated the same patients. The
possible evaluations are terrible, poor, marginal, and clear. Table 2.2 contains the
data.

Agreement Study Example

Table 2.2. Skin Condition Data

Dermatologist 1
Terrible
Poor
Marginal
Clear

Dermatologist 2
Terrible Poor Marginal
10
4
1
5
10
12
2
4
12
0
2
6

Clear
0
2
5
13

The dermatologists evaluations of the patients are contained in the variables derm1
and derm2; the variable count is the number of patients given a particular pair of
ratings. In order to evaluate the agreement of the diagnoses (a possible contribution
to measurement error in the study), the kappa coefcient is computed. You specify
the AGREE option in the TABLES statement and use the TEST statement to request
a test for the null hypothesis that their agreement is purely by chance. You specify
the keyword KAPPA to perform this test for the kappa coefcient. The results are
shown in Figure 2.6.
data SkinCondition;
input derm1 $ derm2 $ count;
datalines;
terrible terrible 10
terrible
poor 4
terrible marginal 1
terrible
clear 0
poor
terrible 5
poor
poor 10
poor
marginal 12
poor
clear 2
marginal terrible 2
marginal
poor 4
marginal marginal 12
marginal
clear 5
clear
terrible 0
clear
poor 2
clear
marginal 6
clear
clear 13
;

proc freq data=SkinCondition order=data;

weight count;
tables derm1*derm2 / agree noprint;
test kappa;
run;

Chapter 2. The FREQ Procedure

The FREQ Procedure

Statistics for Table of derm1 by derm2
Simple Kappa Coefficient
-------------------------------Kappa
0.3449
ASE
0.0724
95% Lower Conf Limit
0.2030
95% Upper Conf Limit
0.4868
Test of H0: Kappa = 0
ASE under H0
Z
One-sided Pr > Z
Two-sided Pr > |Z|

0.0612
5.6366
<.0001
<.0001

Sample Size = 88

Figure 2.6. Agreement Study

The kappa coefcient has the value 0.3449, which indicates slight agreement between the dermatologists, and the hypothesis test conrms that you can reject the
null hypothesis of no agreement. This conclusion is further supported by the condence interval of (0.2030, 0.4868), which suggests that the true kappa is greater than
zero. The AGREE option also produces Bowkers test for symmetry and the weighted
kappa coefcient, but that output is not shown.

Syntax
The following statements are available in PROC FREQ.

PROC FREQ < options > ;

BY variables ;
EXACT statistic-options < / computation-options > ;
OUTPUT < OUT=SAS-data-set > options ;
TABLES requests < / options > ;
TEST options ;
WEIGHT variable < / option > ;
The PROC FREQ statement is the only required statement for the FREQ procedure.
If you specify the following statements, PROC FREQ produces a one-way frequency
table for each variable in the most recently created data set.
proc freq;
run;

The rest of this section gives detailed syntax information for the BY, EXACT,
OUTPUT, TABLES, TEST, and WEIGHT statements in alphabetical order after the
description of the PROC FREQ statement. Table 2.3 summarizes the basic functions
of each statement.

PROC FREQ Statement

Table 2.3. Summary of PROC FREQ Statements

Statement
BY
EXACT
OUTPUT
TABLES
TEST
WEIGHT

Description
calculates separate frequency or crosstabulation tables for each BY
group.
requests exact tests for specied statistics.
creates an output data set that contains specied statistics.
species frequency or crosstabulation tables and requests tests and
measures of association.
requests asymptotic tests for measures of association and agreement.
identies a variable with values that weight each observation.

PROC FREQ Statement

PROC FREQ < options > ;
The PROC FREQ statement invokes the procedure.
The following table lists the options available in the PROC FREQ statement.
Descriptions follow in alphabetical order.
Table 2.4. PROC FREQ Statement Options

Option
DATA=
COMPRESS
FORMCHAR=
NLEVELS
NOPRINT
ORDER=
PAGE

Description
species the input data set.
begins the next one-way table on the current page
species the outline and cell divider characters for the cells of the
crosstabulation table.
displays the number of levels for all TABLES variables
suppresses all displayed output.
species the order for listing variable values.
displays one table per page.

You can specify the following options in the PROC FREQ statement.
COMPRESS

begins display of the next one-way frequency table on the same page as the preceding
one-way table if there is enough space to begin the table. By default, the next oneway table begins on the current page only if the entire table ts on that page. The
COMPRESS option is not valid with the PAGE option.
DATA=SAS-data-set

names the SAS data set to be analyzed by PROC FREQ. If you omit the DATA=
option, the procedure uses the most recently created SAS data set.
FORMCHAR (1,2,7) =formchar-string

denes the characters to be used for constructing the outlines and dividers for the
cells of contingency tables. The FORMCHAR= option can specify 20 different SAS
formatting characters used to display output; however, PROC FREQ uses only the
rst, second, and seventh formatting characters. Therefore, the proper specication
for PROC FREQ is FORMCHAR(1,2,7)= formchar-string. The formchar-string

Chapter 2. The FREQ Procedure

should be three characters long. The characters are used to denote (1) vertical separator, (2) horizontal separator, and (7) vertical-horizontal intersection. You can use
any character in formchar-string, including hexadecimal characters. If you use hexadecimal characters, you must put an x after the closing quote. For information on
which hexadecimal codes to use for which characters, consult the documentation for
your hardware.
Specifying all blanks for formchar-string produces tables with no outlines or dividers:
formchar (1,2,7)=

If you do not specify the FORMCHAR= option, PROC FREQ uses the default
formchar (1,2,7)=|-+

Refer to the CALENDAR, PLOT, and TABULATE procedures in the Base SAS 9.1
Procedures Guide for more information on form characters.
Table 2.5. Formatting Characters Used by PROC FREQ

Position
1
2
7

Default
|
+

Used to Draw
vertical separators
horizontal separators
intersections of vertical and horizontal separators

NLEVELS

displays the Number of Variable Levels table. This table provides the number of
levels for each variable named in the TABLES statements. See the section Number
of Variable Levels Table on page 151 for more information. PROC FREQ determines the variable levels from the formatted variable values, as described in the section Grouping with Formats on page 99.
NOPRINT

suppresses the display of all output. Note that this option temporarily disables the
Output Delivery System (ODS). For more information, see Chapter 14, Using the
Output Delivery System. (SAS/STAT Users Guide) .
Note: A NOPRINT option is also available in the TABLES statement. It suppresses
display of the crosstabulation tables but allows display of the requested statistics.
ORDER=DATA | FORMATTED | FREQ | INTERNAL

species the order in which the values of the frequency and crosstabulation table
variables are to be reported. The following table shows how PROC FREQ interprets
values of the ORDER= option.
DATA

orders values according to their order in the input data set.

FORMATTED

orders values by their formatted values. This order is operatingenvironment dependent. By default, the order is ascending.

FREQ

orders values by descending frequency count.

INTERNAL

orders values by their unformatted values, which yields the same

order that the SORT procedure does. This order is operatingenvironment dependent.

By default, ORDER=INTERNAL. The ORDER= option does not apply to missing

values, which are always ordered rst.

EXACT Statement
PAGE

displays only one table per page. Otherwise, PROC FREQ displays multiple tables
per page as space permits. The PAGE option is not valid with the COMPRESS option.

BY Statement
BY variables ;
You can specify a BY statement with PROC FREQ to obtain separate analyses on
observations in groups dened by the BY variables. When a BY statement appears,
the procedure expects the input data set to be sorted in order of the BY variables.
If your input data set is not sorted in ascending order, use one of the following alternatives:
Sort the data using the SORT procedure with a similar BY statement.
Specify the BY statement option NOTSORTED or DESCENDING in the BY
statement for the FREQ procedure. The NOTSORTED option does not mean
that the data are unsorted but rather that the data are arranged in groups (according to values of the BY variables) and that these groups are not necessarily
in alphabetical or increasing numeric order.
Create an index on the BY variables using the DATASETS procedure.
For more information on the BY statement, refer to the discussion in SAS Language
Reference: Concepts. For more information on the DATASETS procedure, refer to
the discussion in the Base SAS 9.1 Procedures Guide.

EXACT Statement
EXACT statistic-options < / computation-options > ;
The EXACT statement requests exact tests or condence limits for the specied
statistics. Optionally, PROC FREQ computes Monte Carlo estimates of the exact
p-values. The statistic-options specify the statistics for which to provide exact tests
or condence limits. The computation-options specify options for the computation
of exact statistics.
CAUTION: PROC FREQ computes exact tests with fast and efcient algorithms
that are superior to direct enumeration. Exact tests are appropriate when a data set
is small, sparse, skewed, or heavily tied. For some large problems, computation of
exact tests may require a large amount of time and memory. Consider using asymptotic tests for such problems. Alternatively, when asymptotic methods may not be
sufcient for such large problems, consider using Monte Carlo estimation of exact
p-values. See the section Computational Resources on page 145 for more information.

Chapter 2. The FREQ Procedure

Statistic-Options
The statistic-options specify the statistics for which exact tests or condence limits are computed. PROC FREQ can compute exact p-values for the following
hypothesis tests: chi-square goodness-of-t test for one-way tables; Pearson chisquare, likelihood-ratio chi-square, Mantel-Haenszel chi-square, Fishers exact test,
Jonckheere-Terpstra test, Cochran-Armitage test for trend, and McNemars test for
two-way tables. PROC FREQ can also compute exact p-values for tests of the following statistics: Pearson correlation coefcient, Spearman correlation coefcient,
simple kappa coefcient, weighted kappa coefcient, and common odds ratio. PROC
FREQ can compute exact p-values for the binomial proportion test for one-way tables, as well as exact condence limits for the binomial proportion. Additionally,
PROC FREQ can compute exact condence limits for the odds ratio for 2 2 tables,
as well as exact condence limits for the common odds ratio for stratied 22 tables.
Table 2.6 lists the available statistic-options and the exact statistics computed. Most
of the option names are identical to the corresponding options in the TABLES statement and the OUTPUT statement. You can request exact computations for groups
of statistics by using options that are identical to the following TABLES statement
options: CHISQ, MEASURES, and AGREE. For example, when you specify the
CHISQ option in the EXACT statement, PROC FREQ computes exact p-values for
the Pearson chi-square, likelihood-ratio chi-square, and Mantel-Haenszel chi-square
tests. You request exact p-values for an individual test by specifying one of the
statistic-options shown in Table 2.6 .
Table 2.6. EXACT Statement Statistic-Options

Option
AGREE
BINOMIAL
CHISQ

COMOR
FISHER
JT
KAPPA
LRCHI
MCNEM
MEASURES
MHCHI
OR
PCHI
PCORR
SCORR
TREND
WTKAP

Exact Statistics Computed

McNemars test for 2 2 tables, simple kappa coefcient, and
weighted kappa coefcient
binomial proportion test for one-way tables
chi-square goodness-of-t test for one-way tables; Pearson chisquare, likelihood-ratio chi-square, and Mantel-Haenszel chisquare tests for two-way tables
condence limits for the common odds ratio for h 2 2 tables;
common odds ratio test
Fishers exact test
Jonckheere-Terpstra test
test for the simple kappa coefcient
likelihood-ratio chi-square test
McNemars test
tests for the Pearson correlation and the Spearman correlation, and
the odds ratio condence limits for 2 2 tables
Mantel-Haenszel chi-square test
condence limits for the odds ratio for 2 2 tables
Pearson chi-square test
test for the Pearson correlation coefcient
test for the Spearman correlation coefcient
Cochran-Armitage test for trend
test for the weighted kappa coefcient

EXACT Statement
Computation-Options
The computation-options specify options for computation of exact statistics. You can
specify the following computation-options in the EXACT statement.
ALPHA=
species the level of the condence limits for Monte Carlo p-value estimates. The
value of the ALPHA= option must be between 0 and 1, and the default is 0.01.
A condence level of produces 100(1 )% condence limits. The default of
ALPHA=.01 produces 99% condence limits for the Monte Carlo estimates. The
ALPHA= option invokes the MC option.
MAXTIME=value

species the maximum clock time (in seconds) that PROC FREQ can use to compute
an exact p-value. If the procedure does not complete the computation within the specied time, the computation terminates. The value of the MAXTIME= option must be
a positive number. The MAXTIME= option is valid for Monte Carlo estimation of
exact p-values, as well as for direct exact p-value computation.
See the section Computational Resources on page 145 for more information.
MC

requests Monte Carlo estimation of exact p-values instead of direct exact p-value
computation. Monte Carlo estimation can be useful for large problems that require a
great amount of time and memory for exact computations but for which asymptotic
approximations may not be sufcient. See the section Monte Carlo Estimation on
page 146 for more information.
The MC option is available for all EXACT statistic-options except BINOMIAL,
COMOR, MCNEM, and OR. PROC FREQ computes only exact tests or condence
limits for those statistics.
The ALPHA=, N=, and SEED= options also invoke the MC option.
N=n

species the number of samples for Monte Carlo estimation. The value of the N=
option must be a positive integer, and the default is 10000 samples. Larger values
of n produce more precise estimates of exact p-values. Because larger values of n
generate more samples, the computation time increases. The N= option invokes the
MC option.
POINT

requests exact point probabilities for the test statistics.

The POINT option is available for all the EXACT statement statistic-options except
the OR option, which provides exact condence limits as opposed to an exact test.
The POINT option is not available with the MC option.
SEED=number

species the initial seed for random number generation for Monte Carlo estimation.
The value of the SEED= option must be an integer. If you do not specify the SEED=
option, or if the SEED= value is negative or zero, PROC FREQ uses the time of day
from the computers clock to obtain the initial seed. The SEED= option invokes the
MC option.

Chapter 2. The FREQ Procedure

Using TABLES Statement Options with the EXACT Statement

If you use only one TABLES statement, you do not need to specify options in the
TABLES statement that are identical to options appearing in the EXACT statement.
PROC FREQ automatically invokes the corresponding TABLES statement option
when you specify the option in the EXACT statement. However, when you use multiple TABLES statements and want exact computations, you must specify options in
the TABLES statement to compute the desired statistics. PROC FREQ then performs
exact computations for all statistics that are also specied in the EXACT statement.

OUTPUT Statement
OUTPUT < OUT= SAS-data-set > options ;
The OUTPUT statement creates a SAS data set containing statistics computed by
PROC FREQ. The variables contain statistics for each two-way table or stratum, as
well as summary statistics across all strata.
Only one OUTPUT statement is allowed for each execution of PROC FREQ. You
must specify a TABLES statement with the OUTPUT statement. If you use multiple
TABLES statements, the contents of the OUTPUT data set correspond to the last
TABLES statement. If you use multiple table requests in a TABLES statement, the
contents of the OUTPUT data set correspond to the last table request.
For more information, see the section Output Data Sets on page 148.
Note that you can use the Output Delivery System (ODS) to create a SAS data set
from any piece of PROC FREQ output. For more information, see Table 2.11 on page
159 and Chapter 14, Using the Output Delivery System. (SAS/STAT Users Guide)
You can specify the following options in an OUTPUT statement.
OUT=SAS-data-set

names the output data set. If you omit the OUT= option, the data set is named DATAn,
where n is the smallest integer that makes the name unique.
options

specify the statistics that you want in the output data set. Available statistics are those
produced by PROC FREQ for each one-way or two-way table, as well as the summary
statistics across all strata. When you request a statistic, the OUTPUT data set contains
that estimate or test statistic plus any associated standard error, condence limits, pvalues, and degrees of freedom. You can output statistics by using group options
identical to those specied in the TABLES statement: AGREE, ALL, CHISQ, CMH,
and MEASURES. Alternatively, you can request an individual statistic by specifying
one of the options shown in the following table.

OUTPUT Statement
Table 2.7. OUTPUT Statement Options and Required TABLES Statement Options

Option

Output Data Set Statistics

AGREE

McNemars test for 2 2 tables, simple

kappa coefcient, and weighted kappa
coefcient; for square tables with more
than two response categories, Bowkers
test of symmetry; for multiple strata,
overall simple and weighted kappa statistics, and tests for equal kappas among
strata; for multiple strata with two response categories, Cochrans Q test
continuity-adjusted chi-square for 2 2
tables
all statistics under CHISQ, MEASURES,
and CMH, and the number of nonmissing
subjects
Breslow-Day test

AJCHI
ALL

BDCHI
BIN | BINOMIAL
CHISQ

CMH

CMH1

CMH2
CMHCOR
CMHGA
CMHRMS
COCHQ

for one-way tables, binomial proportion

statistics
chi-square goodness-of-t test for oneway tables; for two-way tables, Pearson
chi-square, likelihood-ratio chi-square,
continuity-adjusted chi-square for 2
2 tables, Mantel-Haenszel chi-square,
Fishers exact test for 2 2 tables, phi
coefcient, contingency coefcient, and
Cramers V
Cochran-Mantel-Haenszel correlation,
row mean scores (ANOVA), and general
association statistics; for 2 2 tables,
logit and Mantel-Haenszel adjusted odds
ratios, relative risks, and Breslow-Day
test
same as CMH, but excludes general association and row mean scores (ANOVA)
statistics
same as CMH, but excludes the general
association statistic
Cochran-Mantel-Haenszel correlation
statistic
Cochran-Mantel-Haenszel general association statistic
Cochran-Mantel-Haenszel row mean
scores (ANOVA) statistic
Cochrans Q

Required TABLES
Statement Option
AGREE

ALL or CHISQ
ALL

ALL or CMH or CMH1

or CMH2
BINOMIAL
ALL or CHISQ

ALL or CMH

ALL or CMH or CMH1

ALL or CMH or CMH2

ALL or CMH or CMH1
or CMH2
ALL or CMH
ALL or CMH or CMH2
AGREE

Chapter 2. The FREQ Procedure

Table 2.7. (continued)

Option

Output Data Set Statistics

CONTGY
CRAMV
EQKAP
EQWKP
FISHER | EXACT
GAMMA
JT
KAPPA
KENTB
LAMCR
LAMDAS
LAMRC
LGOR

contingency coefcient
Cramers V
test for equal simple kappas
test for equal weighted kappas
Fishers exact test
gamma
Jonckheere-Terpstra test
simple kappa coefcient
Kendalls tau-b
lambda asymmetric (C|R)
lambda symmetric
lambda asymmetric (R|C)
adjusted logit odds ratio

LGRRC1

adjusted column 1 logit relative risk

LGRRC2

adjusted column 2 logit relative risk

LRCHI
MCNEM
MEASURES

likelihood-ratio chi-square
McNemars test
gamma, Kendalls tau-b, Stuarts
tau-c, Somers D(C|R), Somers
D(R|C), Pearson correlation coefcient,
Spearman correlation coefcient, lambda
asymmetric (C|R), lambda asymmetric
(R|C), lambda symmetric, uncertainty
coefcient (C|R), uncertainty coefcient
(R|C), and symmetric uncertainty coefcient; for 2 2 tables, odds ratio and
relative risks
Mantel-Haenszel chi-square
adjusted Mantel-Haenszel odds ratio

MHCHI
MHOR
MHRRC1
MHRRC2
N
NMISS

adjusted column 1 Mantel-Haenszel relative risk

adjusted column 2 Mantel-Haenszel relative risk
number of nonmissing subjects for the
stratum
number of missing subjects for the stratum

Required TABLES
Statement Option
ALL or CHISQ
ALL or CHISQ
AGREE
AGREE
ALL or CHISQ
ALL or MEASURES
JT
AGREE
ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
ALL or CMH or CMH1
or CMH2
ALL or CMH or CMH1
or CMH2
ALL or CMH or CMH1
or CMH2
ALL or CHISQ
AGREE
ALL or MEASURES

ALL or CHISQ
ALL or CMH or CMH1
or CMH2
ALL or CMH or CMH1
or CMH2
ALL or CMH or CMH1
or CMH2

ALL and CHISQ compute Fishers exact test for 2 2 tables. Use the FISHER option to compute
Fishers exact test for general rxc tables.

OUTPUT Statement
Table 2.7. (continued)

Option

Output Data Set Statistics

odds ratio

PCHI

RISKDIFF
RISKDIFF1
RISKDIFF2
RRC1

chi-square goodness-of-t test for oneway tables; for two-way tables, Pearson
chi-square
Pearson correlation coefcient
phi coefcient
polychoric correlation coefcient
column 1 risk difference (row 1 - row 2)
column 2 risk difference (row 1 - row 2)
odds ratio and relative risks for 2 2 tables
risks and risk differences
column 1 risks and risk difference
column 2 risks and risk difference
column 1 relative risk

RRC2

column 2 relative risk

RSK1
RSK11
RSK12
RSK2
RSK21
RSK22
SCORR
SMDCR
SMDRC
STUTC
TREND
TSYMM
U
UCR
URC
WTKAP

column 1 risk (overall)

column 1 risk, for row 1
column 2 risk, for row 1
column 2 risk (overall)
column 1 risk, for row 2
column 2 risk, for row 2
Spearman correlation coefcient
Somers D(C|R)
Somers D(R|C)
Stuarts tau-c
Cochran-Armitage test for trend
Bowkers test of symmetry
symmetric uncertainty coefcient
uncertainty coefcient (C|R)
uncertainty coefcient (R|C)
weighted kappa coefcient

PCORR
PHI
PLCORR
RDIF1
RDIF2
RELRISK

Required TABLES
Statement Option
ALL or MEASURES
or RELRISK
ALL or CHISQ

ALL or MEASURES
ALL or CHISQ
PLCORR
RISKDIFF
RISKDIFF
ALL or MEASURES
or RELRISK
RISKDIFF
RISKDIFF
RISKDIFF
ALL or MEASURES
or RELRISK
ALL or MEASURES
or RELRISK
RISKDIFF
RISKDIFF
RISKDIFF
RISKDIFF
RISKDIFF
RISKDIFF
ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
TREND
AGREE
ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
AGREE

Using the TABLES Statement with the OUTPUT Statement

In order to specify that the OUTPUT data set contain a particular statistic, you must
have PROC FREQ compute the statistic by using the corresponding option in the
TABLES statement or the EXACT statement. For example, you cannot specify the
option PCHI (Pearson chi-square) in the OUTPUT statement without also specifying
a TABLES statement option or an EXACT statement option to compute the Pearson
chi-square. The TABLES statement option ALL or CHISQ computes the Pearson chisquare. Additionally, if you have only one TABLES statement, the EXACT statement
option CHISQ or PCHI computes the Pearson chi-square.

Chapter 2. The FREQ Procedure

TABLES Statement
TABLES requests < / options > ;
The TABLES statement requests one-way to n-way frequency and crosstabulation
tables and statistics for those tables.
If you omit the TABLES statement, PROC FREQ generates one-way frequency tables
for all data set variables that are not listed in the other statements.
The following argument is required in the TABLES statement.
requests

specify the frequency and crosstabulation tables to produce. A request is composed

of one variable name or several variable names separated by asterisks. To request a
one-way frequency table, use a single variable. To request a two-way crosstabulation
table, use an asterisk between two variables. To request a multiway table (an n-way
table, where n>2), separate the desired variables with asterisks. The unique values of
these variables form the rows, columns, and strata of the table.
For two-way to multiway tables, the values of the last variable form the crosstabulation table columns, while the values of the next-to-last variable form the rows.
Each level (or combination of levels) of the other variables forms one stratum. PROC
FREQ produces a separate crosstabulation table for each stratum. For example, a
specication of A*B*C*D in a TABLES statement produces k tables, where k is the
number of different combinations of values for A and B. Each table lists the values
for C down the side and the values for D across the top.
You can use multiple TABLES statements in the PROC FREQ step. PROC FREQ
builds all the table requests in one pass of the data, so that there is essentially no
loss of efciency. You can also specify any number of table requests in a single
TABLES statement. To specify multiple table requests quickly, use a grouping syntax
by placing parentheses around several variables and joining other variables or variable
combinations. For example, the following statements illustrate grouping syntax.
Table 2.8. Grouping Syntax

Request
tables A*(B C);
tables (A B)*(C D);
tables (A B C)*D;
tables A C;
tables (A C)*D;

Equivalent to
tables A*B A*C;
tables A*C B*C A*D B*D;
tables A*D B*D C*D;
tables A B C;
tables A*D B*D C*D;

Without Options
If you request a one-way frequency table for a variable without specifying options,
PROC FREQ produces frequencies, cumulative frequencies, percentages of the total
frequency, and cumulative percentages for each value of the variable. If you request a
two-way or an n-way crosstabulation table without specifying options, PROC FREQ
produces crosstabulation tables that include cell frequencies, cell percentages of the
total frequency, cell percentages of row frequencies, and cell percentages of column

TABLES Statement
frequencies. The procedure excludes observations with missing values from the table
but displays the total frequency of missing observations below each table.

Options
The following table lists the options available with the TABLES statement.
Descriptions follow in alphabetical order.
Table 2.9. TABLES Statement Options

Option
Description
Control Statistical Analysis
AGREE
requests tests and measures of classication agreement
ALL
requests tests and measures of association produced by CHISQ,
MEASURES, and CMH
ALPHA=
sets the condence level for condence limits
BDT
requests Tarones adjustment for the Breslow-Day test
BINOMIAL
requests binomial proportion, condence limits and test for oneway tables
BINOMIALC
requests BINOMIAL statistics with a continuity correction
CHISQ
requests chi-square tests and measures of association based on chisquare
CL
requests condence limits for the MEASURES statistics
CMH
requests all Cochran-Mantel-Haenszel statistics
CMH1
requests the CMH correlation statistic, and adjusted relative risks
and odds ratios
CMH2
requests CMH correlation and row mean scores (ANOVA) statistics, and adjusted relative risks and odds ratios
CONVERGE=
species convergence criterion to compute polychoric correlation
FISHER
requests Fishers exact test for tables larger than 2 2
JT
requests Jonckheere-Terpstra test
MAXITER=
species maximum number of iterations to compute polychoric
correlation
MEASURES
requests measures of association and their asymptotic standard errors
MISSING
treats missing values as nonmissing
PLCORR
requests polychoric correlation
RELRISK
requests relative risk measures for 2 2 tables
RISKDIFF
requests risks and risk differences for 2 2 tables
RISKDIFFC
requests RISKDIFF statistics with a continuity correction
SCORES=
species the type of row and column scores
TESTF=
species expected frequencies for a one-way table chi-square test
TESTP=
species expected proportions for a one-way table chi-square test
TREND
requests Cochran-Armitage test for trend

Chapter 2. The FREQ Procedure

Table 2.9. (continued)

Option
Description
Control Additional Table Information
CELLCHI2
displays each cells contribution to the total Pearson chi-square
statistic
CUMCOL
displays the cumulative column percentage in each cell
DEVIATION
displays the deviation of the cell frequency from the expected
value for each cell
EXPECTED
displays the expected cell frequency for each cell
MISSPRINT
displays missing value frequencies
SPARSE
lists all possible combinations of variable levels even when a combination does not occur
TOTPCT
displays percentage of total frequency on n-way tables when n > 2
Control Displayed Output
CONTENTS=
species the HTML contents link for crosstabulation tables
CROSSLIST
displays crosstabulation tables in ODS column format
FORMAT=
formats the frequencies in crosstabulation tables
LIST
displays two-way to n-way tables in list format
NOCOL
suppresses display of the column percentage for each cell
NOCUM
suppresses display of cumulative frequencies and cumulative percentages in one-way frequency tables and in list format
NOFREQ
suppresses display of the frequency count for each cell
NOPERCENT
suppresses display of the percentage, row percentage, and column
percentage in crosstabulation tables, or percentages and cumulative percentages in one-way frequency tables and in list format
NOPRINT
suppresses display of tables but displays statistics
NOROW
suppresses display of the row percentage for each cell
NOSPARSE
suppresses zero cell frequencies in the list display and in the OUT=
data set when ZEROS is specied
NOWARN
suppresses log warning message for the chi-square test
PRINTKWT
displays kappa coefcient weights
SCOROUT
displays the row and the column scores
Create an Output Data Set
OUT=
species an output data set to contain variable values and frequency counts
OUTCUM
includes the cumulative frequency and cumulative percentage in
the output data set for one-way tables
OUTEXPECT
includes the expected frequency of each cell in the output data set
OUTPCT
includes the percentage of column frequency, row frequency, and
two-way table frequency in the output data set

TABLES Statement
You can specify the following options in a TABLES statement.
AGREE < (WT=FC) >

requests tests and measures of classication agreement for square tables. The
AGREE option provides McNemars test for 2 2 tables and Bowkers test of symmetry for tables with more than two response categories. The AGREE option also
produces the simple kappa coefcient, the weighted kappa coefcient, the asymptotic
standard errors for the simple and weighted kappas, and the corresponding condence
limits. When there are multiple strata, the AGREE option provides overall simple and
weighted kappas as well as tests for equal kappas among strata. When there are multiple strata and two response categories, PROC FREQ computes Cochrans Q test.
For more information, see the section Tests and Measures of Agreement on page
127.
The (WT=FC) specication requests that PROC FREQ use Fleiss-Cohen weights to
compute the weighted kappa coefcient. By default, PROC FREQ uses CicchettiAllison weights. See the section Weighted Kappa Coefcient on page 130 for
more information. You can specify the PRINTKWT option to display the kappa
coefcient weights.
AGREE statistics are computed only for square tables, where the number of rows
equals the number of columns. If your table is not square due to observations with
zero weights, you can use the ZEROS option in the WEIGHT statement to include
these observations. For more details, see the section Tables with Zero Rows and
Columns on page 133.
ALL

requests all of the tests and measures that are computed by the CHISQ, MEASURES,
and CMH options. The number of CMH statistics computed can be controlled by the
CMH1 and CMH2 options.
ALPHA=

species the level of condence limits. The value of the ALPHA= option must be between 0 and 1, and the default is 0.05. A condence level of produces 100(1 )%
condence limits. The default of ALPHA=0.05 produces 95% condence limits.
ALPHA= applies to condence limits requested by TABLES statement options.
There is a separate ALPHA= option in the EXACT statement that sets the level of
condence limits for Monte Carlo estimates of exact p-values, which are requested
in the EXACT statement.
BDT

requests Tarones adjustment in the Breslow-Day test for homogeneity of odds ratios.
(You must specify the CMH option to compute the Breslow-Day test.) See the section Breslow-Day Test for Homogeneity of the Odds Ratios on page 142 for more
information.
BINOMIAL < (P= value) | (LEVEL= level-number | level-value) >

requests the binomial proportion for one-way tables. The BINOMIAL option also
provides the asymptotic standard error, asymptotic and exact condence intervals,

Chapter 2. The FREQ Procedure

and the asymptotic test for the binomial proportion. To request an exact test for the
binomial proportion, use the BINOMIAL option in the EXACT statement.
To specify the null hypothesis proportion for the test, use P=. If you omit P=value,
PROC FREQ uses 0.5 as the default for the test. By default, BINOMIAL computes
the proportion of observations for the rst variable level that appears in the output.
To specify a different level, use LEVEL=level-number or LEVEL=level-value, where
level-number is the variable levels number or order in the output, and level-value is
the formatted value of the variable level.
To include a continuity correction in the asymptotic condence interval and test, use
the BINOMIALC option instead of the BINOMIAL option.
See the section Binomial Proportion on page 118 for more information.
BINOMIALC < (P= value) | (LEVEL= level-number | level-value) >

requests the BINOMIAL option statistics for one-way tables, and includes a continuity correction in the asymptotic condence interval and the asymptotic test. The
BINOMIAL option statistics include the binomial proportion, the asymptotic standard error, asymptotic and exact condence intervals, and the asymptotic test for the
binomial proportion. To request an exact test for the binomial proportion, use the
BINOMIAL option in the EXACT statement.
To specify the null hypothesis proportion for the test, use P=. If you omit P=value,
PROC FREQ uses 0.5 as the default for the test. By default BINOMIALC computes
the proportion of observations for the rst variable level that appears in the output.
To specify a different level, use LEVEL=level-number or LEVEL=level-value, where
level-number is the variable levels number or order in the output, and level-value is
the formatted value of the variable level.
See the section Binomial Proportion on page 118 for more information.
CELLCHI2

displays each crosstabulation table cells contribution to the total Pearson chi-square
statistic, which is computed as
(frequency expected)2
expected
The CELLCHI2 option has no effect for one-way tables or for tables that are displayed with the LIST option.
CHISQ

requests chi-square tests of homogeneity or independence and measures of association based on chi-square. The tests include the Pearson chi-square, likelihood-ratio
chi-square, and Mantel-Haenszel chi-square. The measures include the phi coefcient, the contingency coefcient, and Cramers V . For 2 2 tables, the CHISQ
option includes Fishers exact test and the continuity-adjusted chi-square. For oneway tables, the CHISQ option requests a chi-square goodness-of-t test for equal
proportions. If you specify the null hypothesis proportions with the TESTP= option,
then PROC FREQ computes a chi-square goodness-of-t test for the specied proportions. If you specify null hypothesis frequencies with the TESTF= option, PROC

TABLES Statement
FREQ computes a chi-square goodness-of-t test for the specied frequencies. See
the section Chi-Square Tests and Statistics on page 103 for more information.
CL

requests condence limits for the MEASURES statistics. If you omit the
MEASURES option, the CL option invokes MEASURES. The FREQ procedure
determines the condence coefcient using the ALPHA= option, which, by default,
equals 0.05 and produces 95% condence limits.
For more information, see the section Condence Limits on page 109.
CMH

requests Cochran-Mantel-Haenszel statistics, which test for association between the

row and column variables after adjusting for the remaining variables in a multiway
table. In addition, for 2 2 tables, PROC FREQ computes the adjusted MantelHaenszel and logit estimates of the odds ratios and relative risks and the corresponding condence limits. For the stratied 2 2 case, PROC FREQ computes the
Breslow-Day test for homogeneity of odds ratios. (To request Tarones adjustment
for the Breslow-Day test, use the BDT option.) The CMH1 and CMH2 options control the number of CMH statistics that PROC FREQ computes. For more information,
see the section Cochran-Mantel-Haenszel Statistics on page 134.
CMH1

requests the Cochran-Mantel-Haenszel correlation statistic and, for 2 2 tables, the

adjusted Mantel-Haenszel and logit estimates of the odds ratios and relative risks
and the corresponding condence limits. For the stratied 2 2 case, PROC FREQ
computes the Breslow-Day test for homogeneity of odds ratios. Except for 2 2
tables, the CMH1 option requires less memory than the CMH option, which can
require an enormous amount for large tables.
CMH2

requests the Cochran-Mantel-Haenszel correlation statistic, row mean scores

(ANOVA) statistic, and, for 2 2 tables, the adjusted Mantel-Haenszel and logit
estimates of the odds ratios and relative risks and the corresponding condence
limits. For the stratied 2 2 case, PROC FREQ computes the Breslow-Day test for
homogeneity of odds ratios. Except for tables with two columns, the CMH2 option
requires less memory than the CMH option, which can require an enormous amount
for large tables.
CONTENTS=link-text

species the text for the HTML contents le links to crosstabulation tables. For
information on HTML output, refer to the SAS Output Delivery System Users Guide.
The CONTENTS= option affects only the HTML contents le, and not the HTML
body le.
If you omit the CONTENTS= option, by default, the HTML link text for crosstabulation tables is Cross-Tabular Freq Table.
Note that links to all crosstabulation tables produced by a single TABLES statement
use the same text. To specify different text for different crosstabulation table links,

Chapter 2. The FREQ Procedure

request the tables in separate TABLES statements and use the CONTENTS= option
in each TABLES statement.
The CONTENTS= option affects only links to crosstabulation tables. It does not
affect links to other PROC FREQ tables. To specify link text for any other PROC
FREQ table, you can use PROC TEMPLATE to create a customized table denition. The CONTENTS LABEL attribute in the DEFINE TABLE statement of PROC
TEMPLATE species the contents le link for the table. For detailed information,
refer to the chapter titled The TEMPLATE Procedure in the SAS Output Delivery
System Users Guide.
CONVERGE=value

species the convergence criterion for computing the polychoric correlation when
you specify the PLCORR option. The value of the CONVERGE= option must be a
positive number; by default, CONVERGE=0.0001. Iterative computation of the polychoric correlation stops when the convergence measure falls below the value of the
CONVERGE= option or when the number of iterations exceeds the value specied
in the MAXITER= option, whichever happens rst.
See the section Polychoric Correlation on page 116 for more information.
CROSSLIST

displays crosstabulation tables in ODS column format, instead of the default crosstabulation cell format. In a CROSSLIST table display, the rows correspond to the
crosstabulation table cells, and the columns correspond to descriptive statistics such
as Frequency, Percent, and so on. See the section Multiway Tables on page 152 for
details on the contents of the CROSSLIST table.
The CROSSLIST table displays the same information as the default crosstabulation
table, but uses an ODS column format instead of the table cell format. Unlike the
default crosstabulation table, the CROSSLIST table has a table denition that you
can customize with PROC TEMPLATE. For more information, refer to the chapter
titled The TEMPLATE Procedure in the SAS Output Delivery System Users Guide.
You can control the contents of a CROSSLIST table with the same options available
for the default crosstabulation table. These include the NOFREQ, NOPERCENT,
NOROW, and NOCOL options. You can request additional information in a
CROSSLIST table with the CELLCHI2, DEVIATION, EXPECTED, MISSPRINT,
and TOTPCT options.
The FORMAT= option and the CUMCOL option have no effect for CROSSLIST
tables. You cannot specify both the LIST option and the CROSSLIST option in the
same TABLES statement.
You can use the NOSPARSE option to suppress display of variable levels with zero
frequency in CROSSLIST tables. By default for CROSSLIST tables, PROC FREQ
displays all levels of the column variable within each level of the row variable, including any column variable levels with zero frequency for that row. And for multiway
tables displayed with the CROSSLIST option, the procedure displays all levels of
the row variable for each stratum of the table by default, including any row variable
levels with zero frequency for the stratum.

TABLES Statement
CUMCOL

displays the cumulative column percentages in the cells of the crosstabulation table.
DEVIATION

displays the deviation of the cell frequency from the expected frequency for each cell
of the crosstabulation table. The DEVIATION option is valid for contingency tables
but has no effect on tables produced with the LIST option.
EXPECTED

displays the expected table cell frequencies under the hypothesis of independence (or
homogeneity). The EXPECTED option is valid for crosstabulation tables but has no
effect on tables produced with the LIST option.
FISHER | EXACT

requests Fishers exact test for tables that are larger than 2 2. This test is also
known as the Freeman-Halton test. For more information, see the section Fishers
Exact Test on page 106 and the EXACT Statement section on page 77.
If you omit the CHISQ option in the TABLES statement, the FISHER option invokes
CHISQ. You can also request Fishers exact test by specifying the FISHER option in
the EXACT statement.
CAUTION: For tables with many rows or columns or with large total frequency,
PROC FREQ may require a large amount of time or memory to compute exact pvalues. See the section Computational Resources on page 145 for more information.
FORMAT=format-name

species a format for the following crosstabulation table cell values: frequency, expected frequency, and deviation. PROC FREQ also uses this format to display the
total row and column frequencies for crosstabulation tables.
You can specify any standard SAS numeric format or a numeric format dened
with the FORMAT procedure. The format length must not exceed 24. If you omit
FORMAT=, by default, PROC FREQ uses the BEST6. format to display frequencies
less than 1E6, and the BEST7. format otherwise.
To change formats for all other FREQ tables, you can use PROC TEMPLATE. For information on this procedure, refer to the chapter titled The TEMPLATE Procedure
in the SAS Output Delivery System Users Guide.
JT

performs the Jonckheere-Terpstra test.

Jonckheere-Terpstra Test on page 125.

For more information, see the section

LIST

displays two-way to n-way tables in a list format rather than as crosstabulation tables.
PROC FREQ ignores the LIST option when you request statistical tests or measures
of association.

Chapter 2. The FREQ Procedure

MAXITER=number

species the maximum number of iterations for computing the polychoric correlation
when you specify the PLCORR option. The value of the MAXITER= option must be
a positive integer; by default, MAXITER=20. Iterative computation of the polychoric
correlation stops when the number of iterations exceeds the value of the MAXITER=
option, or when the convergence measure falls below the value of the CONVERGE=
option, whichever happens rst. For more information see the section Polychoric
Correlation on page 116.
MEASURES

requests several measures of association and their asymptotic standard errors (ASE).
The measures include gamma, Kendalls tau-b, Stuarts tau-c, Somers D(C|R),
Somers D(R|C), the Pearson and Spearman correlation coefcients, lambda (symmetric and asymmetric), uncertainty coefcients (symmetric and asymmetric). To
request condence limits for these measures of association, you can specify the CL
option.
For 2 2 tables, the MEASURES option also provides the odds ratio, column
1 relative risk, column 2 relative risk, and the corresponding condence limits.
Alternatively, you can obtain the odds ratio and relative risks, without the other measures of association, by specifying the RELRISK option.
For more information, see the section Measures of Association on page 108.
MISSING

treats missing values as nonmissing and includes them in calculations of percentages

and other statistics.
For more information, see the section Missing Values on page 100.
MISSPRINT

displays missing value frequencies for all tables, even though PROC FREQ does not
use the frequencies in the calculation of statistics. For more information, see the
section Missing Values on page 100.
NOCOL

suppresses the display of column percentages in cells of the crosstabulation table.

NOCUM

suppresses the display of cumulative frequencies and cumulative percentages for oneway frequency tables and for crosstabulation tables in list format.
NOFREQ

suppresses the display of cell frequencies for crosstabulation tables. This also suppresses frequencies for row totals.
NOPERCENT

suppresses the display of cell percentages, row total percentages, and column total
percentages for crosstabulation tables. For one-way frequency tables and crosstabulation tables in list format, the NOPERCENT option suppresses the display of percentages and cumulative percentages.

TABLES Statement
NOPRINT

suppresses the display of frequency and crosstabulation tables but displays all requested tests and statistics. Use the NOPRINT option in the PROC FREQ statement
to suppress the display of all tables.
NOROW

suppresses the display of row percentages in cells of the crosstabulation table.

NOSPARSE

requests that PROC FREQ not invoke the SPARSE option when you specify the
ZEROS option in the WEIGHT statement. The NOSPARSE option suppresses the
display of cells with a zero frequency count in the list output, and it also omits them
from the OUT= data set. By default, the ZEROS option invokes the SPARSE option, which displays table cells with a zero frequency count in the LIST output and
includes them in the OUT= data set. For more information, see the description of the
ZEROS option.
For CROSSLIST tables, the NOSPARSE option suppresses display of variable levels
with zero frequency. By default for CROSSLIST tables, PROC FREQ displays all
levels of the column variable within each level of the row variable, including any
column variable levels with zero frequency for that row. And for multiway tables
displayed with the CROSSLIST option, the procedure displays all levels of the row
variable for each stratum of the table by default, including any row variable levels
with zero frequency for the stratum.
NOWARN

suppresses the log warning message that the asymptotic chi-square test may not be
valid. By default, PROC FREQ displays this log message when more than 20 percent
of the table cells have expected frequencies less than ve.
OUT=SAS-data-set

names the output data set that contains variable values and frequency counts. The
variable COUNT contains the frequencies and the variable PERCENT contains the
percentages. If more than one table request appears in the TABLES statement, the
contents of the data set correspond to the last table request in the TABLES statement.
For more information, see the section Output Data Sets on page 148 and see the
following descriptions for the options OUTCUM, OUTEXPECT, and OUTPCT.
OUTCUM

includes the cumulative frequency and the cumulative percentage for one-way tables
in the output data set when you specify the OUT= option in the TABLES statement.
The variable CUM FREQ contains the cumulative frequency for each level of the
analysis variable, and the variable CUM PCT contains the cumulative percentage
for each level. The OUTCUM option has no effect for two-way or multiway tables.
For more information, see the section Output Data Sets on page 148.
OUTEXPECT

includes the expected frequency in the output data set for crosstabulation tables when
you specify the OUT= option in the TABLES statement. The variable EXPECTED

Chapter 2. The FREQ Procedure

contains the expected frequency for each table cell. The EXPECTED option is valid
for two-way or multiway tables, and has no effect for one-way tables.
For more information, see the section Output Data Sets on page 148.
OUTPCT

includes the following additional variables in the output data set when you specify
the OUT= option in the TABLES statement for crosstabulation tables:

PCT COL

the percentage of column frequency

PCT ROW

the percentage of row frequency

PCT TABL

the percentage of stratum frequency, for n-way tables where n > 2

The OUTPCT option is valid for two-way or multiway tables, and has no effect for
one-way tables.
For more information, see the section Output Data Sets on page 148.
PLCORR

requests the polychoric correlation coefcient. For 2 2 tables, this statistic is more
commonly known as the tetrachoric correlation coefcient, and it is labeled as such
in the displayed output. If you omit the MEASURES option, the PLCORR option invokes MEASURES. For more information, see the section Polychoric Correlation
on page 116 and the descriptions for the CONVERGE= and MAXITER= options in
this list.
PRINTKWT

displays the weights PROC FREQ uses to compute the weighted kappa coefcient.
You must also specify the AGREE option, which requests the weighted kappa coefcient. You can specify (WT=FC) with the AGREE option to request Fleiss-Cohen
weights. By default, PROC FREQ uses Cicchetti-Allison weights.
See the section Weighted Kappa Coefcient on page 130 for more information.
RELRISK

requests relative risk measures and their condence limits for 2 2 tables. These
measures include the odds ratio and the column 1 and 2 relative risks. For more
information, see the section Odds Ratio and Relative Risks for 2 x 2 Tables on page
122. You can also obtain the RELRISK measures by specifying the MEASURES
option, which produces other measures of association in addition to the relative risks.
RISKDIFF

requests column 1 and 2 risks (or binomial proportions), risk differences, and their
condence limits for 2 2 tables. See the section Risks and Risk Differences on
page 120 for more information.
RISKDIFFC

requests the RISKDIFF option statistics for 2 2 tables, and includes a continuity
correction in the asymptotic condence limits. The RISKDIFF option statistics include the column 1 and 2 risks (or binomial proportions), risk differences, and their

TABLES Statement
condence limits. See the section Risks and Risk Differences on page 120 for more
information.
SCORES=type

species the type of row and column scores that PROC FREQ uses with the MantelHaenszel chi-square, Pearson correlation, Cochran-Armitage test for trend, weighted
kappa coefcient, and Cochran-Mantel-Haenszel statistics, where type is one of the
following (the default is SCORE=TABLE):
MODRIDIT
RANK
RIDIT
TABLE
By default, the row or column scores are the integers 1,2,... for character variables
and the actual variable values for numeric variables. Using other types of scores
yields nonparametric analyses. For more information, see the section Scores on
page 102.
To display the row and column scores, you can use the SCOROUT option.
SCOROUT

displays the row and the column scores. You specify the score type with the
SCORES= option. PROC FREQ uses the scores when it calculates the MantelHaenszel chi-square, Pearson correlation, Cochran-Armitage test for trend, weighted
kappa coefcient, or Cochran-Mantel-Haenszel statistics. The SCOROUT option
displays the row and column scores only when statistics are computed for two-way
tables. To store the scores in an output data set, use the Output Delivery System.
For more information, see the section Scores on page 102.
SPARSE

lists all possible combinations of the variable values for an n-way table when n > 1,
even if a combination does not occur in the data. The SPARSE option applies only
to crosstabulation tables displayed in list format and to the OUT= output data set.
Otherwise, if you do not use the LIST option or the OUT= option, the SPARSE
option has no effect.
When you specify the SPARSE and LIST options, PROC FREQ displays all combinations of variable variables in the table listing, including those values with a frequency
count of zero. By default, without the SPARSE option, PROC FREQ does not display
zero-frequency values in list output. When you use the SPARSE and OUT= options,
PROC FREQ includes empty crosstabulation table cells in the output data set. By
default, PROC FREQ does not include zero-frequency table cells in the output data
set.
For more information, see the section Missing Values on page 100.

Chapter 2. The FREQ Procedure

TESTF=(values)

species the null hypothesis frequencies for a one-way chi-square test for specied
frequencies. You can separate values with blanks or commas. The sum of the frequency values must equal the total frequency for the one-way table. The number of
TESTF= values must equal the number of variable levels in the one-way table. List
these values in the order in which the corresponding variable levels appear in the
output. If you omit the CHISQ option, the TESTF= option invokes CHISQ.
For more information, see the section Chi-Square Test for One-Way Tables on page
104.
TESTP=(values)

species the null hypothesis proportions for a one-way chi-square test for specied
proportions. You can separate values with blanks or commas. Specify values in probability form as numbers between 0 and 1, where the proportions sum to 1. Or specify
values in percentage form as numbers between 0 and 100, where the percentages sum
to 100. The number of TESTP= values must equal the number of variable levels in
the one-way table. List these values in the order in which the corresponding variable levels appear in the output. If you omit the CHISQ option, the TESTP= option
invokes CHISQ.
For more information, see the section Chi-Square Test for One-Way Tables on page
104.
TOTPCT

displays the percentage of total frequency in crosstabulation tables, for n-way tables where n > 2. This percentage is also available with the LIST option or as the
PERCENT variable in the OUT= output data set.
TREND

performs the Cochran-Armitage test for trend. The table must be 2 C or R 2. For
more information, see the section Cochran-Armitage Test for Trend on page 124.

TEST Statement
TEST options ;
The TEST statement requests asymptotic tests for the specied measures of association and measures of agreement. You must use a TABLES statement with the TEST
statement.
options

specify the statistics for which to provide asymptotic tests. The available statistics are
those measures of association and agreement listed in Table 2.10 . The option names
are identical to those in the TABLES statement and the OUTPUT statement. You can
request all available tests for groups of statistics by using group options MEASURES
or AGREE. Or you can request tests individually by using one of the options shown
in Table 2.10 .
For each measure of association or agreement that you specify, the TEST statement provides an asymptotic test that the measure equals zero. When you request
an asymptotic test, PROC FREQ gives the asymptotic standard error under the null

WEIGHT Statement
hypothesis, the test statistic, and the p-values. Additionally, PROC FREQ reports the
condence limits for that measure. The ALPHA= option in the TABLES statement
determines the condence level, which, by default, equals 0.05 and provides 95%
condence limits. For more information, see the sections Asymptotic Tests on
page 109 and Condence Limits on page 109, and see Statistical Computations
beginning on page 102 for sections describing the individual measures.
In addition to these asymptotic tests, exact tests for selected measures of association
and agreement are available with the EXACT statement. See the section EXACT
Statement on page 77 for more information.
Table 2.10. TEST Statement Options and Required TABLES Statement Options

Option

Asymptotic Tests Computed

AGREE

simple kappa coefcient and weighted

kappa coefcient
gamma
simple kappa coefcient
Kendalls tau-b
gamma, Kendalls tau-b, Stuarts tau-c,
Somers D(C|R), Somers D(R|C), the
Pearson correlation, and the Spearman
correlation
Pearson correlation coefcient
Spearman correlation coefcient
Somers D(C|R)
Somers D(R|C)
Stuarts tau-c
weighted kappa coefcient

GAMMA
KAPPA
KENTB
MEASURES

PCORR
SCORR
SMDCR
SMDRC
STUTC
WTKAP

Required TABLES
Statement Option
AGREE
ALL or MEASURES
AGREE
ALL or MEASURES
ALL or MEASURES

ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
ALL or MEASURES
AGREE

WEIGHT Statement
WEIGHT variable < / option > ;
The WEIGHT statement species a numeric variable with a value that represents the
frequency of the observation. The WEIGHT statement is most commonly used to
input cell count data. See the Inputting Frequency Counts section on page 98 for
more information. If you use the WEIGHT statement, PROC FREQ assumes that an
observation represents n observations, where n is the value of variable. The value of
the weight variable need not be an integer. When a weight value is missing, PROC
FREQ ignores the corresponding observation. When a weight value is zero, PROC
FREQ ignores the corresponding observation unless you specify the ZEROS option,
which includes observations with zero weights. If a WEIGHT statement does not
appear, each observation has a default weight of 1. The sum of the weight variable
values represents the total number of observations.
If any value of the weight variable is negative, PROC FREQ displays the frequencies
(as measured by the weighted values) but does not compute percentages and other
statistics. If you create an output data set using the OUT= option in the TABLES

Chapter 2. The FREQ Procedure

statement, PROC FREQ creates the PERCENT variable and assigns a missing value
for each observation. PROC FREQ also assigns missing values to the variables that
the OUTEXPECT and OUTPCT options create. You cannot create an output data
set using the OUTPUT statement since statistics are not computed when there are
negative weights.

Option
ZEROS

includes observations with zero weight values. By default, PROC FREQ ignores
observations with zero weights.
If you specify the ZEROS option, frequency and and crosstabulation tables display
any levels corresponding to observations with zero weights. Without the ZEROS
option, PROC FREQ does not process observations with zero weights, and so does
not display levels that contain only observations with zero weights.
With the ZEROS option, PROC FREQ includes levels with zero weights in the chisquare goodness-of-t test for one-way tables. Also, PROC FREQ includes any levels
with zero weights in binomial computations for one-way tables. This enables computation of binomial estimates and tests when there are no observations with positive
weights in the specied level.
For two-way tables, the ZEROS option enables computation of kappa statistics when
there are levels containing no observations with positive weight. For more information, see the section Tables with Zero Rows and Columns on page 133.
Note that even with the ZEROS option, PROC FREQ does not compute the CHISQ
or MEASURES statistics for two-way tables when the table has a zero row or zero
column, because most of these statistics are undened in this case.
The ZEROS option invokes the SPARSE option in the TABLES statement, which
includes table cells with a zero frequency count in the list output and the OUT=
data set. By default, without the SPARSE option, PROC FREQ does not include
zero frequency cells in the list output or in the OUT= data set. If you specify the
ZEROS option in the WEIGHT statement but do not want the SPARSE option, you
can specify the NOSPARSE option in the TABLES statement.

Details
Inputting Frequency Counts
PROC FREQ can use either raw data or cell count data to produce frequency and
crosstabulation tables. Raw data, also known as case-record data, report the data
as one record for each subject or sample member. Cell count data report the data
as a table, listing all possible combinations of data values along with the frequency
counts. This way of presenting data often appears in published results.

Grouping with Formats

The following DATA step statements store raw data in a SAS data set:
data Raw;
input Subject $
datalines;
01 1 1 02 1 1 03
06 1 2 07 1 2 08
11 2 1 12 2 1 13
;

R C @@;
1 1
1 2
2 2

04 1 1
09 2 1
14 2 2

05 1 1
10 2 1
14 2 2

You can store the same data as cell counts using the following DATA step statements:
data CellCounts;
input R C Count @@;
datalines;
1 1 5
1 2 3
2 1 4
2 2 3
;

The variable R contains the values for the rows, and the variable C contains the values
for the columns. The Count variable contains the cell count for each row and column
combination.
Both the Raw data set and the CellCounts data set produce identical frequency
counts, two-way tables, and statistics. With the CellCounts data set, you must use
a WEIGHT statement to specify that the Count variable contains cell counts. For
example, to create a two-way crosstabulation table, submit the following statements:
proc freq data=CellCounts;
weight Count;
tables R*C;
run;

Grouping with Formats

PROC FREQ groups a variables values according to its formatted values. If you
assign a format to a variable with a FORMAT statement, PROC FREQ formats the
variable values before dividing observations into the levels of a frequency or crosstabulation table.
For example, suppose that a variable X has the values 1.1, 1.4, 1.7, 2.1, and 2.3. Each
of these values appears as a level in the frequency table. If you decide to round each
value to a single digit, include the following statement in the PROC FREQ step:
format X 1.;

Now the table lists the frequency count for formatted level 1 as two and formatted
level 2 as three.

100

Chapter 2. The FREQ Procedure

PROC FREQ treats formatted character variables in the same way. The formatted values are used to group the observations into the levels of a frequency table or crosstabulation table. PROC FREQ uses the entire value of a character format to classify an
observation.
You can also use the FORMAT statement to assign formats that were created with
the FORMAT procedure to the variables. User-written formats determine the number
of levels for a variable and provide labels for a table. If you use the same data with
different formats, then you can produce frequency counts and statistics for different
classications of the variable values.
When you use PROC FORMAT to create a user-written format that combines missing
and nonmissing values into one category, PROC FREQ treats the entire category of
formatted values as missing. For example, a questionnaire codes 1 as yes, 2 as no,
and 8 as a no answer. The following PROC FORMAT step creates a user-written
format:
proc format;
value Questfmt 1 =Yes
2 =No
8,.=Missing;
run;

When you use a FORMAT statement to assign Questfmt. to a variable, the variables
frequency table no longer includes a frequency count for the response of 8. You
must use the MISSING or MISSPRINT option in the TABLES statement to list the
frequency for no answer. The frequency count for this level includes observations
with either a value of 8 or a missing value (.).
The frequency or crosstabulation table lists the values of both character and numeric
variables in ascending order based on internal (unformatted) variable values unless
you change the order with the ORDER= option. To list the values in ascending order
by formatted values, use ORDER=FORMATTED in the PROC FREQ statement.
For more information on the FORMAT statement, refer to SAS Language Reference:
Concepts.

Missing Values
By default, PROC FREQ excludes missing values before it constructs the frequency
and crosstabulation tables. PROC FREQ also excludes missing values before computing statistics. However, the total frequency of observations with missing values is
displayed below each table. The following options change the way in which PROC
FREQ handles missing values:

MISSPRINT includes missing value frequencies in frequency or crosstabulation tables.

MISSING

includes missing values in percentage and statistical calculations.

Missing Values
The OUT= option in the TABLES statement includes an observation in the output
data set that contains the frequency of missing values. The NMISS option in the
OUTPUT statement creates a variable in the output data set that contains the number
of missing values.
Figure 2.7 shows three ways in which PROC FREQ handles missing values. The rst
table uses the default method; the second table uses the MISSPRINT option; and the
third table uses the MISSING option.
*** Default ***
The FREQ Procedure
Cumulative
Cumulative
A
Frequency
Percent
Frequency
Percent
-----------------------------------------------------1
2
50.00
2
50.00
2
2
50.00
4
100.00
Frequency Missing = 2

* MISSPRINT Option *

The FREQ Procedure
Cumulative
Cumulative
A
Frequency
Percent
Frequency
Percent
-----------------------------------------------------.
2
.
.
.
1
2
50.00
2
50.00
2
2
50.00
4
100.00
Frequency Missing = 2

* MISSING Option *

The FREQ Procedure
Cumulative
Cumulative
A
Frequency
Percent
Frequency
Percent
-----------------------------------------------------.
2
33.33
2
33.33
1
2
33.33
4
66.67
2
2
33.33
6
100.00

Figure 2.7. Missing Values in Frequency Tables

When a combination of variable values for a crosstabulation is missing, PROC FREQ

assigns zero to the frequency count for the table cell. By default, PROC FREQ omits
missing combinations in list format and in the output data set that is created in a
TABLES statement. To include the missing combinations, use the SPARSE option
with the LIST or OUT= option in the TABLES statement.
PROC FREQ treats missing BY variable values like any other BY variable value. The
missing values form a separate BY group. When the value of a WEIGHT variable is
missing, PROC FREQ excludes the observation from the analysis.

101

102

Chapter 2. The FREQ Procedure

Statistical Computations
Denitions and Notation
In this chapter, a two-way table represents the crosstabulation of variables X and Y.
Let the rows of the table be labeled by the values Xi , i = 1, 2, . . . , R, and the columns
by Yj , j = 1, 2, . . . , C. Let nij denote the cell frequency in the ith row and the jth
column and dene the following:
nij

(row totals)

nij

ni =

(column totals)

=
i

nij

n =
i

pij

(overall total)

= nij /n

(cell percentages)

pi = ni /n

(row percentages)

(column percentages)

= nj /n

Ri = score for row i

= score for column j

R =

ni Ri /n

(average row score)

nj Cj /n

(average column score)

C =
j

Aij

k<i l<j

k>i l>j

Dij

nkl +
k>i l<j

nkl

nkl +

nkl
k<i l>j

nij Aij
i

(twice the number of discordances)

(twice the number of concordances)

nij Dij

Q =

Scores

PROC FREQ uses scores for the variable values when computing the MantelHaenszel chi-square, Pearson correlation, Cochran-Armitage test for trend, weighted
kappa coefcient, and Cochran-Mantel-Haenszel statistics. The SCORES= option in
the TABLES statement species the score type that PROC FREQ uses. The available
score types are TABLE, RANK, RIDIT, and MODRIDIT scores. The default score
type is TABLE.

Statistical Computations
For numeric variables, table scores are the values of the row and column levels. If the
row or column variables are formatted, then the table score is the internal numeric
value corresponding to that level. If two or more numeric values are classied into
the same formatted level, then the internal numeric value for that level is the smallest
of these values. For character variables, table scores are dened as the row numbers
and column numbers (that is, 1 for the rst row, 2 for the second row, and so on).
Rank scores, which you can use to obtain nonparametric analyses, are dened by
Row scores:

R1i =

nk + (ni + 1)/2

i = 1, 2, . . . , R

nl + (nj + 1)/2

j = 1, 2, . . . , C

k<i

Column scores:

C1j

=
l<j

Note that rank scores yield midranks for tied values.

Ridit scores (Bross 1958; Mack and Skillings 1980) also yield nonparametric analyses, but they are standardized by the sample size. Ridit scores are derived from rank
scores as
R2i = R1i /n
C2j

= C1j /n

Modied ridit (MODRIDIT) scores (van Elteren 1960; Lehmann 1975), which also
yield nonparametric analyses, represent the expected values of the order statistics for
the uniform distribution on (0,1). Modied ridit scores are derived from rank scores
as
R3i = R1i /(n + 1)
C3j

= C1j /(n + 1)

Chi-Square Tests and Statistics

When you specify the CHISQ option in the TABLES statement, PROC FREQ performs the following chi-square tests for each two-way table: Pearson chi-square,
continuity-adjusted chi-square for 2 2 tables, likelihood-ratio chi-square, MantelHaenszel chi-square, and Fishers exact test for 2 2 tables. Also, PROC FREQ
computes the following statistics derived from the Pearson chi-square: the phi coefcient, the contingency coefcient, and Cramers V . PROC FREQ computes Fishers
exact test for general R C tables when you specify the FISHER (or EXACT) option
in the TABLES statement, or, equivalently, when you specify the FISHER option in
the EXACT statement.

103

104

Chapter 2. The FREQ Procedure

For one-way frequency tables, PROC FREQ performs a chi-square goodness-of-t
test when you specify the CHISQ option. The other chi-square tests and statistics
described in this section are dened only for two-way tables and so are not computed
for one-way frequency tables.
All the two-way test statistics described in this section test the null hypothesis of
no association between the row variable and the column variable. When the sample
size n is large, these test statistics are distributed approximately as chi-square when
the null hypothesis is true. When the sample size is not large, exact tests may be
useful. PROC FREQ computes exact tests for the following chi-square statistics when
you specify the corresponding option in the EXACT statement: Pearson chi-square,
likelihood-ratio chi-square, and Mantel-Haenszel chi-square. See the section Exact
Statistics beginning on page 142 for more information.
Note that the Mantel-Haenszel chi-square statistic is appropriate only when both variables lie on an ordinal scale. The other chi-square tests and statistics in this section
are appropriate for either nominal or ordinal variables. The following sections give
the formulas that PROC FREQ uses to compute the chi-square tests and statistics. For
further information on the formulas and on the applicability of each statistic, refer to
Agresti (1996), Stokes, Davis, and Koch (1995), and the other references cited for
each statistic.

Chi-Square Test for One-Way Tables

For one-way frequency tables, the CHISQ option in the TABLES statement computes
a chi-square goodness-of-t test. Let C denote the number of classes, or levels, in the
one-way table. Let fi denote the frequency of class i (or the number of observations
in class i) for i = 1, 2, ..., C. Then PROC FREQ computes the chi-square statistic as
C

=
i=1

(fi ei )2
ei

where ei is the expected frequency for class i under the null hypothesis.
In the test for equal proportions, which is the default for the CHISQ option, the null
hypothesis species equal proportions of the total sample size for each class. Under
this null hypothesis, the expected frequency for each class equals the total sample
size divided by the number of classes,
ei = n / C

for i = 1, 2, . . . , C

In the test for specied frequencies, which PROC FREQ computes when you input null hypothesis frequencies using the TESTF= option, the expected frequencies
are those TESTF= values. In the test for specied proportions, which PROC FREQ
computes when you input null hypothesis proportions using the TESTP= option, the
expected frequencies are determined from the TESTP= proportions pi , as
ei = pi n

for i = 1, 2, . . . , C

Statistical Computations
Under the null hypothesis (of equal proportions, specied frequencies, or specied
proportions), this test statistic has an asymptotic chi-square distribution, with C 1
degrees of freedom. In addition to the asymptotic test, PROC FREQ computes the
exact one-way chi-square test when you specify the CHISQ option in the EXACT
statement.
Chi-Square Test for Two-Way Tables

The Pearson chi-square statistic for two-way tables involves the differences between
the observed and expected frequencies, where the expected frequencies are computed
under the null hypothesis of independence. The chi-square statistic is computed as
(nij eij )2
eij

QP =
i

where
eij =

ni nj
n

When the row and column variables are independent, QP has an asymptotic chisquare distribution with (R 1)(C 1) degrees of freedom. For large values of QP ,
this test rejects the null hypothesis in favor of the alternative hypothesis of general
association. In addition to the asymptotic test, PROC FREQ computes the exact chisquare test when you specify the PCHI or CHISQ option in the EXACT statement.
For a 2 2 table, the Pearson chi-square is also appropriate for testing the equality
of two binomial proportions or, for R 2 and 2 C tables, the homogeneity of
proportions. Refer to Fienberg (1980).
Likelihood-Ratio Chi-Square Test

The likelihood-ratio chi-square statistic involves the ratios between the observed and
expected frequencies. The statistic is computed as
G2 = 2

nij ln
i

nij
eij

When the row and column variables are independent, G2 has an asymptotic chisquare distribution with (R 1)(C 1) degrees of freedom. In addition to the
asymptotic test, PROC FREQ computes the exact test when you specify the LRCHI
or CHISQ option in the EXACT statement.
Continuity-Adjusted Chi-Square Test

The continuity-adjusted chi-square statistic for 2 2 tables is similar to the Pearson

chi-square, except that it is adjusted for the continuity of the chi-square distribution.
The continuity-adjusted chi-square is most useful for small sample sizes. The use of
the continuity adjustment is controversial; this chi-square test is more conservative,
and more like Fishers exact test, when your sample size is small. As the sample size
increases, the statistic becomes more and more like the Pearson chi-square.

105

106

Chapter 2. The FREQ Procedure

The statistic is computed as
[ max(0, |nij eij | 0.5) ]2
eij

QC =
i

Under the null hypothesis of independence, QC has an asymptotic chi-square distribution with (R 1)(C 1) degrees of freedom.
Mantel-Haenszel Chi-Square Test

The Mantel-Haenszel chi-square statistic tests the alternative hypothesis that there is
a linear association between the row variable and the column variable. Both variables
must lie on an ordinal scale. The statistic is computed as
QM H = (n 1)r2
where r2 is the Pearson correlation between the row variable and the column variable. For a description of the Pearson correlation, see the Pearson Correlation
Coefcient section on page 113. The Pearson correlation and, thus, the MantelHaenszel chi-square statistic use the scores that you specify in the SCORES= option
in the TABLES statement.
Under the null hypothesis of no association, QM H has an asymptotic chi-square
distribution with one degree of freedom. In addition to the asymptotic test, PROC
FREQ computes the exact test when you specify the MHCHI or CHISQ option in the
EXACT statement.
Refer to Mantel and Haenszel (1959) and Landis, Heyman, and Koch (1978).
Fishers Exact Test

Fishers exact test is another test of association between the row and column variables. This test assumes that the row and column totals are xed, and then uses the
hypergeometric distribution to compute probabilities of possible tables with these
observed row and column totals. Fishers exact test does not depend on any largesample distribution assumptions, and so it is appropriate even for small sample sizes
and for sparse tables.
2 2 Tables

For 2 2 tables, PROC FREQ gives the following information for Fishers exact test:
table probability, two-sided p-value, left-sided p-value, and right-sided p-value. The
table probability equals the hypergeometric probability of the observed table, and is
in fact the value of the test statistic for Fishers exact test.
Where p is the hypergeometric probability of a specic table with the observed row
and column totals, Fishers exact p-values are computed by summing probabilities p
over dened sets of tables,
PROB =

p
A

Statistical Computations
The two-sided p-value is the sum of all possible table probabilties (for tables having
the observed row and column totals) that are less than or equal to the observed table
probability. So, for the two-sided p-value, the set A includes all possible tables with
hypergeometric probabilities less than or equal to the probability of the observed
table. A small two-sided p-value supports the alternative hypothesis of association
between the row and column variables.
One-sided tests are dened in terms of the frequency of the cell in the rst row and
rst column of the table, the (1,1) cell. Denoting the observed (1,1) cell frequency
by F , the left-sided p-value for Fishers exact test is probability that the (1,1) cell
frequency is less than or equal to F . So, for the left-sided p-value, the set A includes
those tables with a (1,1) cell frequency less than or equal to F . A small left-sided pvalue supports the alternative hypothesis that the probability of an observation being
in the rst cell is less than expected under the null hypothesis of independent row and
column variables.
Similarly, for a right-sided alternative hypothesis, A is the set of tables where the
frequency of the (1,1) cell is greater than or equal to that in the observed table. A
small right-sided p-value supports the alternative that the probability of the rst cell
is greater than that expected under the null hypothesis.
Because the (1,1) cell frequency completely determines the 2 2 table when the
marginal row and column sums are xed, these one-sided alternatives can be equivalently stated in terms of other cell probabilities or ratios of cell probabilities. The
left-sided alternative is equivalent to an odds ratio greater than 1, where the odds ratio equals (n11 n22 / n12 n21 ). Additionally, the left-sided alternative is equivalent to
the column 1 risk for row 1 being less than the column 1 risk for row 2, p1|1 < p1|2 .
Similarly, the right-sided alternative is equivalent to the column 1 risk for row 1 being
greater than the column 1 risk for row 2, p1|1 > p1|2 . Refer to Agresti (1996).
R C Tables

Fishers exact test was extended to general R C tables by Freeman and Halton
(1951), and this test is also known as the Freeman-Halton test. For R C tables, the
two-sided p-value is dened the same as it is for 2 2 tables. The set A contains all
tables with p less than or equal to the probability of the observed table. A small pvalue supports the alternative hypothesis of association between the row and column
variables. For RC tables, Fishers exact test is inherently two-sided. The alternative
hypothesis is dened only in terms of general, and not linear, association. Therefore,
PROC FREQ does not provide right-sided or left-sided p-values for general R C
tables.
For R C tables, PROC FREQ computes Fishers exact test using the network algorithm of Mehta and Patel (1983), which provides a faster and more efcient solution
than direct enumeration. See the section Exact Statistics beginning on page 142 for
more details.

107

108

Chapter 2. The FREQ Procedure

Phi Coefcient

The phi coefcient is a measure of association derived from the Pearson chi-square
statistic. It has range 1 1 for 2 2 tables. Otherwise, the range is
the

0 min( R 1, C 1) (Liebetrau 1983). The phi coefcient is computed

as
n11 n22 n12 n21
=
n1 n2 n1 n2
=

QP /n

for 2 2 tables

otherwise

Refer to Fleiss (1981, pp. 5960).

Contingency Coefcient

The contingency coefcient is a measure of association derived from the Pearson chisquare. It has the range 0 P (m 1)/m, where m = min(R, C) (Liebetrau
1983). The contingency coefcient is computed as

P =

QP
QP + n

Refer to Kendall and Stuart (1979, pp. 587588).

Cramers V

Cramers V is a measure of association derived from the Pearson chi-square. It is

designed so that the attainable upper bound is always 1. It has the range 1 V 1
for 2 2 tables; otherwise, the range is 0 V 1. Cramers V is computed as
V =

V =

for 2 2 tables
QP /n
min(R 1, C 1)

otherwise

Refer to Kendall and Stuart (1979, p. 588).

Measures of Association
When you specify the MEASURES option in the TABLES statement, PROC FREQ
computes several statistics that describe the association between the two variables
of the contingency table. The following are measures of ordinal association that
consider whether the variable Y tends to increase as X increases: gamma, Kendalls
tau-b, Stuarts tau-c, and Somers D. These measures are appropriate for ordinal
variables, and they classify pairs of observations as concordant or discordant. A pair
is concordant if the observation with the larger value of X also has the larger value of
Y. A pair is discordant if the observation with the larger value of X has the smaller

Statistical Computations
value of Y. Refer to Agresti (1996) and the other references cited in the discussion of
each measure of association.
The Pearson correlation coefcient and the Spearman rank correlation coefcient are
also appropriate for ordinal variables. The Pearson correlation describes the strength
of the linear association between the row and column variables, and it is computed
using the row and column scores specied by the SCORES= option in the TABLES
statement. The Spearman correlation is computed with rank scores. The polychoric
correlation (requested by the PLCORR option) also requires ordinal variables and
assumes that the variables have an underlying bivariate normal distribution. The following measures of association do not require ordinal variables, but they are appropriate for nominal variables: lambda asymmetric, lambda symmetric, and uncertainty
coefcients.
PROC FREQ computes estimates of the measures according to the formulas given in
the discussion of each measure of association. For each measure, PROC FREQ computes an asymptotic standard error (ASE), which is the square root of the asymptotic
variance denoted by var in the following sections.
Condence Limits

If you specify the CL option in the TABLES statement, PROC FREQ computes
asymptotic condence limits for all MEASURES statistics. The condence coefcient is determined according to the value of the ALPHA= option, which, by default,
equals 0.05 and produces 95% condence limits.
The condence limits are computed as
est ( z/2 ASE )
where est is the estimate of the measure, z/2 is the 100(1 /2) percentile of
the standard normal distribution, and ASE is the asymptotic standard error of the
estimate.
Asymptotic Tests

For each measure that you specify in the TEST statement, PROC FREQ computes
an asymptotic test of the null hypothesis that the measure equals zero. Asymptotic
tests are available for the following measures of association: gamma, Kendalls tau-b,
Stuarts tau-c, Somers D(R|C), Somers D(C|R), the Pearson correlation coefcient, and the Spearman rank correlation coefcient. To compute an asymptotic test,
PROC FREQ uses a standardized test statistic z, which has an asymptotic standard
normal distribution under the null hypothesis. The standardized test statistic is computed as
z =

est
var0 (est)

where est is the estimate of the measure and var0 (est) is the variance of the estimate
under the null hypothesis. Formulas for var0 (est) are given in the discussion of each
measure of association.

109

110

Chapter 2. The FREQ Procedure

Note that the ratio of est to var0 (est) is the same for the following measures:
gamma, Kendalls tau-b, Stuarts tau-c, Somers D(R|C), and Somers D(C|R).
Therefore, the tests for these measures are identical. For example, the p-values for
the test of H0 : gamma = 0 equal the p-values for the test of H0 : tau-b = 0.
PROC FREQ computes one-sided and two-sided p-values for each of these tests.
When the test statistic z is greater than its null hypothesis expected value of zero,
PROC FREQ computes the right-sided p-value, which is the probability of a larger
value of the statistic occurring under the null hypothesis. A small right-sided p-value
supports the alternative hypothesis that the true value of the measure is greater than
zero. When the test statistic is less than or equal to zero, PROC FREQ computes the
left-sided p-value, which is the probability of a smaller value of the statistic occurring
under the null hypothesis. A small left-sided p-value supports the alternative hypothesis that the true value of the measure is less than zero. The one-sided p-value P1 can
be expressed as
P1 = Prob ( Z > z )

if z > 0

P1 = Prob ( Z < z )

if z 0

where Z has a standard normal distribution. The two-sided p-value P2 is computed

as
P2 = Prob ( |Z| > |z| )
Exact Tests

Exact tests are available for two measures of association, the Pearson correlation coefcient and the Spearman rank correlation coefcient. If you specify the PCORR
option in the EXACT statement, PROC FREQ computes the exact test of the hypothesis that the Pearson correlation equals zero. If you specify the SCORR option in the
EXACT statement, PROC FREQ computes the exact test of the hypothesis that the
Spearman correlation equals zero. See the section Exact Statistics beginning on
page 142 for information on exact tests.
Gamma

The estimator of gamma is based only on the number of concordant and discordant
pairs of observations. It ignores tied pairs (that is, pairs of observations that have
equal values of X or equal values of Y ). Gamma is appropriate only when both
variables lie on an ordinal scale. It has the range 1 1. If the two variables
are independent, then the estimator of gamma tends to be close to zero. Gamma is
estimated by
G=

P Q
P +Q

with asymptotic variance

var =

16
(P + Q)4

nij (QAij P Dij )2

Statistical Computations
The variance of the estimator under the null hypothesis that gamma equals zero is
computed as

var0 (G) =
(P + Q)2

nij (Aij Dij )2 (P Q)2 /n

For 2 2 tables, gamma is equivalent to Yules Q. Refer to Goodman and Kruskal

(1979), Agresti (1990), and Brown and Benedetti (1977).
Kendalls Tau-b

Kendalls tau-b is similar to gamma except that tau-b uses a correction for ties. Tau-b
is appropriate only when both variables lie on an ordinal scale. Tau-b has the range
1 b 1. It is estimated by
P Q
tb =
wr wc
with

1
var = 4
w

nij (2wdij + tb vij )2 n3 t2 (wr + wc )2

b
i

where
w =

wr wc
n2
i

wr = n2
i

n2
j

wc = n2
j

dij

= Aij Dij

vij

= ni wc + nj wr

The variance of the estimator under the null hypothesis that tau-b equals zero is computed as

var0 (tb ) =

4
wr wc

nij (Aij Dij )2 (P Q)2 /n

Refer to Kendall (1955) and Brown and Benedetti (1977).

111

112

Chapter 2. The FREQ Procedure

Stuarts Tau-c

Stuarts tau-c makes an adjustment for table size in addition to a correction for ties.
Tau-c is appropriate only when both variables lie on an ordinal scale. Tau-c has the
range 1 c 1. It is estimated by
tc =

m(P Q)
n2 (m 1)

with

var =

4m2
(m 1)2 n4

nij d2 (P Q)2 /n
ij
i

where
m = min(R, C)
dij

= Aij Dij

The variance of the estimator under the null hypothesis that tau-c equals zero is
var0 (tc ) = var
Refer to Brown and Benedetti (1977).
Somers D (C |R ) and D (R |C )

Somers D(C|R) and Somers D(R|C) are asymmetric modications of tau-b. C|R
denotes that the row variable X is regarded as an independent variable, while the
column variable Y is regarded as dependent. Similarly, R|C denotes that the column
variable Y is regarded as an independent variable, while the row variable X is regarded
as dependent. Somers D differs from tau-b in that it uses a correction only for pairs
that are tied on the independent variable. Somers D is appropriate only when both
variables lie on an ordinal scale. It has the range 1 D 1. Formulas for Somers
D(R|C) are obtained by interchanging the indices.
D(C|R) =

P Q
wr

with
var =

4
4
wr

nij (wr dij (P Q)(n ni ))2

Statistical Computations
where
n2
i

wr = n2
i

dij

= Aij Dij

The variance of the estimator under the null hypothesis that D(C|R) equals zero is
computed as

var0 (D(C|R)) =

4
2
wr

nij (Aij Dij )2 (P Q)2 /n

Refer to Somers (1962), Goodman and Kruskal (1979), and Liebetrau (1983).
Pearson Correlation Coefcient

PROC FREQ computes the Pearson correlation coefcient using the scores specied
in the SCORES= option. The Pearson correlation is appropriate only when both variables lie on an ordinal scale. It has the range 1 1. The Pearson correlation
coefcient is computed as
r=

v
ssrc
=
w
ssr ssc

with
var =

1
w4

bij v

nij w(Ri R)(Cj C)

The row scores Ri and the column scores Cj are determined by the SCORES= option
in the TABLES statement, and

nij (Ri R)2

ssr =
i

nij (Cj C)2

ssc =
i

nij (Ri R)(Cj C)

ssrc =
bij

= (Ri R)2 ssc + (Cj C)2 ssr

v = ssrc

w =
ssr ssc

113

114

Chapter 2. The FREQ Procedure

Refer to Snedecor and Cochran (1989) and Brown and Benedetti (1977).
To compute an asymptotic test for the Pearson correlation, PROC FREQ uses a standardized test statistic r , which has an asymptotic standard normal distribution under
the null hypothesis that the correlation equals zero. The standardized test statistic is
computed as
r

r =

var0 (r)

where var0 (r) is the variance of the correlation under the null hypothesis.
i

var0 (r) =

nij (Ri R)2 (Cj C)2 ss2 /n

ssr ssc

The asymptotic variance is derived for multinomial sampling in a contingency table

framework, and it differs from the form obtained under the assumption that both
variables are continuous and normally distributed. Refer to Brown and Benedetti
(1977).
PROC FREQ also computes the exact test for the hypothesis that the Pearson correlation equals zero when you specify the PCORR option in the EXACT statement. See
the section Exact Statistics beginning on page 142 for information on exact tests.
Spearman Rank Correlation Coefcient

The Spearman correlation coefcient is computed using rank scores R1i and C1j ,
dened in the section Scores beginning on page 102. It is appropriate only when
both variables lie on an ordinal scale. It has the range 1 s 1. The Spearman
correlation coefcient is computed as
rs =

v
w

with
var =

nij (zij z )2

n2 w 4

where
nij R(i)C(j)

v =
i

w =
F

1
FG
12
n3
i

= n3
i

Statistical Computations
n3
j

G = n3
j

R(i) = R1i n/2

C(j) = C1j n/2
1
n

z =

nij zij
i

zij

= wvij vwij

vij

= n R(i)C(j) +

1
2

nil C(l) +
l

1
2

nkj R(k)+
k

nkl C(l) +
l

wij

k>i

nkl R(k)
k

l>j

n
F n2 + Gn2
i
j
96w

Refer to Snedecor and Cochran (1989) and Brown and Benedetti (1977).
To compute an asymptotic test for the Spearman correlation, PROC FREQ uses a

standardized test statistic rs , which has an asymptotic standard normal distribution

under the null hypothesis that the correlation equals zero. The standardized test statistic is computed as
rs

rs =

var0 (rs )

where var0 (rs ) is the variance of the correlation under the null hypothesis.
var0 (rs ) =

1
n2 w 2

nij (vij v )2

where
nij vij /n

The asymptotic variance is derived for multinomial sampling in a contingency table

framework, and it differs from the form obtained under the assumption that both
variables are continuous and normally distributed. Refer to Brown and Benedetti
(1977).
PROC FREQ also computes the exact test for the hypothesis that the Spearman rank
correlation equals zero when you specify the SCORR option in the EXACT statement. See the section Exact Statistics beginning on page 142 for information on
exact tests.

115

116

Chapter 2. The FREQ Procedure

Polychoric Correlation

When you specify the PLCORR option in the TABLES statement, PROC FREQ
computes the polychoric correlation. This measure of association is based on the
assumption that the ordered, categorical variables of the frequency table have an underlying bivariate normal distribution. For 2 2 tables, the polychoric correlation is
also known as the tetrachoric correlation. Refer to Drasgow (1986) for an overview
of polychoric correlation. The polychoric correlation coefcient is the maximum
likelihood estimate of the product-moment correlation between the normal variables,
estimating thresholds from the observed table frequencies. The range of the polychoric correlation is from -1 to 1. Olsson (1979) gives the likelihood equations and
an asymptotic covariance matrix for the estimates.
To estimate the polychoric correlation, PROC FREQ iteratively solves the likelihood
equations by a Newton-Raphson algorithm using the Pearson correlation coefcient
as the initial approximation. Iteration stops when the convergence measure falls below the convergence criterion or when the maximum number of iterations is reached,
whichever occurs rst. The CONVERGE= option sets the convergence criterion, and
the default value is 0.0001. The MAXITER= option sets the maximum number of
iterations, and the default value is 20.
Lambda Asymmetric

Asymmetric lambda, (C|R), is interpreted as the probable improvement in predicting the column variable Y given knowledge of the row variable X. Asymmetric
lambda has the range 0 (C|R) 1. It is computed as
(C|R) =

r
nr
i ri

with
var =

n i ri
(n r)3

(ri | li = l)

ri + r 2
i

where
ri = max(nij )
j

r = max(nj )
j

Also, let li be the unique value of j such that ri = nij , and let l be the unique value
of j such that r = nj .
Because of the uniqueness assumptions, ties in the frequencies or in the marginal
totals must be broken in an arbitrary but consistent manner. In case of ties, l is dened
here as the smallest value of j such that r = nj . For a given i, if there is at least one
value j such that nij = ri = cj , then li is dened here to be the smallest such value

Statistical Computations

117

of j. Otherwise, if nil = ri , then li is dened to be equal to l. If neither condition is

true, then li is taken to be the smallest value of j such that nij = ri . The formulas for
lambda asymmetric (R|C) can be obtained by interchanging the indices.
Refer to Goodman and Kruskal (1979).
Lambda Symmetric

The nondirectional lambda is the average of the two asymmetric lambdas, (C|R)
and (R|C). Lambda symmetric has the range 0 1. Lambda symmetric is
dened as
i ri

j cj

2n r c

wv
w

with

var =

1
wvy 2w2 n
w4

(nij | j = li , i = kj ) 2v 2 (n nkl )
i

where
cj

= max(nij )
i

c = max(ni )
i

w = 2n r c
ri

v = 2n
i

cj
j

(cj | kj = k) + rk + cl

(ri | li = l) +

x =
i

y = 8n w v 2x

Refer to Goodman and Kruskal (1979).

Uncertainty Coefcients (C |R ) and (R |C )

The uncertainty coefcient, U (C|R), is the proportion of uncertainty (entropy) in

the column variable Y that is explained by the row variable X. It has the range
0 U (C|R) 1. The formulas for U (R|C) can be obtained by interchanging
the indices.

118

Chapter 2. The FREQ Procedure

U (C|R) =

H(X) + H(Y ) H(XY )

v
=
H(Y )
w

with
var =

1
2 w4
n

nij H(Y ) ln
i

nij
ni

+ (H(X) H(XY )) ln

nj
n

where
v = H(X) + H(Y ) H(XY )
w = H(Y )
H(X) =
i

H(Y ) =
j

ni
ni
ln
n
n
nj
nj
ln
n
n

H(XY ) =
i

nij
nij
ln
n
n

Refer to Theil (1972, pp. 115120) and Goodman and Kruskal (1979).
Uncertainty Coefcient (U )

The uncertainty coefcient, U , is the symmetric version of the two asymmetric coefcients. It has the range 0 U 1. It is dened as
U=

2(H(X) + H(Y ) H(XY ))

H(X) + H(Y )

with
nij H(XY ) ln

var = 4
i

ni nj
(H(X) + H(Y
n2
2 (H(X) + H(Y ))4
n

)) ln

nij
n

Refer to Goodman and Kruskal (1979).

Binomial Proportion
When you specify the BINOMIAL option in the TABLES statement, PROC FREQ
computes a binomial proportion for one-way tables. By default this is the proportion
of observations in the rst variable level, or class, that appears in the output. To
specify a different level, use the LEVEL= option.
p = n1 / n

Statistical Computations
where n1 is the frequency for the rst level and n is the total frequency for the oneway table. The standard error for the binomial proportion is computed as
se() =
p

p (1 p) / n

Using the normal approximation to the binomial distribution, PROC FREQ constructs
asymptotic condence limits for p according to
p ( z/2 se() )

p
where z/2 is the 100(1 /2) percentile of the standard normal distribution. The
condence level is determined by the ALPHA= option, which, by default, equals
0.05 and produces 95% condence limits.
If you specify the BINOMIALC option, PROC FREQ includes a continuity correction of 1/2n in the asymptotic condence limits for p. The purpose of this correction
is to adjust for the difference between the normal approximation and the binomial distribution, which is a discrete distribution. Refer to Fleiss (1981). With the continuity
correction, the asymptotic condence limits for p are
p ( z/2 se() + (1/2n) )

p
Additionally, PROC FREQ computes exact condence limits for the binomial proportion using the F distribution method given in Collett (1991) and also described by
Leemis and Trivedi (1996).
PROC FREQ computes an asymptotic test of the hypothesis that the binomial proportion equals p0 , where the value of p0 is specied by the P= option in the TABLES
statement. If you do not specify a value for the P= option, PROC FREQ uses p0 = 0.5
by default. The asymptotic test statistic is
z =

p p0

p0 (1 p0 ) / n

If you specify the BINOMIALC option, PROC FREQ includes a continuity correction in the asymptotic test statistic, towards adjusting for the difference between the
normal approximation and the discrete binomial distribution. Refer to Fleiss (1981).
The continuity correction of (1/2n) is subtracted from ( p0 ) in the numerator of
p
the test statistic z if ( p0 ) is positive; otherwise, the continuity correction is added
p
to the numerator.
PROC FREQ computes one-sided and two-sided p-values for this test. When the test
statistic z is greater than zero, its expected value under the null hypothesis, PROC
FREQ computes the right-sided p-value, which is the probability of a larger value
of the statistic occurring under the null hypothesis. A small right-sided p-value supports the alternative hypothesis that the true value of the proportion is greater than
p0 . When the test statistic is less than or equal to zero, PROC FREQ computes the
left-sided p-value, which is the probability of a smaller value of the statistic occurring

119

120

Chapter 2. The FREQ Procedure

under the null hypothesis. A small left-sided p-value supports the alternative hypothesis that the true value of the proportion is less than p0 . The one-sided p-value P1 can
be expressed as
P1 = Prob ( Z > z )

if z > 0

P1 = Prob ( Z < z )

if z 0

where Z has a standard normal distribution. The two-sided p-value P2 is computed

as
P2 = Prob ( |Z| > |z| )
When you specify the BINOMIAL option in the EXACT statement, PROC FREQ
also computes an exact test of the null hypothesis H0 : p = p0 . To compute this exact
test, PROC FREQ uses the binomial probability function
n
x

Prob (X = x | p0 ) =

p0x (1 p0 ) (nx)

x = 0, 1, 2, . . . , n

where the variable X has a binomial distribution with parameters n and p0 . To compute Prob(X n1 ), PROC FREQ sums these binomial probabilities over x from
zero to n1 . To compute Prob(X n1 ), PROC FREQ sums these binomial probabilities over x from n1 to n. Then the exact one-sided p-value is
P1 = min ( Prob(X n1 | p0 ), Prob(X n1 | p0 ) )
and the exact two-sided p-value is
P2 = 2 P1

Risks and Risk Differences

The RISKDIFF option in the TABLES statement provides estimates of risks (or binomial proportions) and risk differences for 2 2 tables. This analysis may be appropriate when comparing the proportion of some characteristic for two groups, where
row 1 and row 2 correspond to the two groups, and the columns correspond to two
possible characteristics or outcomes. For example, the row variable might be a treatment or dose, and the column variable might be the response. Refer to Collett (1991),
Fleiss (1981), and Stokes, Davis, and Koch (1995).
Let the frequencies of the 2 2 table be represented as follows.

Row 1
Row 2
Total

Column 1
n11
n21
n1

Column 2
n12
n22
n2

Total
n1
n2
n

Statistical Computations
The column 1 risk for row 1 is the proportion of row 1 observations classied in
column 1,
p1|1 = n11 / n1
This estimates the conditional probability of the column 1 response, given the rst
level of the row variable.
The column 1 risk for row 2 is the proportion of row 2 observations classied in
column 1,
p1|2 = n21 / n2
and the overall column 1 risk is the proportion of all observations classied in
column 1,
p1 = n1 / n
The column 1 risk difference compares the risks for the two rows, and it is computed
as the column 1 risk for row 1 minus the column 1 risk for row 2,
(pdi )1 = p1|1 p1|2
The risks and risk difference are dened similarly for column 2.
The standard error of the column 1 risk estimate for row i is computed as
se(p1|i ) =

p1|i (1 p1|i ) / ni

The standard error of the overall column 1 risk estimate is computed as

se(p1 ) =

p1 (1 p1 ) / n

If the two rows represent independent binomial samples, the standard error for the
column 1 risk difference is computed as
se ( (pdi )1 ) =

var(p1|1 ) + var(p1|2 )

The standard errors are computed in a similar manner for the column 2 risks and risk
difference.
Using the normal approximation to the binomial distribution, PROC FREQ constructs
asymptotic condence limits for the risks and risk differences according to
est ( z/2 se(est) )

121

122

Chapter 2. The FREQ Procedure

where est is the estimate, z/2 is the 100(1 /2) percentile of the standard normal
distribution, and se(est) is the standard error of the estimate. The condence level
is determined from the value of the ALPHA= option, which, by default, equals 0.05
and produces 95% condence limits.
If you specify the RISKDIFFC option, PROC FREQ includes continuity corrections
in the asymptotic condence limits for the risks and risk differences. Continuity corrections adjust for the difference between the normal approximation and the discrete
binomial distribution. Refer to Fleiss (1981). Including a continuity correction, the
asymptotic condence limits become
est ( z/2 se(est) + cc )
where cc is the continuity correction. For the column 1 risk for row 1, cc = (1/2n1 );
for the column 1 risk for row 2, cc = (1/2n2 ); for the overall column 1 risk, cc =
(1/2n); and for the column 1 risk difference, cc = ((1/n1 + 1/n2 )/2). Continuity
corrections are computed similarly for the column 2 risks and risk difference.
PROC FREQ computes exact condence limits for the column 1, column 2, and overall risks using the F distribution method given in Collett (1991) and also described by
Leemis and Trivedi (1996). PROC FREQ does not provide exact condence limits
for the risk differences. Refer to Agresti (1992) for a discussion of issues involved in
constructing exact condence limits for differences of proportions.

Odds Ratio and Relative Risks for 2 x 2 Tables

Odds Ratio (Case-Control Studies)

The odds ratio is a useful measure of association for a variety of study designs. For a
retrospective design called a case-control study, the odds ratio can be used to estimate
the relative risk when the probability of positive response is small (Agresti 1990). In a
case-control study, two independent samples are identied based on a binary (yes-no)
response variable, and the conditional distribution of a binary explanatory variable is
examined, within xed levels of the response variable. Refer to Stokes, Davis, and
Koch (1995) and Agresti (1996).
The odds of a positive response (column 1) in row 1 is n11 /n12 . Similarly, the odds
of a positive response in row 2 is n21 /n22 . The odds ratio is formed as the ratio of
the row 1 odds to the row 2 odds. The odds ratio for 2 2 tables is dened as
OR =

n11 n22
n11 /n12
=
n21 /n22
n12 n21

The odds ratio can be any nonnegative number. When the row and column variables
are independent, the true value of the odds ratio equals 1. An odds ratio greater than 1
indicates that the odds of a positive response are higher in row 1 than in row 2. Values
less than 1 indicate the odds of positive response are higher in row 2. The strength of
association increases with the deviation from 1.
The transformation G = (OR 1)/(OR + 1) transforms the odds ratio to the range
(1, 1) with G = 0 when OR = 1; G = 1 when OR = 0; and G approaches 1

Statistical Computations
as OR approaches innity. G is the gamma statistic, which PROC FREQ computes
when you specify the MEASURES option.
The asymptotic 100(1 )% condence limits for the odds ratio are

OR exp(z v), OR exp(z v)

where
v = var(ln OR) =

1
1
1
1
+
+
+
n11 n12 n21 n22

and z is the 100(1 /2) percentile of the standard normal distribution. If any of the
four cell frequencies are zero, the estimates are not computed.
When you specify option OR in the EXACT statement, PROC FREQ computes exact
condence limits for the odds ratio. Because this is a discrete problem, the condence
coefcient for these exact condence limits is not exactly 1 but is at least 1 .
Thus, these condence limits are conservative. Refer to Agresti (1992).
PROC FREQ computes exact condence limits for the odds ratio with an algorithm
based on that presented by Thomas (1971). Refer also to Gart (1971). The following
two equations are solved iteratively for the lower and upper condence limits, 1 and
2 .
n1
i=n11
n11
i=0

n1
i
n1
i

n2
n1 i
n2
n1 i

i /
1
i=0
n1

i
2

/
i=0

n1
i
n1
i

n2
n1 i
n2
n1 i

i
1

i
2

When the odds ratio equals zero, which occurs when either n11 = 0 or n22 = 0, then
PROC FREQ sets the lower exact condence limit to zero and determines the upper
limit with level . Similarly, when the odds ratio equals innity, which occurs when
either n12 = 0 or n21 = 0, then PROC FREQ sets the upper exact condence limit
to innity and determines the lower limit with level .
Relative Risks (Cohort Studies)

These measures of relative risk are useful in cohort (prospective) study designs,
where two samples are identied based on the presence or absence of an explanatory
factor. The two samples are observed in future time for the binary (yes-no) response
variable under study. Relative risk measures are also useful in cross-sectional studies,
where two variable are observed simultaneously. Refer to Stokes, Davis, and Koch
(1995) and Agresti (1996).
The column 1 relative risk is the ratio of the column 1 risks for row 1 to row 2.
The column 1 risk for row 1 is the proportion of the row 1 observations classied in
column 1,
p1|1 = n11 / n1

123

124

Chapter 2. The FREQ Procedure

Similarly, the column 1 risk for row 2 is
p1|2 = n21 / n2
The column 1 relative risk is then computed as
RR1 =

p1|1
p1|2

A relative risk greater than 1 indicates that the probability of positive response is
greater in row 1 than in row 2. Similarly, a relative risk less than 1 indicates that
the probability of positive response is less in row 1 than in row 2. The strength of
association increases with the deviation from 1.
The asymptotic 100(1 )% condence limits for the column 1 relative risk are

RR1 exp(z v), RR1 exp(z v)

where
v = var(ln RR1 ) =

1 p1|1 1 p1|2
+
n11
n21

and z is the 100(1 /2) percentile of the standard normal distribution. If either n11
or n21 is zero, the estimates are not computed.
PROC FREQ computes the column 2 relative risks in a similar manner.

Cochran-Armitage Test for Trend

The TREND option in the TABLES statement requests the Cochran-Armitage test
for trend, which tests for trend in binomial proportions across levels of a single factor or covariate. This test is appropriate for a contingency table where one variable
has two levels and the other variable is ordinal. The two-level variable represents
the response, and the other variable represents an explanatory variable with ordered
levels. When the contingency table has two columns and R rows, PROC FREQ tests
for trend across the R levels of the row variable, and the binomial proportion is computed as the proportion of observations in the rst column. When the table has two
rows and C columns, PROC FREQ tests for trend across the C levels of the column
variable, and the binomial proportion is computed as the proportion of observations
in the rst row.
The trend test is based upon the regression coefcient for the weighted linear regression of the binomial proportions on the scores of the levels of the explanatory
variable. Refer to Margolin (1988) and Agresti (1990). If the contingency table has
two columns and R rows, the trend test statistic is computed as

R
i=1 ni1 (Ri

p1 (1 p1 )s2

Statistical Computations
where
R

ni (Ri R)2

s2 =
i=1

The row scores Ri are determined by the value of the SCORES= option in the
TABLES statement. By default, PROC FREQ uses table scores. For character variables, the table scores for the row variable are the row numbers (for example, 1 for
the rst row, 2 for the second row, and so on). For numeric variables, the table score
for each row is the numeric value of the row level. When you perform the trend
test, the explanatory variable may be numeric (for example, dose of a test substance),
and these variable values may be appropriate scores. If the explanatory variable has
ordinal levels that are not numeric, you can assign meaningful scores to the variable
levels. Sometimes equidistant scores, such as the table scores for a character variable,
may be appropriate. For more information on choosing scores for the trend test, refer
to Margolin (1988).
The null hypothesis for the Cochran-Armitage test is no trend, which means that the
binomial proportion pi1 = ni1 /ni is the same for all levels of the explanatory variable. Under this null hypothesis, the trend test statistic is asymptotically distributed as
a standard normal random variable. In addition to this asymptotic test, PROC FREQ
can compute the exact trend test, which you request by specifying the TREND option
in the EXACT statement. See the section Exact Statistics beginning on page 142
for information on exact tests.
PROC FREQ computes one-sided and two-sided p-values for the trend test. When the
test statistic is greater than its null hypothesis expected value of zero, PROC FREQ
computes the right-sided p-value, which is the probability of a larger value of the
statistic occurring under the null hypothesis. A small right-sided p-value supports
the alternative hypothesis of increasing trend in binomial proportions from row 1 to
row R. When the test statistic is less than or equal to zero, PROC FREQ outputs the
left-sided p-value. A small left-sided p-value supports the alternative of decreasing
trend.
The one-sided p-value P1 can be expressed as
P1 = Prob ( Trend Statistic > T )

if T > 0

P1 = Prob ( Trend Statistic < T )

if T 0

The two-sided p-value P2 is computed as

P2 = Prob ( |Trend Statistic| > |T | )

Jonckheere-Terpstra Test
The JT option in the TABLES statement requests the Jonckheere-Terpstra test, which
is a nonparametric test for ordered differences among classes. It tests the null hypothesis that the distribution of the response variable does not differ among classes. It is

125

126

Chapter 2. The FREQ Procedure

designed to detect alternatives of ordered class differences, which can be expressed
as 1 2 R (or 1 2 R ), with at least one of the inequalities
being strict, where i denotes the effect of class i. For such ordered alternatives, the
Jonckheere-Terpstra test can be preferable to tests of more general class difference alternatives, such as the KruskalWallis test (requested by the option WILCOXON in
the NPAR1WAY procedure). Refer to Pirie (1983) and Hollander and Wolfe (1973)
for more information about the Jonckheere-Terpstra test.
The Jonckheere-Terpstra test is appropriate for a contingency table in which an ordinal column variable represents the response. The row variable, which can be nominal
or ordinal, represents the classication variable. The levels of the row variable should
be ordered according to the ordering you want the test to detect. The order of variable levels is determined by the ORDER= option in the PROC FREQ statement. The
default is ORDER=INTERNAL, which orders by unformatted values. If you specify
ORDER=DATA, PROC FREQ orders values according to their order in the input data
set. For more information on how to order variable levels, see the ORDER= option
on page 76.
The Jonckheere-Terpstra test statistic is computed by rst forming R(R1)/2 MannWhitney counts Mi,i , where i < i , for pairs of rows in the contingency table,
Mi,i

{ number of times Xi,j < Xi ,j ,

j = 1, . . . , ni. ; j = 1, . . . , ni . }
+

1
2

{ number of times Xi,j = Xi ,j ,

j = 1, . . . , ni. ; j = 1, . . . , ni . }

where Xi,j is response j in row i. Then the Jonckheere-Terpstra test statistic is computed as
J

Mi,i
1i< i R

This test rejects the null hypothesis of no difference among classes for large values
of J. Asymptotic p-values for the Jonckheere-Terpstra test are obtained by using
the normal approximation for the distribution of the standardized test statistic. The
standardized test statistic is computed as
J =

J E0 (J)
var0 (J)

where E0 (J) and var0 (J) are the expected value and variance of the test statistic
under the null hypothesis.

E0 (J) =

n2
i

var0 (J) = A/72 + B/ [36n(n 1)(n 2)] + C/ [8n(n 1)]

Statistical Computations

127

where
nj (nj 1)(2nj + 5)

ni (ni 1)(2ni + 5)

A = n(n 1)(2n + 5)

ni (ni 1)(ni 2)

nj (nj 1)(nj 2)
j

ni (ni 1)

nj (nj 1)

In addition to this asymptotic test, PROC FREQ can compute the exact JonckheereTerpstra test, which you request by specifying the JT option in the EXACT statement.
See the section Exact Statistics beginning on page 142 for information on exact
tests.
PROC FREQ computes one-sided and two-sided p-values for the Jonckheere-Terpstra
test. When the standardized test statistic is greater than its null hypothesis expected
value of zero, PROC FREQ computes the right-sided p-value, which is the probability
of a larger value of the statistic occurring under the null hypothesis. A small rightsided p-value supports the alternative hypothesis of increasing order from row 1 to
row R. When the standardized test statistic is less than or equal to zero, PROC FREQ
computes the left-sided p-value. A small left-sided p-value supports the alternative
of decreasing order from row 1 to row R.
The one-sided p-value P1 can be expressed as
P1 = Prob ( Std JT Statistic > J )

if J > 0

P1 = Prob ( Std JT Statistic < J )

if J 0

The two-sided p-value P2 is computed as

P2 = Prob ( |Std JT Statistic| > |J | )

Tests and Measures of Agreement

When you specify the AGREE option in the TABLES statement, PROC FREQ computes tests and measures of agreement for square tables (that is, for tables where
the number of rows equals the number of columns). For two-way tables, these tests
and measures include McNemars test for 2 2 tables, Bowkers test of symmetry,
the simple kappa coefcient, and the weighted kappa coefcient. For multiple strata
(n-way tables, where n > 2), PROC FREQ computes the overall simple kappa coefcient and the overall weighted kappa coefcient, as well as tests for equal kappas
(simple and weighted) among strata. Cochrans Q is computed for multi-way tables
when each variable has two levels, that is, for 2 2 2 tables.

128

Chapter 2. The FREQ Procedure

PROC FREQ computes the kappa coefcients (simple and weighted), their asymptotic standard errors, and their condence limits when you specify the AGREE option
in the TABLES statement. If you also specify the KAPPA option in the TEST statement, then PROC FREQ computes the asymptotic test of the hypothesis that simple
kappa equals zero. Similarly, if you specify the WTKAP option in the TEST statement, PROC FREQ computes the asymptotic test for weighted kappa.
In addition to the asymptotic tests described in this section, PROC FREQ computes
the exact p-value for McNemars test when you specify the option MCNEM in the
EXACT statement. For the kappa statistics, PROC FREQ computes the exact test
of the hypothesis that kappa (or weighted kappa) equals zero when you specify
the option KAPPA (or WTKAP) in the EXACT statement. See the section Exact
Statistics beginning on page 142 for information on exact tests.
The discussion of each test and measures of agreement provides the formulas that
PROC FREQ uses to compute the AGREE statistics. For information on the use and
interpretation of these statistics, refer to Agresti (1990), Agresti (1996), Fleiss (1981),
and the other references cited for each statistic.

McNemars Test

PROC FREQ computes McNemars test for 22 tables when you specify the AGREE
option. McNemars test is appropriate when you are analyzing data from matched
pairs of subjects with a dichotomous (yes-no) response. It tests the null hypothesis of
marginal homogeneity, or p1 = p1 . McNemars test is computed as
QM =

(n12 n21 )2
n12 + n21

Under the null hypothesis, QM has an asymptotic chi-square distribution with one
degree of freedom. Refer to McNemar (1947), as well as the references cited in the
preceding section. In addition to the asymptotic test, PROC FREQ also computes
the exact p-value for McNemars test when you specify the MCNEM option in the
EXACT statement.
Bowkers Test of Symmetry

For Bowkers test of symmetry, the null hypothesis is that the probabilities in the
square table satisfy symmetry or that pij = pji for all pairs of table cells. When there
are more than two categories, Bowkers test of symmetry is calculated as
QB =
i<j

(nij nji )2
nij + nji

For large samples, QB has an asymptotic chi-square distribution with R(R 1)/2
degrees of freedom under the null hypothesis of symmetry of the expected counts.
Refer to Bowker (1948). For two categories, this test of symmetry is identical to
McNemars test.

Statistical Computations
Simple Kappa Coefcient

The simple kappa coefcient, introduced by Cohen (1960), is a measure of interrater

agreement:
=

Po Pe
1 Pe

where Po = i pii and Pe = i pi. p.i . If the two response variables are viewed as
two independent ratings of the n subjects, the kappa coefcient equals +1 when there
is complete agreement of the raters. When the observed agreement exceeds chance
agreement, kappa is positive, with its magnitude reecting the strength of agreement.
Although this is unusual in practice, kappa is negative when the observed agreement
is less than chance agreement. The minimum value of kappa is between 1 and 0,
depending on the marginal proportions.
The asymptotic variance of the simple kappa coefcient can be estimated by the following, according to Fleiss, Cohen, and Everitt (1969):
A+BC
(1 Pe )2 n

var =
where

pii 1 (pi + pi )(1 )

A=
i

B = (1 )2

pij (pi + pj )2
i=j

and
2

C = Pe (1 )

PROC FREQ computes condence limits for the simple kappa coefcient according
to
( z/2

var )

where z/2 is the 100(1 /2) percentile of the standard normal distribution. The
value of is determined by the value of the ALPHA= option, which, by default,
equals 0.05 and produces 95% condence limits.
To compute an asymptotic test for the kappa coefcient, PROC FREQ uses a standardized test statistic , which has an asymptotic standard normal distribution under

129

130

Chapter 2. The FREQ Procedure

the null hypothesis that kappa equals zero. The standardized test statistic is computed
as

var0 ( )

where var0 ( ) is the variance of the kappa coefcient under the null hypothesis.

var0 ( ) =

2
Pe + Pe i pi pi (pi + pi )
(1 Pe )2 n

Refer to Fleiss (1981).

In addition to the asymptotic test for kappa, PROC FREQ computes the exact test
when you specify the KAPPA or AGREE option in the EXACT statement. See the
section Exact Statistics beginning on page 142 for information on exact tests.
Weighted Kappa Coefcient

The weighted kappa coefcient is a generalization of the simple kappa coefcient,

using weights to quantify the relative difference between categories. For 2 2 tables,
the weighted kappa coefcient equals the simple kappa coefcient. PROC FREQ
displays the weighted kappa coefcient only for tables larger than 2 2. PROC
FREQ computes the weights from the column scores, using either the CicchettiAllison weight type or the Fleiss-Cohen weight type, both of which are described
in the following section. The weights wij are constructed so that 0 wij < 1 for all
i = j, wii = 1 for all i, and wij = wji . The weighted kappa coefcient is dened as
w =

Po(w) Pe(w)
1 Pe(w)

where
wij pij

Po(w) =
i

and
wij pi pj

Pe(w) =

The asymptotic variance of the weighted kappa coefcient can be estimated by the
following, according to Fleiss, Cohen, and Everitt (1969):
2

2
i

var =

pij wij (wi + wj )(1 w )

(1 Pe(w) )2 n

w Pe(w) (1 w )

Statistical Computations
where
pj wij

wi =
j

and
pi wij

wj =
i

PROC FREQ computes condence limits for the weighted kappa coefcient according to
w ( z/2

var )

where z/2 is the 100(1 /2) percentile of the standard normal distribution. The
value of is determined by the value of the ALPHA= option, which, by default,
equals 0.05 and produces 95% condence limits.
To compute an asymptotic test for the weighted kappa coefcient, PROC FREQ uses
a standardized test statistic , which has an asymptotic standard normal distribution
w
under the null hypothesis that weighted kappa equals zero. The standardized test
statistic is computed as
=
w

var0 ( w )

where var0 ( w ) is the variance of the weighted kappa coefcient under the null hy
pothesis.
2
i

var0 ( w ) =

j pi pj wij (w i + w j )

2
Pe(w)

(1 Pe(w) )2 n

Refer to Fleiss (1981).

In addition to the asymptotic test for weighted kappa, PROC FREQ computes the
exact test when you specify the WTKAP or AGREE option in the EXACT statement.
See the section Exact Statistics beginning on page 142 for information on exact
tests.
Weights

PROC FREQ computes kappa coefcient weights using the column scores and one of
two available weight types. The column scores are determined by the SCORES= option in the TABLES statement. The two available weight types are Cicchetti-Allison
and Fleiss-Cohen, and PROC FREQ uses the Cicchetti-Allison type by default. If you
specify (WT=FC) with the AGREE option, then PROC FREQ uses the Fleiss-Cohen
weight type to construct kappa weights.

131

132

Chapter 2. The FREQ Procedure

PROC FREQ computes Cicchetti-Allison kappa coefcient weights using a form similar to that given by Cicchetti and Allison (1971).
wij = 1

|Ci Cj |
CC C1

where Ci is the score for column i, and C is the number of categories or columns.
You can specify the score type using the SCORES= option in the TABLES statement; if you do not specify the SCORES= option, PROC FREQ uses table scores.
For numeric variables, table scores are the values of the numeric row and column
headings. You can assign numeric values to the categories in a way that reects their
level of similarity. For example, suppose you have four categories and order them
according to similarity. If you assign them values of 0, 2, 4, and 10, the following
weights are used for computing the weighted kappa coefcient: w12 = 0.8, w13 = 0.6,
w14 = 0, w23 = 0.8, w24 = 0.2, and w34 = 0.4. Note that when there are only two
categories (that is, C = 2), the weighted kappa coefcient is identical to the simple
kappa coefcient.
If you specify (WT=FC) with the AGREE option in the TABLES statement, PROC
FREQ computes Fleiss-Cohen kappa coefcient weights using a form similar to that
given by Fleiss and Cohen (1973).
wij = 1

(Ci Cj )2
(CC C1 )2

For the preceding example, the weights used for computing the weighted kappa coefcient are: w12 = 0.96, w13 = 0.84, w14 = 0, w23 = 0.96, w24 = 0.36, and w34 =
0.64.
Overall Kappa Coefcient

When there are multiple strata, PROC FREQ combines the stratum-level estimates of
kappa into an overall estimate of the supposed common value of kappa. Assume there
are q strata, indexed by h = 1, 2, . . . , q, and let var( h ) denote the squared standard

error of h . Then the estimate of the overall kappa, according to Fleiss (1981), is

computed as
q

overall =

h=1

/
var( h )

h=1

1
var( h )

PROC FREQ computes an estimate of the overall weighted kappa in a similar manner.
Tests for Equal Kappa Coefcients

When there are multiple strata, the following chi-square statistic tests whether the
stratum-level values of kappa are equal.
q

QK =
h=1

( h overall )2

var( h )

Statistical Computations
Under the null hypothesis of equal kappas over the q strata, QK has an asymptotic
chi-square distribution with q 1 degrees of freedom. PROC FREQ computes a test
for equal weighted kappa coefcients in a similar manner.
Cochrans Q Test

Cochrans Q is computed for multi-way tables when each variable has two levels,
that is, for 2 2 2 tables. Cochrans Q statistic is used to test the homogeneity
of the one-dimensional margins. Let m denote the number of variables and N denote
the total number of subjects. Then Cochrans Q statistic is computed as
QC = (m 1)

m
mT

m
2
2
j=1 Tj T
2
N Sk
k=1

where Tj is the number of positive responses for variable j, T is the total number
of positive responses over all variables, and Sk is the number of positive responses
for subject k. Under the null hypothesis, Cochrans Q is an approximate chi-square
statistic with m 1 degrees of freedom. Refer to Cochran (1950). When there are
only two binary response variables (m = 2), Cochrans Q simplies to McNemars
test. When there are more than two response categories, you can test for marginal
homogeneity using the repeated measures capabilities of the CATMOD procedure.
Tables with Zero Rows and Columns

The AGREE statistics are dened only for square tables, where the number of rows
equals the number of columns. If the table is not square, PROC FREQ does not
compute AGREE statistics. In the kappa statistic framework, where two independent
raters assign ratings to each of n subjects, suppose one of the raters does not use all
possible r rating levels. If the corresponding table has r rows but only r 1 columns,
then the table is not square, and PROC FREQ does not compute the AGREE statistics.
To create a square table in this situation, use the ZEROS option in the WEIGHT
statement, which requests that PROC FREQ include observations with zero weights
in the analysis. And input zero-weight observations to represent any rating levels that
are not used by a rater, so that the input data set has at least one observation for each
possible rater and rating combination. This includes all rating levels in the analysis,
whether or not all levels are actually assigned by both raters. The resulting table is a
square table, r r, and so all AGREE statistics can be computed.
For more information, see the description of the ZEROS option. By default, PROC
FREQ does not process observations that have zero weights, because these observations do not contribute to the total frequency count, and because any resulting zeroweight row or column causes many of the tests and measures of association to be
undened. However, kappa statistics are dened for tables with a zero-weight row or
column, and the ZEROS option allows input of zero-weight observations so you can
construct the tables needed to compute kappas.

133

134

Chapter 2. The FREQ Procedure

Cochran-Mantel-Haenszel Statistics
For n-way crosstabulation tables, consider the following example:
proc freq;
tables A*B*C*D / cmh;
run;

The CMH option in the TABLES statement gives a stratied statistical analysis of the
relationship between C and D, after controlling for A and B. The stratied analysis
provides a way to adjust for the possible confounding effects of A and B without being forced to estimate parameters for them. The analysis produces Cochran-MantelHaenszel statistics, and for 2 2 tables, it includes estimation of the common odds
ratio, common relative risks, and the Breslow-Day test for homogeneity of the odds
ratios.
Let the number of strata be denoted by q, indexing the strata by h = 1, 2, . . . , q.
Each stratum contains a contingency table with X representing the row variable and
Y representing the column variable. For table h, denote the cell frequency in row i
and column j by nhij , with corresponding row and column marginal totals denoted
by nhi. and nh.j , and the overall stratum total by nh .
Because the formulas for the Cochran-Mantel-Haenszel statistics are more easily dened in terms of matrices, the following notation is used. Vectors are presumed to be
column vectors unless they are transposed ( ).
(1 C)

nhi

(nhi1 , nhi2 , . . . , nhiC )

(nh1 , nh2 , . . . , nhR )

phi

(1 1)

phj

nhi
nh
nhj
nh

(ph1 , ph2 , . . . , phR )

(1 R)

(ph1 , ph2 , . . . , phC )

(1 C)

(1 RC)

(1 1)

Assume that the strata are independent and that the marginal totals of each stratum
are xed. The null hypothesis, H0 , is that there is no association between X and Y
in any of the strata. The corresponding model is the multiple hypergeometric; this
implies that, under H0 , the expected value and covariance matrix of the frequencies
are, respectively,
mh = E[nh | H0 ] = nh (Ph Ph )
and
var[nh | H0 ] = c (DPh Ph Ph ) (DPh Ph Ph )

Statistical Computations
where
c=

n2
h
nh 1

and where denotes Kronecker product multiplication and Da is a diagonal matrix

with elements of a on the main diagonal.
The generalized CMH statistic (Landis, Heyman, and Koch 1978) is dened as
QCMH = G VG 1 G
where
Bh (nh mh )

G =
h

Bh (Var(nh | H0 )) Bh

VG =
h

and where
Bh = Ch Rh
is a matrix of xed constants based on column scores Ch and row scores Rh . When
the null hypothesis is true, the CMH statistic has an asymptotic chi-square distribution
with degrees of freedom equal to the rank of Bh . If VG is found to be singular, PROC
FREQ prints a message and sets the value of the CMH statistic to missing.
PROC FREQ computes three CMH statistics using this formula for the generalized
CMH statistic, with different row and column score denitions for each statistic. The
CMH statistics that PROC FREQ computes are the correlation statistic, the ANOVA
(row mean scores) statistic, and the general association statistic. These statistics test
the null hypothesis of no association against different alternative hypotheses. The
following sections describe the computation of these CMH statistics.
CAUTION: The CMH statistics have low power for detecting an association in which
the patterns of association for some of the strata are in the opposite direction of the
patterns displayed by other strata. Thus, a nonsignicant CMH statistic suggests
either that there is no association or that no pattern of association has enough strength
or consistency to dominate any other pattern.
Correlation Statistic

The correlation statistic, popularized by Mantel and Haenszel (1959) and Mantel
(1963), has one degree of freedom and is known as the Mantel-Haenszel statistic.
The alternative hypothesis for the correlation statistic is that there is a linear association between X and Y in at least one stratum. If either X or Y does not lie on an
ordinal (or interval) scale, then this statistic is not meaningful.

135

136

Chapter 2. The FREQ Procedure

To compute the correlation statistic, PROC FREQ uses the formula for the generalized CMH statistic with the row and column scores determined by the SCORES=
option in the TABLES statement. See the section Scores on page 102 for more
information on the available score types. The matrix of row scores Rh has dimension
1 R, and the matrix of column scores Ch has dimension 1 C.
When there is only one stratum, this CMH statistic reduces to (n 1)r2 , where r is
the Pearson correlation coefcient between X and Y . When nonparametric (RANK
2
or RIDIT) scores are specied, then the statistic reduces to (n 1)rs , where rs is the
Spearman rank correlation coefcient between X and Y. When there is more than one
stratum, then this CMH statistic becomes a stratum-adjusted correlation statistic.

ANOVA (Row Mean Scores) Statistic

The ANOVA statistic can be used only when the column variable Y lies on an ordinal
(or interval) scale so that the mean score of Y is meaningful. For the ANOVA statistic,
the mean score is computed for each row of the table, and the alternative hypothesis
is that, for at least one stratum, the mean scores of the R rows are unequal. In other
words, the statistic is sensitive to location differences among the R distributions of
Y.
The matrix of column scores Ch has dimension 1 C, the column scores are determined by the SCORES= option.
The matrix of row scores Rh has dimension (R 1) R and is created internally by
PROC FREQ as
Rh = [IR1 , JR1 ]
where IR1 is an identity matrix of rank R 1, and JR1 is an (R 1) 1 vector
of ones. This matrix has the effect of forming R 1 independent contrasts of the R
mean scores.
When there is only one stratum, this CMH statistic is essentially an analysis of variance (ANOVA) statistic in the sense that it is a function of the variance ratio F statistic that would be obtained from a one-way ANOVA on the dependent variable Y. If
nonparametric scores are specied in this case, then the ANOVA statistic is a KruskalWallis test.
If there is more than one stratum, then this CMH statistic corresponds to a stratumadjusted ANOVA or Kruskal-Wallis test. In the special case where there is one subject
per row and one subject per column in the contingency table of each stratum, this
CMH statistic is identical to Friedmans chi-square. See Example 2.8 on page 180
for an illustration.
General Association Statistic

The alternative hypothesis for the general association statistic is that, for at least one
stratum, there is some kind of association between X and Y. This statistic is always
interpretable because it does not require an ordinal scale for either X or Y.

Statistical Computations
For the general association statistic, the matrix Rh is the same as the one used for the
ANOVA statistic. The matrix Ch is dened similarly as
Ch = [IC1 , JC1 ]
PROC FREQ generates both score matrices internally. When there is only one stratum, then the general association CMH statistic reduces to QP (n 1)/n, where QP
is the Pearson chi-square statistic. When there is more than one stratum, then the
CMH statistic becomes a stratum-adjusted Pearson chi-square statistic. Note that a
similar adjustment can be made by summing the Pearson chi-squares across the strata.
However, the latter statistic requires a large sample size in each stratum to support the
resulting chi-square distribution with q(R1)(C 1) degrees of freedom. The CMH
statistic requires only a large overall sample size since it has only (R 1)(C 1)
degrees of freedom.
Refer to Cochran (1954); Mantel and Haenszel (1959); Mantel (1963); Birch (1965);
Landis, Heyman, and Koch (1978).
Adjusted Odds Ratio and Relative Risk Estimates

The CMH option provides adjusted odds ratio and relative risk estimates for stratied
22 tables. For each of these measures, PROC FREQ computes the Mantel-Haenszel
estimate and the logit estimate. These estimates apply to n-way table requests in the
TABLES statement, when the row and column variables both have only two levels.
For example,
proc freq;
tables A*B*C*D / cmh;
run;

In this example, if the row and columns variables C and D both have two levels,
PROC FREQ provides odds ratio and relative risk estimates, adjusting for the confounding variables A and B.
The choice of an appropriate measure depends on the study design. For case-control
(retrospective) studies, the odds ratio is appropriate. For cohort (prospective) or crosssectional studies, the relative risk is appropriate. See the section Odds Ratio and
Relative Risks for 2 x 2 Tables beginning on page 122 for more information on
these measures.
Throughout this section, z denotes the 100(1/2) percentile of the standard normal
distribution.
Odds Ratio, Case-Control Studies
Mantel-Haenszel Estimator
The Mantel-Haenszel estimate of the common odds ratio is computed as
ORMH =

h nh11

nh22 /nh
h nh12 nh21 /nh

137

138

Chapter 2. The FREQ Procedure

It is always computed unless the denominator is zero. Refer to Mantel and Haenszel
(1959) and Agresti (1990).
Using the estimated variance for log(ORMH ) given by Robins, Breslow, and
Greenland (1986), PROC FREQ computes the corresponding 100(1 )% condence limits for the odds ratio as
( ORMH exp(z ), ORMH exp(z ) )

where
2 = var[ ln(ORMH ) ]

h (nh11

+ nh22 )(nh11 nh22 )/n2

h
h nh11

2(
+

h [(nh11

h (nh12

nh22 /nh )2

+ nh22 )(nh12 nh21 ) + (nh12 + nh21 )(nh11 nh22 )]/n2

h
2 ( h nh11 nh22 /nh ) ( h nh12 nh21 /nh )

+ nh21 )(nh12 nh21 )/n2

h nh12

nh21 /nh )2

Note that the Mantel-Haenszel odds ratio estimator is less sensitive to small nh than
the logit estimator.
Logit Estimator
The adjusted logit estimate of the odds ratio (Woolf 1955) is computed as
ORL = exp

h wh ln(ORh )
h wh

and the corresponding 100(1 )% condence limits are

ORL exp

z
h wh

, ORL exp

z
h wh

where ORh is the odds ratio for stratum h, and

wh =

1
var(ln ORh )

If any cell frequency in a stratum h is zero, then PROC FREQ adds 0.5 to each cell
of the stratum before computing ORh and wh (Haldane 1955), and prints a warning.
Exact Condence Limits for the Common Odds Ratio
When you specify the COMOR option in the EXACT statement, PROC FREQ computes exact condence limits for the common odds ratio for stratied 2 2 tables.

Statistical Computations
This computation assumes that the odds ratio is constant over all the 2 2 tables.
Exact condence limits are constructed from the distribution of S =
h nh11 , conditional on the marginal totals of the 2 2 tables.
Because this is a discrete problem, the condence coefcient for these exact condence limits is not exactly 1 but is at least 1 . Thus, these condence limits
are conservative. Refer to Agresti (1992).
PROC FREQ computes exact condence limits for the common odds ratio with an
algorithm based on that presented by Vollset, Hirji, and Elashoff (1991). Refer also
to Mehta, Patel, and Gray (1985).
Conditional on the marginal totals of 2 2 table h, let the random variable Sh denote
the frequency of table cell (1, 1). Given the row totals nh1 and nh2 and column
totals nh1 and nh2 , the lower and upper bounds for Sh are lh and uh ,
lh

max ( 0, nh1 nh2 )

min ( nh1 , nh1 )

Let Csh denote the hypergeometric coefcient,

Csh

nh2
nh1 sh

nh1
sh

and let denote the common odds ratio. Then the conditional distribution of Sh is
x = uh

P ( Sh = sh | n1 , n1 , n2 ) = Csh sh /

Cx x
x = lh

Summing over all the 2 2 tables, S =

of S are l and u,
lh

l =

Sh , and the lower and upper bounds

u =

and

The conditional distribution of the sum S is

and the corresponding 100(1 )% condence limits are

RRL exp

z
h wh

, RRL exp

z
h wh

141

142

Chapter 2. The FREQ Procedure

where RRh is the column 1 relative risk estimate for stratum h, and
wh =

1
var(ln RRh )

If nh11 or nh21 is zero, then PROC FREQ adds 0.5 to each cell of the stratum before
computing RRh and wh , and prints a warning. Refer to Kleinbaum, Kupper, and
Morgenstern (1982, Sections 17.4 and 17.5).
Breslow-Day Test for Homogeneity of the Odds Ratios

When you specify the CMH option, PROC FREQ computes the Breslow-Day test for
stratied analysis of 22 tables. It tests the null hypothesis that the odds ratios for the
q strata are all equal. When the null hypothesis is true, the statistic has approximately
a chi-square distribution with q 1 degrees of freedom. Refer to Breslow and Day
(1980) and Agresti (1996).
The Breslow-Day statistic is computed as
QBD =
h

(nh11 E(nh11 | ORMH ))2

var(nh11 | ORMH )

where E and var denote expected value and variance, respectively. The summation
does not include any table with a zero row or column. If ORMH equals zero or if it
is undened, then PROC FREQ does not compute the statistic and prints a warning
message.
For the Breslow-Day test to be valid, the sample size should be relatively large in
each stratum, and at least 80% of the expected cell counts should be greater than
5. Note that this is a stricter sample size requirement than the requirement for the
Cochran-Mantel-Haenszel test for q 2 2 tables, in that each stratum sample size
(not just the overall sample size) must be relatively large. Even when the BreslowDay test is valid, it may not be very powerful against certain alternatives, as discussed
in Breslow and Day (1980).
If you specify the BDT option, PROC FREQ computes the Breslow-Day test with
Tarones adjustment, which subtracts an adjustment factor from QBD to make the
resulting statistic asymptotically chi-square.
QBDT = QBD

E(nh11 | ORMH )))2

h var(nh11 | ORMH )

h (nh11

Refer to Tarone (1985), Jones et al. (1989), and Breslow (1996).

Exact Statistics
Exact statistics can be useful in situations where the asymptotic assumptions are not
met, and so the asymptotic p-values are not close approximations for the true pvalues. Standard asymptotic methods involve the assumption that the test statistic
follows a particular distribution when the sample size is sufciently large. When the

Statistical Computations
sample size is not large, asymptotic results may not be valid, with the asymptotic
p-values differing perhaps substantially from the exact p-values. Asymptotic results
may also be unreliable when the distribution of the data is sparse, skewed, or heavily tied. Refer to Agresti (1996) and Bishop, Fienberg, and Holland (1975). Exact
computations are based on the statistical theory of exact conditional inference for
contingency tables, reviewed by Agresti (1992).
In addition to computation of exact p-values, PROC FREQ provides the option of
estimating exact p-values by Monte Carlo simulation. This can be useful for problems
that are so large that exact computations require a great amount of time and memory,
but for which asymptotic approximations may not be sufcient.
PROC FREQ provides exact p-values for the following tests for two-way tables:
Pearson chi-square, likelihood-ratio chi-square, Mantel-Haenszel chi-square, Fishers
exact test, Jonckheere-Terpstra test, Cochran-Armitage test for trend, and McNemars
test. PROC FREQ also computes exact p-values for tests of hypotheses that the following statistics equal zero: Pearson correlation coefcient, Spearman correlation
coefcient, simple kappa coefcient, and weighted kappa coefcient. Additionally,
PROC FREQ computes exact condence limits for the odds ratio for 2 2 tables. For
stratied 2 2 tables, PROC FREQ computes exact condence limits for the common odds ratio, as well as an exact test that the common odds ratio equals one. For
one-way frequency tables, PROC FREQ provides the exact chi-square goodness-oft test (for equal proportions or for proportions or frequencies that you specify). Also
for one-way tables, PROC FREQ provides exact condence limits for the binomial
proportion and an exact test for the binomial proportion value.
The following sections summarize the exact computational algorithms, dene the
exact p-values that PROC FREQ computes, discuss the computational resource requirements, and describe the Monte Carlo estimation option.
Computational Algorithms

PROC FREQ computes exact p-values for general R C tables using the network algorithm developed by Mehta and Patel (1983). This algorithm provides a substantial
advantage over direct enumeration, which can be very time-consuming and feasible
only for small problems. Refer to Agresti (1992) for a review of algorithms for computation of exact p-values, and refer to Mehta, Patel, and Tsiatis (1984) and Mehta,
Patel, and Senchaudhuri (1991) for information on the performance of the network
algorithm.
The reference set for a given contingency table is the set of all contingency tables
with the observed marginal row and column sums. Corresponding to this reference
set, the network algorithm forms a directed acyclic network consisting of nodes in a
number of stages. A path through the network corresponds to a distinct table in the
reference set. The distances between nodes are dened so that the total distance of a
path through the network is the corresponding value of the test statistic. At each node,
the algorithm computes the shortest and longest path distances for all the paths that
pass through that node. For statistics that can be expressed as a linear combination
of cell frequencies multiplied by increasing row and column scores, PROC FREQ
computes shortest and longest path distances using the algorithm given in Agresti,
Mehta, and Patel (1990). For statistics of other forms, PROC FREQ computes an

143

144

Chapter 2. The FREQ Procedure

upper bound for the longest path and a lower bound for the shortest path, following
the approach of Valz and Thompson (1994).
The longest and shortest path distances or bounds for a node are compared to the
value of the test statistic to determine whether all paths through the node contribute
to the p-value, none of the paths through the node contribute to the p-value, or neither
of these situations occur. If all paths through the node contribute, the p-value is incremented accordingly, and these paths are eliminated from further analysis. If no paths
contribute, these paths are eliminated from the analysis. Otherwise, the algorithm
continues, still processing this node and the associated paths. The algorithm nishes
when all nodes have been accounted for, incrementing the p-value accordingly, or
eliminated.
In applying the network algorithm, PROC FREQ uses full precision to represent all
statistics, row and column scores, and other quantities involved in the computations.
Although it is possible to use rounding to improve the speed and memory requirements of the algorithm, PROC FREQ does not do this since it can result in reduced
accuracy of the p-values.
For one-way tables, PROC FREQ computes the exact chi-square goodness-of-t test
by the method of Radlow and Alf (1975). PROC FREQ generates all possible oneway tables with the observed total sample size and number of categories. For each
possible table, PROC FREQ compares its chi-square value with the value for the observed table. If the tables chi-square value is greater than or equal to the observed
chi-square, PROC FREQ increments the exact p-value by the probability of that table, which is calculated under the null hypothesis using the multinomial frequency
distribution. By default, the null hypothesis states that all categories have equal proportions. If you specify null hypothesis proportions or frequencies using the TESTP=
or TESTF= option in the TABLES statement, then PROC FREQ calculates the exact
chi-square test based on that null hypothesis.
For binomial proportions in one-way tables, PROC FREQ computes exact condence
limits using the F distribution method given in Collett (1991) and also described by
Leemis and Trivedi (1996). PROC FREQ computes the exact test for a binomial
proportion (H0 : p = p0 ) by summing binomial probabilities over all alternatives. See
the section Binomial Proportion on page 118 for details. By default, PROC FREQ
uses p0 = 0.5 as the null hypothesis proportion. Alternatively, you can specify the
null hypothesis proportion with the P= option in the TABLES statement.
See the section Odds Ratio and Relative Risks for 2 x 2 Tables on page 122 for
details on computation of exact condence limits for the odds ratio for 2 2 tables.
See the section Exact Condence Limits for the Common Odds Ratio on page 138
for details on computation of exact condence limits for the common odds ratio for
stratied 2 2 tables.

Denition of p-Values

For several tests in PROC FREQ, the test statistic is nonnegative, and large values of
the test statistic indicate a departure from the null hypothesis. Such tests include the
Pearson chi-square, the likelihood-ratio chi-square, the Mantel-Haenszel chi-square,
Fishers exact test for tables larger than 2 2 tables, McNemars test, and the one-

Statistical Computations
way chi-square goodness-of-t test. The exact p-value for these nondirectional tests
is the sum of probabilities for those tables having a test statistic greater than or equal
to the value of the observed test statistic.
There are other tests where it may be appropriate to test against either a one-sided or a
two-sided alternative hypothesis. For example, when you test the null hypothesis that
the true parameter value equals 0 (T = 0), the alternative of interest may be one-sided
(T 0, or T 0) or two-sided (T = 0). Such tests include the Pearson correlation coefcient, Spearman correlation coefcient, Jonckheere-Terpstra test, CochranArmitage test for trend, simple kappa coefcient, and weighted kappa coefcient. For
these tests, PROC FREQ outputs the right-sided p-value when the observed value of
the test statistic is greater than its expected value. The right-sided p-value is the sum
of probabilities for those tables having a test statistic greater than or equal to the
observed test statistic. Otherwise, when the test statistic is less than or equal to its
expected value, PROC FREQ outputs the left-sided p-value. The left-sided p-value
is the sum of probabilities for those tables having a test statistic less than or equal to
the one observed. The one-sided p-value P1 can be expressed as
P1 = Prob (Test Statistic t)

if t > E0 (T )

P1 = Prob (Test Statistic t)

if t E0 (T )

where t is the observed value of the test statistic and E0 (T ) is the expected value of
the test statistic under the null hypothesis. PROC FREQ computes the two-sided pvalue as the sum of the one-sided p-value and the corresponding area in the opposite
tail of the distribution of the statistic, equidistant from the expected value. The twosided p-value P2 can be expressed as
P2 = Prob ( | Test Statistic E0 (T ) | | t E0 (T ) | )
If you specify the POINT option in the EXACT statement, PROC FREQ also displays
exact point probabilities for the test statistics. The exact point probability is the exact
probability that the test statistic equals the observed value.
Computational Resources

PROC FREQ uses relatively fast and efcient algorithms for exact computations.
These recently developed algorithms, together with improvements in computer
power, make it feasible now to perform exact computations for data sets where previously only asymptotic methods could be applied. Nevertheless, there are still large
problems that may require a prohibitive amount of time and memory for exact computations, depending on the speed and memory available on your computer. For large
problems, consider whether exact methods are really needed or whether asymptotic
methods might give results quite close to the exact results, while requiring much less
computer time and memory. When asymptotic methods may not be sufcient for
such large problems, consider using Monte Carlo estimation of exact p-values, as
described in the section Monte Carlo Estimation on page 146.

145

146

Chapter 2. The FREQ Procedure

A formula does not exist that can predict in advance how much time and memory
are needed to compute an exact p-value for a certain problem. The time and memory
required depend on several factors, including which test is being performed, the total
sample size, the number of rows and columns, and the specic arrangement of the
observations into table cells. Generally, larger problems (in terms of total sample size,
number of rows, and number of columns) tend to require more time and memory.
Additionally, for a xed total sample size, time and memory requirements tend to
increase as the number of rows and columns increases, since this corresponds to an
increase in the number of tables in the reference set. Also for a xed sample size, time
and memory requirements increase as the marginal row and column totals become
more homogeneous. Refer to Agresti, Mehta, and Patel (1990) and Gail and Mantel
(1977).
At any time while PROC FREQ is computing exact p-values, you can terminate
the computations by pressing the system interrupt key sequence (refer to the SAS
Companion for your system) and choosing to stop computations. After you terminate
exact computations, PROC FREQ completes all other remaining tasks. The procedure produces the requested output and reports missing values for any exact p-values
that were not computed by the time of termination.
You can also use the MAXTIME= option in the EXACT statement to limit the amount
of time PROC FREQ uses for exact computations. You specify a MAXTIME= value
that is the maximum amount of clock time (in seconds) that PROC FREQ can use to
compute an exact p-value. If PROC FREQ does not nish computing an exact p-value
within that time, it terminates the computation and completes all other remaining
tasks.

Monte Carlo Estimation

If you specify the option MC in the EXACT statement, PROC FREQ computes Monte
Carlo estimates of the exact p-values instead of directly computing the exact p-values.
Monte Carlo estimation can be useful for large problems that require a great amount
of time and memory for exact computations but for which asymptotic approximations
may not be sufcient. To describe the precision of each Monte Carlo estimate, PROC
FREQ provides the asymptotic standard error and 100(1)% condence limits. The
condence level is determined by the ALPHA= option in the EXACT statement,
which, by default, equals 0.01, and produces 99% condence limits. The N=n option
in the EXACT statement species the number of samples that PROC FREQ uses for
Monte Carlo estimation; the default is 10000 samples. You can specify a larger value
for n to improve the precision of the Monte Carlo estimates. Because larger values
of n generate more samples, the computation time increases. Alternatively, you can
specify a smaller value of n to reduce the computation time.
To compute a Monte Carlo estimate of an exact p-value, PROC FREQ generates a
random sample of tables with the same total sample size, row totals, and column totals as the observed table. PROC FREQ uses the algorithm of Agresti, Wackerly, and
Boyett (1979), which generates tables in proportion to their hypergeometric probabilities conditional on the marginal frequencies. For each sample table, PROC FREQ
computes the value of the test statistic and compares it to the value for the observed
table. When estimating a right-sided p-value, PROC FREQ counts all sample tables

Computational Resources
for which the test statistic is greater than or equal to the observed test statistic. Then
the p-value estimate equals the number of these tables divided by the total number of
tables sampled.

PMC =

M /N

= number of samples with (Test Statistic t)

= total number of samples

t = observed Test Statistic

PROC FREQ computes left-sided and two-sided p-value estimates in a similar manner. For left-sided p-values, PROC FREQ evaluates whether the test statistic for each
sampled table is less than or equal to the observed test statistic. For two-sided pvalues, PROC FREQ examines the sample test statistics according to the expression
for P2 given in the section Asymptotic Tests on page 109. The variable M is a binomially distributed variable with N trials and success probability p. It follows that
the asymptotic standard error of the Monte Carlo estimate is

se(PMC ) =

PMC (1 PMC )/(N 1)

PROC FREQ constructs asymptotic condence limits for the p-values according to

PMC z/2 se(PMC )

where z/2 is the 100(1/2) percentile of the standard normal distribution, and the
condence level is determined by the ALPHA= option in the EXACT statement.

When the Monte Carlo estimate PMC equals 0, then PROC FREQ computes the
condence limits for the p-value as
( 0, 1 (1/N ) )

When the Monte Carlo estimate PM C equals 1, then PROC FREQ computes the
condence limits as
( (1/N ) , 1 )

Computational Resources
For each variable in a table request, PROC FREQ stores all of the levels in memory.
If all variables are numeric and not formatted, this requires about 84 bytes for each
variable level. When there are character variables or formatted numeric variables,
the memory that is required depends on the formatted variable lengths, with longer
formatted lengths requiring more memory. The number of levels for each variable is
limited only by the largest integer that your operating environment can store.

147

148

Chapter 2. The FREQ Procedure

For any single crosstabulation table requested, PROC FREQ builds the entire table in
memory, regardless of whether the table has zero cell counts. Thus, if the numeric
variables A, B, and C each have 10 levels, PROC FREQ requires 2520 bytes to store
the variable levels for the table request A*B*C, as follows:
3 variables * 10 levels/variable * 84 bytes/level

In addition, PROC FREQ requires 8000 bytes to store the table cell frequencies
1000 cells * 8 bytes/cell

even though there may be only 10 observations.

When the variables have many levels or when there are many multiway tables, your
computer may not have enough memory to construct the tables. If PROC FREQ
runs out of memory while constructing tables, it stops collecting levels for the variable with the most levels and returns the memory that is used by that variable. The
procedure then builds the tables that do not contain the disabled variables.
If there is not enough memory for your table request and if increasing the available
memory is impractical, you can reduce the number of multiway tables or variable
levels. If you are not using the CMH or AGREE option in the TABLES statement
to compute statistics across strata, reduce the number of multiway tables by using
PROC SORT to sort the data set by one or more of the variables or by using the
DATA step to create an index for the variables. Then remove the sorted or indexed
variables from the TABLES statement and include a BY statement that uses these
variables. You can also reduce memory requirements by using a FORMAT statement
in the PROC FREQ step to reduce the number of levels. Additionally, reducing the
formatted variable lengths reduces the amount of memory that is needed to store
the variable levels. For more information on using formats, see the Grouping with
Formats section on page 99.

Output Data Sets

PROC FREQ produces two types of output data sets that you can use with other
statistical and reporting procedures. These data sets are produced as follows:
Specifying a TABLES statement with an OUT= option creates an output data
set that contains frequency or crosstabulation table counts and percentages.
Specifying an OUTPUT statement creates an output data set that contains
statistics.
PROC FREQ does not display the output data sets. Use PROC PRINT, PROC
REPORT, or any other SAS reporting tool to display an output data set.

Output Data Sets

Contents of the TABLES Statement Output Data Set
The OUT= option in the TABLES statement creates an output data set that contains
one observation for each combination of the variable values (or table cell) in the last
table request. By default, each observation contains the frequency and percentage for
the table cell. When the input data set contains missing values, the output data set
also contains an observation with the frequency of missing values. The output data
set includes the following variables:
BY variables
table request variables, such as A, B, C, and D in the table request A*B*C*D
COUNT, a variable containing the cell frequency
PERCENT, a variable containing the cell percentage
If you specify the OUTEXPECT and OUTPCT options in the TABLES statement,
the output data set also contains expected frequencies and row, column, and table
percentages, respectively. The additional variables are
EXPECTED, a variable containing the expected frequency
PCT TABL, a variable containing the percentage of two-way table frequency,
for n-way tables where n > 2
PCT ROW, a variable containing the percentage of row frequency
PCT COL, a variable containing the percentage of column frequency
If you specify the OUTCUM option in the TABLES statement, the output data set
also contains cumulative frequencies and cumulative percentages for one-way tables.
The additional variables are
CUM FREQ, a variable containing the cumulative frequency
CUM PCT, a variable containing the cumulative percentage
The OUTCUM option has no effect for two-way or multiway tables.
When you submit the following statements
proc freq;
tables A A*B / out=D;
run;

the output data set D contains frequencies and percentages for the last table request,
A*B. If A has two levels (1 and 2), B has three levels (1,2, and 3), and no table cell
count is zero or missing, the output data set D includes six observations, one for each
combination of A and B. The rst observation corresponds to A=1 and B=1; the second observation corresponds to A=1 and B=2; and so on. The data set includes the
variables COUNT and PERCENT. The value of COUNT is the number of observations with the given combination of A and B values. The value of PERCENT is the
percent of the total number of observations having that A and B combination.
When PROC FREQ combines different variable values into the same formatted level,
the output data set contains the smallest internal value for the formatted level. For

149

150

Chapter 2. The FREQ Procedure

example, suppose a variable X has the values 1.1., 1.4, 1.7, 2.1, and 2.3. When you
submit the statement
format X 1.;

in a PROC FREQ step, the formatted levels listed in the frequency table for X are 1
and 2. If you create an output data set with the frequency counts, the internal values
of X are 1.1 and 1.7. To report the internal values of X when you display the output
data set, use a format of 3.1 with X.

Contents of the OUTPUT Statement Output Data Set

The OUTPUT statement creates a SAS data set containing the statistics that PROC
FREQ computes for the last table request. You specify which statistics to store in the
output data set. There is an observation with the specied statistics for each stratum
or two-way table. If PROC FREQ computes summary statistics for a stratied table,
the output data set also contains a summary observation with those statistics.
The OUTPUT data set can include the following variables.
BY variables
variables that identify the stratum, such as A and B in the table request
A*B*C*D
variables that contain the specied statistics
The output data set also includes variables with the p-values and degrees of freedom,
asymptotic standard error (ASE), or condence limits when PROC FREQ computes
these values for a specied statistic.
The variable names for the specied statistics in the output data set are the names
of the options enclosed in underscores. PROC FREQ forms variable names for the
corresponding p-values, degrees of freedom, or condence limits by combining the
name of the option with the appropriate prex from the following list:

DF
E
L
U
E0
Z
P
P2
PL
PR
XP
XP2
XPL
XPR
XPT
XL
XR

degrees of freedom
asymptotic standard error (ASE)
lower condence limit
upper condence limit
ASE under the null hypothesis
standardized value
p-value
two-sided p-value
left-sided p-value
right-sided p-value
exact p-value
exact two-sided p-value
exact left-sided p-value
exact right-sided p-value
exact point probability
exact lower condence limit
exact upper condence limit

Displayed Output
For example, variable names created for the Pearson chi-square, its degrees of freedom, its p-values are PCHI , DF PCHI, and P PCHI, respectively.
If the length of the prex plus the statistic option exceeds eight characters, PROC
FREQ truncates the option so that the name of the new variable is eight characters
long.

Displayed Output
Number of Variable Levels Table
If you specify the NLEVELS option in the PROC FREQ statement, PROC FREQ
displays the Number of Variable Levels table. This table provides the number of
levels for all variables named in the TABLES statements. PROC FREQ determines
the variable levels from the formatted variable values. See Grouping with Formats
for details. The Number of Variable Levels table contains the following information:
Variable name
Levels, which is the total number of levels of the variable
Number of Nonmissing Levels, if there are missing levels for any of the variables
Number of Missing Levels, if there are missing levels for any of the variables

One-Way Frequency Tables

PROC FREQ displays one-way frequency tables for all one-way table requests in the
TABLES statements, unless you specify the NOPRINT option in the PROC statement
or the NOPRINT option in the TABLES statement. For a one-way table showing
the frequency distribution of a single variable, PROC FREQ displays the following
information:
the name of the variable and its values
Frequency counts, giving the number of observations that have each value
specied Test Frequency counts, if you specify the CHISQ and TESTF= options to request a chi-square goodness-of-t test for specied frequencies
Percent, giving the percentage of the total number of observations with that
value. (The NOPERCENT option suppresses this information.)
specied Test Percents, if you specify the CHISQ and TESTP= options
to request a chi-square goodness-of-t test for specied percents. (The
NOPERCENT option suppresses this information.)
Cumulative Frequency counts, giving the sum of the frequency counts of that
value and all other values listed above it in the table. The last cumulative frequency is the total number of nonmissing observations. (The NOCUM option
suppresses this information.)
Cumulative Percent values, giving the percentage of the total number of observations with that value and all others previously listed in the table. (The
NOCUM or the NOPERCENT option suppresses this information.)

151

152

Chapter 2. The FREQ Procedure

The one-way table also displays the Frequency Missing, or the number of observations with missing values.

Statistics for One-Way Frequency Tables

For one-way tables, two statistical options are available in the TABLES statement.
The CHISQ option provides a chi-square goodness-of-t test, and the BINOMIAL
option provides binomial proportion statistics. PROC FREQ displays the following
information, unless you specify the NOPRINT option in the PROC statement:
If you specify the CHISQ option for a one-way table, PROC FREQ provides a
chi-square goodness-of-t test, displaying the Chi-Square statistic, the degrees
of freedom (DF), and the probability value (Pr > ChiSq). If you specify the
CHISQ option in the EXACT statement, PROC FREQ also displays the exact probability value for this test. If you specify the POINT option with the
CHISQ option in the EXACT statement, PROC FREQ displays the exact point
probability for the test statistic.
If you specify the BINOMIAL option for a one-way table, PROC FREQ displays the estimate of the binomial Proportion, which is the proportion of observations in the rst class listed in the one-way table. PROC FREQ also displays
the asymptotic standard error (ASE) and the asymptotic and exact condence
limits for this estimate. For the binomial proportion test, PROC FREQ displays
the asymptotic standard error under the null hypothesis (ASE Under H0), the
standardized test statistic (Z), and the one-sided and two-sided probability values. If you specify the BINOMIAL option in the EXACT statement, PROC
FREQ also displays the exact one-sided and two-sided probability values for
this test. If you specify the POINT option with the BINOMIAL option in the
EXACT statement, PROC FREQ displays the exact point probability for the
test.

Multiway Tables
PROC FREQ displays all multiway table requests in the TABLES statements, unless
you specify the NOPRINT option in the PROC statement or the NOPRINT option in
the TABLES statement.
For two-way to multiway crosstabulation tables, the values of the last variable in the
table request form the table columns. The values of the next-to-last variable form the
rows. Each level (or combination of levels) of the other variables forms one stratum.
There are three ways to display multiway tables in PROC FREQ. By default, PROC
FREQ displays multiway tables as separate two-way crosstabulation tables for each
stratum of the multiway table. Also by default, PROC FREQ displays these twoway crosstabulation tables in table cell format. Alternatively, if you specify the
CROSSLIST option, PROC FREQ displays the two-way crosstabulation tables in
ODS column format. If you specify the LIST option, PROC FREQ displays multiway tables in list format.

Displayed Output
Crosstabulation Tables

By default, PROC FREQ displays two-way crosstabulation tables in table cell format.
The row variable values are listed down the side of the table, the column variable
values are listed across the top of the table, and each row and column variable level
combination forms a table cell.
Each cell of a crosstabulation table may contain the following information:
Frequency, giving the number of observations that have the indicated values of
the two variables. (The NOFREQ option suppresses this information.)
the Expected cell frequency under the hypothesis of independence, if you specify the EXPECTED option
the Deviation of the cell frequency from the expected value, if you specify the
DEVIATION option
Cell Chi-Square, which is the cells contribution to the total chi-square statistic,
if you specify the CELLCHI2 option
Tot Pct, or the cells percentage of the total frequency, for n-way tables when
n > 2, if you specify the TOTPCT option
Percent, the cells percentage of the total frequency. (The NOPERCENT option
suppresses this information.)
Row Pct, or the row percentage, the cells percentage of the total frequency
count for that cells row. (The NOROW option suppresses this information.)
Col Pct, or column percentage, the cells percentage of the total frequency
count for that cells column. (The NOCOL option suppresses this information.)
Cumulative Col%, or cumulative column percent, if you specify the CUMCOL
option
The table also displays the Frequency Missing, or the number of observations with
missing values.
CROSSLIST Tables

If you specify the CROSSLIST option, PROC FREQ displays two-way crosstabulation tables with ODS column format. Using column format, a CROSSLIST table
provides the same information (frequencies, percentages, and other statistics) as the
default crosstabulation table with cell format. But unlike the default crosstabulation
table, a CROSSLIST table has a table denition that you can customize with PROC
TEMPLATE. For more information, refer to the chapter titled The TEMPLATE
Procedure in the SAS Output Delivery System Users Guide.
In the CROSSLIST table format, the rows of the display correspond to the crosstabulation table cells, and the columns of the display correspond to descriptive statistics
such as frequencies and percentages. Each table cell is identied by the values of
its TABLES row and column variable levels, with all column variable levels listed
within each row variable level. The CROSSLIST table also provides row totals, column totals, and overall table totals.

153

154

Chapter 2. The FREQ Procedure

For a crosstabulation table in the CROSSLIST format, PROC FREQ displays the
following information:
the row variable name and values
the column variable name and values
Frequency, giving the number of observations that have the indicated values of
the two variables. (The NOFREQ option suppresses this information.)
the Expected cell frequency under the hypothesis of independence, if you specify the EXPECTED option
the Deviation of the cell frequency from the expected value, if you specify the
DEVIATION option
Cell Chi-Square, which is the cells contribution to the total chi-square statistic,
if you specify the CELLCHI2 option
Total Percent, or the cells percentage of the total frequency, for n-way tables
when n > 2, if you specify the TOTPCT option
Percent, the cells percentage of the total frequency. (The NOPERCENT option
suppresses this information.)
Row Percent, the cells percentage of the total frequency count for that cells
row. (The NOROW option suppresses this information.)
Column Percent, the cells percentage of the total frequency count for that cells
column. (The NOCOL option suppresses this information.)
The table also displays the Frequency Missing, or the number of observations with
missing values.

LIST Tables

If you specify the LIST option in the TABLES statement, PROC FREQ displays
multiway tables in a list format rather than as crosstabulation tables. The LIST option
displays the entire multiway table in one table, instead of displaying a separate twoway table for each stratum. The LIST option is not available when you also request
statistical options. Unlike the default crosstabulation output, the LIST output does
not display row percentages, column percentages, and optional information such as
expected frequencies and cell chi-squares.
For a multiway table in list format, PROC FREQ displays the following information:
the variable names and values
Frequency counts, giving the number of observations with the indicated combination of variable values
Percent, the cells percentage of the total number of observations.
NOPERCENT option suppresses this information.)

(The

Cumulative Frequency counts, giving the sum of the frequency counts of that
cell and all other cells listed above it in the table. The last cumulative frequency is the total number of nonmissing observations. (The NOCUM option
suppresses this information.)

Displayed Output
Cumulative Percent values, giving the percentage of the total number of observations for that cell and all others previously listed in the table. (The NOCUM
or the NOPERCENT option suppresses this information.)
The table also displays the Frequency Missing, or the number of observations with
missing values.

Statistics for Multiway Tables

PROC FREQ computes statistical tests and measures for crosstabulation tables, depending on which statements and options you specify. You can suppress the display
of all these results by specifying the NOPRINT option in the PROC statement. With
any of the following information, PROC FREQ also displays the Sample Size and the
Frequency Missing.
If you specify the SCOROUT option, PROC FREQ displays the Row Scores
and Column Scores that it uses for statistical computations. The Row Scores table displays the row variable values and the Score corresponding to each value.
The Column Scores table displays the column variable values and the corresponding Scores. PROC FREQ also identies the score type used to compute
the row and column scores. You can specify the score type with the SCORES=
option in the TABLES statement.
If you specify the CHISQ option, PROC FREQ displays the following statistics
for each two-way table: Pearson Chi-Square, Likelihood-Ratio Chi-Square,
Continuity-Adjusted Chi-Square (for 2 2 tables), Mantel-Haenszel ChiSquare, the Phi Coefcient, the Contingency Coefcient, and Cramers V . For
each test statistic, PROC FREQ also displays the degrees of freedom (DF) and
the probability value (Prob).
If you specify the CHISQ option for 2 2 tables, PROC FREQ also displays
Fishers exact test. The test output includes the cell (1,1) frequency (F), the
exact left-sided and right-sided probability values, the table probability (P),
and the exact two-sided probability value.
If you specify the FISHER option in the TABLES statement (or, equivalently,
the FISHER option in the EXACT statement), PROC FREQ displays Fishers
exact test for tables larger than 2 2. The test output includes the table probability (P) and the probability value. In addition, PROC FREQ displays the
CHISQ output listed earlier, even if you do not also specify the CHISQ option.
If you specify the PCHI, LRCHI, or MHCHI option in the EXACT statement,
PROC FREQ also displays the corresponding exact test: Pearson Chi-Square,
Likelihood-Ratio Chi-Square, or Mantel-Haenszel Chi-Square, respectively.
The test output includes the test statistic, the degrees of freedom (DF), and
the asymptotic and exact probability values. If you also specify the POINT
option in the EXACT statement, PROC FREQ displays the point probability
for each exact test requested. If you specify the CHISQ option in the EXACT
statement, PROC FREQ displays exact probability values for all three of these
chi-square tests.

155

156

Chapter 2. The FREQ Procedure

If you specify the MEASURES option, PROC FREQ displays the following
statistics and their asymptotic standard errors (ASE) for each two-way
table: Gamma, Kendalls Tau-b, Stuarts Tau-c, Somers D(C|R), Somers
D(R|C), Pearson Correlation, Spearman Correlation, Lambda Asymmetric
(C|R), Lambda Asymmetric (R|C), Lambda Symmetric, Uncertainty
Coefcient (C|R), Uncertainty Coefcient (R|C), and Uncertainty
Coefcient Symmetric. If you specify the CL option, PROC FREQ also
displays condence limits for these measures.
If you specify the PLCORR option, PROC FREQ displays the tetrachoric correlation for 2 2 tables or the polychoric correlation for larger tables. In addition, PROC FREQ displays the MEASURES output listed earlier, even if you
do not also specify the MEASURES option.
If you specify the option GAMMA, KENTB, STUTC, SMDCR, SMDRC,
PCORR, or SCORR in the TEST statement, PROC FREQ displays asymptotic
tests for Gamma, Kendalls Tau-b, Stuarts Tau-c, Somers D(C|R), Somers
D(R|C), the Pearson Correlation, or the Spearman Correlation, respectively.
If you specify the MEASURES option in the TEST statement, PROC FREQ
displays all these asymptotic tests. The test output includes the statistic, its
asymptotic standard error (ASE), Condence Limits, the ASE under the null
hypothesis H0, the standardized test statistic (Z), and the one-sided and twosided probability values.
If you specify the PCORR or SCORR option in the EXACT statement, PROC
FREQ displays asymptotic and exact tests for the Pearson Correlation or the
Spearman Correlation, respectively. The test output includes the correlation, its
asymptotic standard error (ASE), Condence Limits, the ASE under the null
hypothesis H0, the standardized test statistic (Z), and the asymptotic and exact
one-sided and two-sided probability values. If you also specify the POINT
option in the EXACT statement, PROC FREQ displays the point probability
for each exact test requested.
If you specify the RISKDIFF option for 2 2 tables, PROC FREQ displays
the Column 1 and Column 2 Risk Estimates. For each column, PROC FREQ
displays Row 1 Risk, Row 2 Risk, Total Risk, and Risk Difference, together
with their asymptotic standard errors (ASE), Asymptotic Condence Limits,
and Exact Condence Limits. Exact condence limits are not available for the
risk difference.
If you specify the MEASURES option or the RELRISK option for 2 2 tables,
PROC FREQ displays Estimates of the Relative Risk for Case-Control and
Cohort studies, together with their Condence Limits. These measures are
also known as the Odds Ratio and the Column 1 and 2 Relative Risks. If you
specify the OR option in the EXACT statement, PROC FREQ also displays
Exact Condence Limits for the Odds Ratio.
If you specify the TREND option, PROC FREQ displays the CochranArmitage Trend Test for tables that are 2 C or R 2. For this test, PROC
FREQ gives the Statistic (Z) and the one-sided and two-sided probability values. If you specify the TREND option in the EXACT statement, PROC FREQ
also displays the exact one-sided and two-sided probability values for this test.

Displayed Output
If you specify the POINT option with the TREND option in the EXACT statement, PROC FREQ displays the exact point probability for the test statistic.
If you specify the JT option, PROC FREQ displays the Jonckheere-Terpstra
Test, showing the Statistic (JT), the standardized test statistic (Z), and the onesided and two-sided probability values. If you specify the JT option in the
EXACT statement, PROC FREQ also displays the exact one-sided and twosided probability values for this test. If you specify the POINT option with
the JT option in the EXACT statement, PROC FREQ displays the exact point
probability for the test statistic.
If you specify the AGREE option and the PRINTKWT option, PROC FREQ
displays the Kappa Coefcient Weights for square tables greater than 2 2.
If you specify the AGREE option, for two-way tables PROC FREQ displays
McNemars Test and the Simple Kappa Coefcient for 2 2 tables. For square
tables larger than 2 2, PROC FREQ displays Bowkers Test of Symmetry,
the Simple Kappa Coefcient, and the Weighted Kappa Coefcient. For
McNemars Test and Bowkers Test of Symmetry, PROC FREQ displays the
Statistic (S), the degrees of freedom (DF), and the probability value (Pr > S).
If you specify the MCNEM option in the EXACT statement, PROC FREQ
also displays the exact probability value for McNemars test. If you specify
the POINT option with the MCNEM option in the EXACT statement, PROC
FREQ displays the exact point probability for the test statistic. For the simple and weighted kappa coefcients, PROC FREQ displays the kappa values,
asymptotic standard errors (ASE), and Condence Limits.
If you specify the KAPPA or WTKAP option in the TEST statement, PROC
FREQ displays asymptotic tests for the simple kappa coefcient or the
weighted kappa coefcient, respectively. If you specify the AGREE option
in the TEST statement, PROC FREQ displays both these asymptotic tests. The
test output includes the kappa coefcient, its asymptotic standard error (ASE),
Condence Limits, the ASE under the null hypothesis H0, the standardized test
statistic (Z), and the one-sided and two-sided probability values.
If you specify the KAPPA or WTKAP option in the EXACT statement, PROC
FREQ displays asymptotic and exact tests for the simple kappa coefcient or
the weighted kappa coefcient, respectively. The test output includes the kappa
coefcient, its asymptotic standard error (ASE), Condence Limits, the ASE
under the null hypothesis H0, the standardized test statistic (Z), and the asymptotic and exact one-sided and two-sided probability values. If you specify the
POINT option in the EXACT statement, PROC FREQ displays the point probability for each exact test requested.
If you specify the MC option in the EXACT statement, PROC FREQ displays
Monte Carlo estimates for all exact p-values requested by statistic-options in
the EXACT statement. The Monte Carlo output includes the p-value Estimate,
its Condence Limits, the Number of Samples used to compute the Monte
Carlo estimate, and the Initial Seed for random number generation.
If you specify the AGREE option, for multiple strata PROC FREQ displays
Overall Simple and Weighted Kappa Coefcients, with their asymptotic standard errors (ASE) and Condence Limits. PROC FREQ also displays Tests for

157

158

Chapter 2. The FREQ Procedure

Equal Kappa Coefcients, giving the Chi-Squares, degrees of freedom (DF),
and probability values (Pr > ChiSq) for the Simple Kappa and Weighted Kappa.
For multiple strata of 2 2 tables, PROC FREQ displays Cochrans Q, giving
the Statistic (Q), the degrees of freedom (DF), and the probability value (Pr >
Q).
If you specify the CMH option, PROC FREQ displays Cochran-MantelHaenszel Statistics for the following three alternative hypotheses: Nonzero
Correlation, Row Mean Scores Differ (ANOVA Statistic), and General
Association. For each of these statistics, PROC FREQ gives the degrees of
freedom (DF) and the probability value (Prob). For 2 2 tables, PROC FREQ
also displays Estimates of the Common Relative Risk for Case-Control and
Cohort studies, together with their condence limits. These include both
Mantel-Haenszel and Logit stratum-adjusted estimates of the common Odds
Ratio, Column 1 Relative Risk, and Column 2 Relative Risk. Also for 2 2
tables, PROC FREQ displays the Breslow-Day Test for Homogeneity of the
Odds Ratios. For this test, PROC FREQ gives the Chi-Square, the degrees of
freedom (DF), and the probability value (Pr > ChiSq).
If you specify the CMH option in the TABLES statement and also specify the
COMOR option in the EXACT statement, PROC FREQ displays exact condence limits for the Common Odds Ratio for multiple strata of 2 2 tables.
PROC FREQ also displays the Exact Test of H0: Common Odds Ratio = 1.
The test output includes the Cell (1,1) Sum (S), Mean of S Under H0,
One-sided Pr <= S, and Point Pr = S. PROC FREQ also provides exact
two-sided probability values for the test, computed according to the following
three methods: 2 * One-sided, Sum of probabilities <= Point probability, and
Pr >= |S - Mean|.

ODS Table Names

PROC FREQ assigns a name to each table it creates. You can use these names to
reference the table when using the Output Delivery System (ODS) to select tables
and create output data sets. For more information on ODS, see Chapter 14, Using
the Output Delivery System. (SAS/STAT Users Guide)
Table 2.11 lists the ODS table names together with their descriptions and the options
required to produce the tables. Note that the ALL option in the TABLES statement
invokes the CHISQ, MEASURES, and CMH options.

ODS Table Names

159

Table 2.11. ODS Tables Produced in PROC FREQ

ODS Table Name

BinomialProp
BinomialPropTest
BreslowDayTest
CMH
ChiSq
CochransQ
ColScores
CommonOddsRatioCL
CommonOddsRatioTest
CommonRelRisks
CrossList
CrossTabFreqs
EqualKappaTest
EqualKappaTests
FishersExact

FishersExactMC
Gamma
GammaTest
JTTest
JTTestMC
KappaStatistics

KappaWeights
List
LRChiSq
LRChiSqMC
McNemarsTest
Measures
MHChiSq
MHChiSqMC
NLevels
OddsRatioCL
OneWayChiSq

Description
Binomial proportion
Binomial proportion test
Breslow-Day test
Cochran-Mantel-Haenszel
statistics
Chi-square tests
Cochrans Q
Column scores
Exact condence limits
for the common odds ratio
Common odds ratio exact test
Common relative risks
Column format
crosstabulation table
Crosstabulation table
Test for equal simple kappas
Tests for equal kappas
Fishers exact test

Monte Carlo estimates

for Fishers exact test
Gamma
Gamma test
Jonckheere-Terpstra test
Monte Carlo estimates
for the JT exact test
Kappa statistics

Kappa weights
List format multiway table
Likelihood-ratio
chi-square exact test
Monte Carlo exact test for
likelihood-ratio chi-square
McNemars test
Measures of association
Mantel-Haenszel
chi-square exact test
Monte Carlo exact test for
Mantel-Haenszel chi-square
Number of variable levels
Exact condence limits
for the odds ratio
One-way chi-square test

Statement
TABLES
TABLES
TABLES
TABLES

Option
BINOMIAL (one-way tables)
BINOMIAL (one-way tables)
CMH (h 2 2 tables)
CMH

TABLES
TABLES
TABLES
EXACT

CHISQ
AGREE (h 2 2 tables)
SCOROUT
COMOR

EXACT
TABLES
TABLES
TABLES
TABLES
TABLES
EXACT
or TABLES
or TABLES
EXACT

COMOR
CMH (h 2 2 tables)
CROSSLIST
(n-way table request, n > 1)
(n-way table request, n > 1)
AGREE (h 2 2 tables)
AGREE (hrr tables, r > 2)
FISHER
FISHER or EXACT
CHISQ (2 2 tables)
FISHER / MC

TEST
TEST
TABLES
EXACT

GAMMA
GAMMA
JT
JT / MC

TABLES

TABLES
TABLES
EXACT

AGREE
(r r tables, r > 2, and
no TEST or EXACT KAPPA)
AGREE and PRINTKWT
LIST
LRCHI

EXACT

LRCHI / MC

TABLES
TABLES
EXACT

AGREE (2 2 tables)
MEASURES
MHCHI

EXACT

MHCHI / MC

PROC
EXACT

NLEVELS
OR

TABLES

CHISQ (one-way tables)

160

Chapter 2. The FREQ Procedure

Table 2.11. (continued)

ODS Table Name

OneWayChiSqMC
OneWayFreqs
OverallKappa
OverallKappas
PearsonChiSq
PearsonChiSqMC
PearsonCorr
PearsonCorrMC

Description
Monte Carlo exact test
for one-way chi-square
One-way frequencies
Overall simple kappa
Overall kappa coefcients
Pearson chi-square
exact test
Monte Carlo exact test
for Pearson chi-square
Pearson correlation

for Spearman correlation
Spearman correlation test

SpearmanCorrTest
SymmetryTest
TauB
TauBTest
TauC
TauCTest
TrendTest

Test of symmetry
Kendalls tau-b
Kendalls tau-b test
Stuarts tau-c
Stuarts tau-c test
Cochran-Armitage test
for trend

Statement
EXACT

Option
CHISQ / MC (one-way tables)

PROC
or TABLES
TABLES
TABLES
EXACT

(with no TABLES stmt)

(one-way table request)
AGREE (h 2 2 tables)
AGREE (hrr tables, r > 2)
PCHI

EXACT

PCHI / MC

TEST
or EXACT
EXACT

PCORR
PCORR
PCORR / MC

TEST
or EXACT
TABLES
TABLES
TABLES
TABLES
TEST
or EXACT
EXACT

PCORR
PCORR
RELRISK or MEASURES
(2 2 tables)
RISKDIFF (2 2 tables)
RISKDIFF (2 2 tables)
SCOROUT
KAPPA
KAPPA
KAPPA / MC

TEST
or EXACT
TEST
TEST
TEST
TEST
TEST
or EXACT
EXACT

KAPPA
KAPPA
SMDCR
SMDCR
SMDRC
SMDRC
SCORR
SCORR
SCORR / MC

TEST
or EXACT
TABLES
TEST
TEST
TEST
TEST
TABLES

SCORR
SCORR
AGREE
KENTB
KENTB
STUTC
STUTC
TREND

Examples

161

Table 2.11. (continued)

ODS Table Name

TrendTestMC
WeightedKappa
WeightedKappaMC
WeightedKappaTest

Description
Monte Carlo exact test
for trend
Weighted kappa
Monte Carlo exact test
for weighted kappa
Weighted kappa test

Statement
EXACT

Option
TREND / MC

TEST
or EXACT
EXACT

WTKAP
WTKAP
WTKAP / MC

TEST
or EXACT

WTKAP
WTKAP

Examples
Example 2.1. Creating an Output Data Set with Table Cell
Frequencies
The eye and hair color of children from two different regions of Europe are recorded
in the data set Color. Instead of recording one observation per child, the data are
recorded as cell counts, where the variable Count contains the number of children
exhibiting each of the 15 eye and hair color combinations. The data set does not
include missing combinations.
data Color;
input Region Eyes $ Hair $ Count @@;
label Eyes =Eye Color
Hair =Hair Color
Region=Geographic Region;
datalines;
1 blue fair
23 1 blue red
7 1
1 blue dark
11 1 green fair
19 1
1 green medium 18 1 green dark
14 1
1 brown red
5 1 brown medium 41 1
1 brown black
3 2 blue fair
46 2
2 blue medium 44 2 blue dark
40 2
2 green fair
50 2 green red
31 2
2 green dark
23 2 brown fair
56 2
2 brown medium 53 2 brown dark
54 2
;

blue
green
brown
brown
blue
blue
green
brown
brown

medium
red
fair
dark
red
black
medium
red
black

24
7
34
40
21
6
37
42
13

The following statements read the Color data set and create an output data set containing the frequencies, percentages, and expected cell frequencies of the Eyes by
Hair two-way table. The TABLES statement requests three tables: Eyes and Hair
frequency tables and an Eyes by Hair crosstabulation table. The OUT= option
creates the FreqCnt data set, which contains the crosstabulation table frequencies.
The OUTEXPECT option outputs the expected cell frequencies to FreqCnt, and the
SPARSE option includes the zero cell counts. The WEIGHT statement species
that Count contains the observation weights. These statements create Output 2.1.1
through Output 2.1.3.

162

Chapter 2. The FREQ Procedure

proc freq data=Color;
weight Count;
tables Eyes Hair Eyes*Hair / out=FreqCnt outexpect sparse;
title1 Eye and Hair Color of European Children;
run;
proc print data=FreqCnt noobs;
title2 Output Data Set from PROC FREQ;
run;

Output 2.1.1 displays the two frequency tables produced, one showing the distribution of eye color, and one showing the distribution of hair color. By default, PROC
FREQ lists the variables values in alphabetical order. The Eyes*Hair specication
produces a crosstabulation table, shown in Output 2.1.2, with eye color dening the
table rows and hair color dening the table columns. A zero cell count for green eyes
and black hair indicates that this eye and hair color combination does not occur in the
data.
The output data set (Output 2.1.3) contains frequency counts and percentages for the
last table. The data set also includes an observation for the zero cell count (SPARSE)
and a variable with the expected cell frequency for each table cell (OUTEXPECT).

Output 2.1.1. Frequency Tables

Eye and Hair Color of European Children
The FREQ Procedure
Eye Color
Cumulative
Cumulative
Eyes
Frequency
Percent
Frequency
Percent
---------------------------------------------------------blue
222
29.13
222
29.13
brown
341
44.75
563
73.88
green
199
26.12
762
100.00

Hair Color
Cumulative
Cumulative
Hair
Frequency
Percent
Frequency
Percent
----------------------------------------------------------black
22
2.89
22
2.89
dark
182
23.88
204
26.77
fair
228
29.92
432
56.69
medium
217
28.48
649
85.17
red
113
14.83
762
100.00

Examples

Output 2.1.2. Crosstabulation Table

Eye and Hair Color of European Children
Table of Eyes by Hair
Eyes(Eye Color)

Hair(Hair Color)

Frequency|
Percent |
Row Pct |
Col Pct |black
|dark
|fair
|medium |red
| Total
---------+--------+--------+--------+--------+--------+
blue
|
6 |
51 |
69 |
68 |
28 |
222
|
0.79 |
6.69 |
9.06 |
8.92 |
3.67 | 29.13
|
2.70 | 22.97 | 31.08 | 30.63 | 12.61 |
| 27.27 | 28.02 | 30.26 | 31.34 | 24.78 |
---------+--------+--------+--------+--------+--------+
brown
|
16 |
94 |
90 |
94 |
47 |
341
|
2.10 | 12.34 | 11.81 | 12.34 |
6.17 | 44.75
|
4.69 | 27.57 | 26.39 | 27.57 | 13.78 |
| 72.73 | 51.65 | 39.47 | 43.32 | 41.59 |
---------+--------+--------+--------+--------+--------+
green
|
0 |
37 |
69 |
55 |
38 |
199
|
0.00 |
4.86 |
9.06 |
7.22 |
4.99 | 26.12
|
0.00 | 18.59 | 34.67 | 27.64 | 19.10 |
|
0.00 | 20.33 | 30.26 | 25.35 | 33.63 |
---------+--------+--------+--------+--------+--------+
Total
22
182
228
217
113
762
2.89
23.88
29.92
28.48
14.83
100.00

Output 2.1.3. OUT= Data Set

Output Data Set from PROC FREQ
Eyes

Hair

blue
blue
blue
blue
blue
brown
brown
brown
brown
brown
green
green
green
green
green

black
dark
fair
medium
red
black
dark
fair
medium
red
black
dark
fair
medium
red

COUNT
6
51
69
68
28
16
94
90
94
47
0
37
69
55
38

EXPECTED

PERCENT

6.409
53.024
66.425
63.220
32.921
9.845
81.446
102.031
97.109
50.568
5.745
47.530
59.543
56.671
29.510

0.7874
6.6929
9.0551
8.9239
3.6745
2.0997
12.3360
11.8110
12.3360
6.1680
0.0000
4.8556
9.0551
7.2178
4.9869

163

164

Chapter 2. The FREQ Procedure

Example 2.2. Computing Chi-Square Tests for One-Way

Frequency Tables
This example examines whether the childrens hair color (from Example 2.1 on page
161) has a specied multinomial distribution for the two regions. The hypothesized
distribution for hair color is 30% fair, 12% red, 30% medium, 25% dark, and 3%
black.
In order to test the hypothesis for each region, the data are rst sorted by Region.
Then the FREQ procedure uses a BY statement to produce a separate table for each
BY group (Region). The option ORDER=DATA orders the frequency table values
(hair color) by their order in the data set. The TABLES statement requests a frequency table for hair color, and the option NOCUM suppresses the display of the
cumulative frequencies and percentages. The TESTP= option species the hypothesized percentages for the chi-square test; the number of percentages specied equals
the number of table levels, and the percentages sum to 100. The following statements
produce Output 2.2.1.
proc sort data=Color;
by Region;
run;
proc freq data=Color order=data;
weight Count;
tables Hair / nocum testp=(30 12 30 25 3);
by Region;
title Hair Color of European Children;
run;

The frequency tables in Output 2.2.1 list the variable values (hair color) in the order
in which they appear in the data set. The Test Percent column lists the hypothesized
percentages for the chi-square test. Always check that you have ordered the TESTP=
percentages to correctly match the order of the variable levels.
PROC FREQ computes a chi-square statistic for each region. The chi-square statistic
is signicant at the 0.05 level for Region 2 (p=0.0003) but not for Region 1. This
indicates a signicant departure from the hypothesized percentages in Region 2.

Examples

Output 2.2.1. One-Way Frequency Table with BY Groups

Hair Color of European Children
----------------------------- Geographic Region=1 -----------------------------The FREQ Procedure
Hair Color
Test
Hair
Frequency
Percent
Percent
------------------------------------------fair
76
30.89
30.00
red
19
7.72
12.00
medium
83
33.74
30.00
dark
65
26.42
25.00
black
3
1.22
3.00

Chi-Square Test
for Specified Proportions
------------------------Chi-Square
7.7602
DF
4
Pr > ChiSq
0.1008
Sample Size = 246

Hair Color of European Children

----------------------------- Geographic Region=2 -----------------------------Hair Color
Test
Hair
Frequency
Percent
Percent
------------------------------------------fair
152
29.46
30.00
red
94
18.22
12.00
medium
134
25.97
30.00
dark
117
22.67
25.00
black
19
3.68
3.00

Chi-Square Test
for Specified Proportions
------------------------Chi-Square
21.3824
DF
4
Pr > ChiSq
0.0003
Sample Size = 516

165

166

Chapter 2. The FREQ Procedure

Example 2.3. Computing Binomial Proportions for One-Way

Frequency Tables
The binomial proportion is computed as the proportion of observations for the rst
level of the variable that you are studying. The following statements compute the
proportion of children with brown eyes (from the data set in Example 2.1 on page
161) and test this value against the hypothesis that the proportion is 50%. Also,
these statements test whether the proportion of children with fair hair is 28%.

proc freq data=Color order=freq;

weight Count;
tables Eyes / binomial alpha=.1;
tables Hair / binomial(p=.28);
title Hair and Eye Color of European Children;
run;

The rst TABLES statement produces a frequency table for eye color. The
BINOMIAL option computes the binomial proportion and condence limits, and it
tests the hypothesis that the proportion for the rst eye color level (brown) is 0.5. The
option ALPHA=.1 species that 90% condence limits should be computed. The
second TABLES statement creates a frequency table for hair color and computes the
binomial proportion and condence limits, but it tests that the proportion for the rst
hair color (fair) is 0.28. These statements produce Output 2.3.1 and Output 2.3.2.
The frequency table in Output 2.3.1 displays the variable values in order of descending frequency count. Since the rst variable level is brown, PROC FREQ computes
the binomial proportion of children with brown eyes. PROC FREQ also computes its
asymptotic standard error (ASE), and asymptotic and exact 90% condence limits.
If you do not specify the ALPHA= option, then PROC FREQ computes the default
95% condence limits.
Because the value of Z is less than zero, PROC FREQ computes a left-sided p-value
(0.0019). This small p-value supports the alternative hypothesis that the true value of
the proportion of children with brown eyes is less than 50%.
Output 2.3.2 displays the results from the second TABLES statement. PROC FREQ
computes the default 95% condence limits since the ALPHA= option is not specied. The value of Z is greater than zero, so PROC FREQ computes a right-sided
p-value (0.1188). This large p-value provides insufcient evidence to reject the null
hypothesis that the proportion of children with fair hair is 28%.

Examples

Output 2.3.1. Binomial Proportion for Eye Color

Hair and Eye Color of European Children
The FREQ Procedure
Eye Color
Cumulative
Cumulative
Eyes
Frequency
Percent
Frequency
Percent
---------------------------------------------------------brown
341
44.75
341
44.75
blue
222
29.13
563
73.88
green
199
26.12
762
100.00

Binomial Proportion
for Eyes = brown
-------------------------------Proportion
0.4475
ASE
0.0180
90% Lower Conf Limit
0.4179
90% Upper Conf Limit
0.4771
Exact Conf Limits
90% Lower Conf Limit
90% Upper Conf Limit

0.4174
0.4779

Test of H0: Proportion = 0.5

ASE under H0
Z
One-sided Pr < Z
Two-sided Pr > |Z|
Sample Size = 762

0.0181
-2.8981
0.0019
0.0038

167

168

Chapter 2. The FREQ Procedure

Output 2.3.2. Binomial Proportion for Hair Color

Hair and Eye Color of European Children
Hair Color
Cumulative
Cumulative
Hair
Frequency
Percent
Frequency
Percent
----------------------------------------------------------fair
228
29.92
228
29.92
medium
217
28.48
445
58.40
dark
182
23.88
627
82.28
red
113
14.83
740
97.11
black
22
2.89
762
100.00

Binomial Proportion
for Hair = fair
-------------------------------Proportion
0.2992
ASE
0.0166
95% Lower Conf Limit
0.2667
95% Upper Conf Limit
0.3317
Exact Conf Limits
95% Lower Conf Limit
95% Upper Conf Limit

0.2669
0.3331

Test of H0: Proportion = 0.28

ASE under H0
Z
One-sided Pr > Z
Two-sided Pr > |Z|
Sample Size = 762

0.0163
1.1812
0.1188
0.2375

Example 2.4. Analyzing a 2x2 Contingency Table

This example computes chi-square tests and Fishers exact test to compare the probability of coronary heart disease for two types of diet. It also estimates the relative
risks and computes exact condence limits for the odds ratio.
The data set FatComp contains hypothetical data for a case-control study of high
fat diet and the risk of coronary heart disease. The data are recorded as cell counts,
where the variable Count contains the frequencies for each exposure and response
combination. The data set is sorted in descending order by the variables Exposure
and Response, so that the rst cell of the 2 2 table contains the frequency of
positive exposure and positive response. The FORMAT procedure creates formats to
identify the type of exposure and response with character values.
proc format;
value ExpFmt 1=High Cholesterol Diet
0=Low Cholesterol Diet;
value RspFmt 1=Yes
0=No;
run;
data FatComp;
input Exposure Response Count;
label Response=Heart Disease;
datalines;
0 0 6
0 1 2
1 0 4
1 1 11
;
proc sort data=FatComp;
by descending Exposure descending Response;
run;

In the following statements, the TABLES statement creates a two-way table, and the
option ORDER=DATA orders the contingency table values by their order in the data
set. The CHISQ option produces several chi-square tests, while the RELRISK option
produces relative risk measures. The EXACT statement creates the exact Pearson chisquare test and exact condence limits for the odds ratio. These statements produce
Output 2.4.1 through Output 2.4.3.
proc freq data=FatComp order=data;
weight Count;
tables Exposure*Response / chisq relrisk;
exact pchi or;
format Exposure ExpFmt. Response RspFmt.;
title Case-Control Study of High Fat/Cholesterol Diet;
run;

169

170

Chapter 2. The FREQ Procedure

Output 2.4.1. Contingency Table

Case-Control Study of High Fat/Cholesterol Diet
The FREQ Procedure
Table of Exposure by Response
Exposure

Response(Heart Disease)

Frequency
|
Percent
|
Row Pct
|
Col Pct
|Yes
|No
| Total
-----------------+--------+--------+
High Cholesterol |
11 |
4 |
15
Diet
| 47.83 | 17.39 | 65.22
| 73.33 | 26.67 |
| 84.62 | 40.00 |
-----------------+--------+--------+
Low Cholesterol |
2 |
6 |
8
Diet
|
8.70 | 26.09 | 34.78
| 25.00 | 75.00 |
| 15.38 | 60.00 |
-----------------+--------+--------+
Total
13
10
23
56.52
43.48
100.00

The contingency table in Output 2.4.1 displays the variable values so that the rst
table cell contains the frequency for the rst cell in the data set, the frequency of
positive exposure and positive response.

Example 2.4. Analyzing a 2x2 Contingency Table

Output 2.4.2. Chi-Square Statistics

Case-Control Study of High Fat/Cholesterol Diet
Statistics for Table of Exposure by Response
Statistic
DF
Value
Prob
-----------------------------------------------------Chi-Square
1
4.9597
0.0259
Likelihood Ratio Chi-Square
1
5.0975
0.0240
Continuity Adj. Chi-Square
1
3.1879
0.0742
Mantel-Haenszel Chi-Square
1
4.7441
0.0294
Phi Coefficient
0.4644
Contingency Coefficient
0.4212
Cramers V
0.4644
WARNING: 50% of the cells have expected counts less than 5.
(Asymptotic) Chi-Square may not be a valid test.

Pearson Chi-Square Test

---------------------------------Chi-Square
4.9597
DF
1
Asymptotic Pr > ChiSq
0.0259
Exact
Pr >= ChiSq
0.0393

Fishers Exact Test

---------------------------------Cell (1,1) Frequency (F)
11
Left-sided Pr <= F
0.9967
Right-sided Pr >= F
0.0367
Table Probability (P)
Two-sided Pr <= P

0.0334
0.0393

Sample Size = 23

Output 2.4.2 displays the chi-square statistics. Since the expected counts in some
of the table cells are small, PROC FREQ gives a warning that the asymptotic chisquare tests may not be appropriate. In this case, the exact tests are appropriate.
The alternative hypothesis for this analysis states that coronary heart disease is more
likely to be associated with a high fat diet, so a one-sided test is desired. Fishers
exact right-sided test analyzes whether the probability of heart disease in the high fat
group exceeds the probability of heart disease in the low fat group; since this p-value
is small, the alternative hypothesis is supported.

171

172

Chapter 2. The FREQ Procedure

Output 2.4.3. Relative Risk

Case-Control Study of High Fat/Cholesterol Diet
Statistics for Table of Exposure by Response
Estimates of the Relative Risk (Row1/Row2)
Type of Study
Value
95% Confidence Limits
----------------------------------------------------------------Case-Control (Odds Ratio)
8.2500
1.1535
59.0029
Cohort (Col1 Risk)
2.9333
0.8502
10.1204
Cohort (Col2 Risk)
0.3556
0.1403
0.9009

Odds Ratio (Case-Control Study)

----------------------------------Odds Ratio
8.2500
Asymptotic Conf Limits
95% Lower Conf Limit
95% Upper Conf Limit
Exact Conf Limits
95% Lower Conf Limit
95% Upper Conf Limit

1.1535
59.0029

0.8677
105.5488

Sample Size = 23

The odds ratio, displayed in Output 2.4.3, provides an estimate of the relative risk
when an event is rare. This estimate indicates that the odds of heart disease is 8.25
times higher in the high fat diet group; however, the wide condence limits indicate
that this estimate has low precision.

Example 2.5. Creating an Output Data Set Containing ChiSquare Statistics

This example uses the Color data from Example 2.1 (page 161) to output the Pearson
chi-square and the likelihood-ratio chi-square statistics to a SAS data set. The following statements create a two-way table of eye color versus hair color.
proc freq data=Color order=data;
weight Count;
tables Eyes*Hair / chisq expected cellchi2 norow nocol;
output out=ChiSqData pchi lrchi n nmiss;
title Chi-Square Tests for 3 by 5 Table of Eye and Hair Color;
run;
proc print data=ChiSqData noobs;
title1 Chi-Square Statistics for Eye and Hair Color;
title2 Output Data Set from the FREQ Procedure;
run;

The CHISQ option produces chi-square tests, the EXPECTED option displays expected cell frequencies in the table, and the CELLCHI2 option displays the cell contribution to the chi-square. The NOROW and NOCOL options suppress the display
of row and column percents in the table.

Example 2.5. Creating an Output Data Set Containing Chi-Square Statistics

The OUTPUT statement creates the ChiSqData data set with eight variables: the N
option stores the number of nonmissing observations, the NMISS option stores the
number of missing observations, and the PCHI and LRCHI options store Pearson
and likelihood-ratio chi-square statistics, respectively, together with their degrees of
freedom and p-values.
The preceding statements produce Output 2.5.1 and Output 2.5.2.
Output 2.5.1. Contingency Table
Chi-Square Tests for 3 by 5 Table of Eye and Hair Color
The FREQ Procedure
Table of Eyes by Hair
Eyes(Eye Color)

Hair(Hair Color)

Frequency
|
Expected
|
Cell Chi-Square|
Percent
|fair
|red
|medium |dark
|black
| Total
---------------+--------+--------+--------+--------+--------+
blue
|
69 |
28 |
68 |
51 |
6 |
222
| 66.425 | 32.921 | 63.22 | 53.024 | 6.4094 |
| 0.0998 | 0.7357 | 0.3613 | 0.0772 | 0.0262 |
|
9.06 |
3.67 |
8.92 |
6.69 |
0.79 | 29.13
---------------+--------+--------+--------+--------+--------+
green
|
69 |
38 |
55 |
37 |
0 |
199
| 59.543 | 29.51 | 56.671 | 47.53 | 5.7454 |
| 1.5019 | 2.4422 | 0.0492 | 2.3329 | 5.7454 |
|
9.06 |
4.99 |
7.22 |
4.86 |
0.00 | 26.12
---------------+--------+--------+--------+--------+--------+
brown
|
90 |
47 |
94 |
94 |
16 |
341
| 102.03 | 50.568 | 97.109 | 81.446 | 9.8451 |
| 1.4187 | 0.2518 | 0.0995 | 1.935 | 3.8478 |
| 11.81 |
6.17 | 12.34 | 12.34 |
2.10 | 44.75
---------------+--------+--------+--------+--------+--------+
Total
228
113
217
182
22
762
29.92
14.83
28.48
23.88
2.89
100.00

The contingency table in Output 2.5.1 displays eye and hair color in the order in
which they appear in the Color data set. The Pearson chi-square statistic in Output
2.5.2 provides evidence of an association between eye and hair color (p=0.0073).
The cell chi-square values show that most of the association is due to more greeneyed children with fair or red hair and fewer with dark or black hair. The opposite
occurs with the brown-eyed children.

173

174

Chapter 2. The FREQ Procedure

Output 2.5.2. Chi-Square Statistics

Chi-Square Tests for 3 by 5 Table of Eye and Hair Color
Statistics for Table of Eyes by Hair
Statistic
DF
Value
Prob
-----------------------------------------------------Chi-Square
8
20.9248
0.0073
Likelihood Ratio Chi-Square
8
25.9733
0.0011
Mantel-Haenszel Chi-Square
1
3.7838
0.0518
Phi Coefficient
0.1657
Contingency Coefficient
0.1635
Cramers V
0.1172
Sample Size = 762

Output 2.5.3. Output Data Set

Chi-Square Statistics for Eye and Hair Color
Output Data Set from the FREQ Procedure
N

NMISS

_PCHI_

DF_PCHI

762

20.9248

P_PCHI

_LRCHI_

.007349898

25.9733

DF_LRCHI
8

P_LRCHI
.001061424

The OUT= data set is displayed in Output 2.5.3. It contains one observation with the
sample size, the number of missing values, and the chi-square statistics and corresponding degrees of freedom and p-values as in Output 2.5.2.

Example 2.6. Computing Cochran-Mantel-Haenszel

Statistics for a Stratied Table
The data set Migraine contains hypothetical data for a clinical trial of migraine treatment. Subjects of both genders receive either a new drug therapy or a placebo. Their
response to treatment is coded as Better or Same. The data are recorded as cell
counts, and the number of subjects for each treatment and response combination is
recorded in the variable Count.
data Migraine;
input Gender $ Treatment $ Response $ Count
datalines;
female Active Better 16
female Active Same
female Placebo Better 5
female Placebo Same
male
Active Better 12
male
Active Same
male
Placebo Better 7
male
Placebo Same
;

@@;
11
20
16
19

Example 2.6. Computing Cochran-Mantel-Haenszel

Statistics for a Stratied Table
The following statements create a three-way table stratied by Gender, where
Treatment forms the rows and Response forms the columns. The CMH option
produces the Cochran-Mantel-Haenszel statistics. For this stratied 2 2 table, estimates of the common relative risk and the Breslow-Day test for homogeneity of the
odds ratios are also displayed. The NOPRINT option suppresses the display of the
contingency tables. These statements produce Output 2.6.1 through Output 2.6.3.
proc freq data=Migraine;
weight Count;
tables Gender*Treatment*Response / cmh noprint;
title Clinical Trial for Treatment of Migraine Headaches;
run;

Output 2.6.1. Cochran-Mantel-Haenszel Statistics

Clinical Trial for Treatment of Migraine Headaches
The FREQ Procedure
Summary Statistics for Treatment by Response
Controlling for Gender
Cochran-Mantel-Haenszel Statistics (Based on Table Scores)
Statistic
Alternative Hypothesis
DF
Value
Prob
--------------------------------------------------------------1
Nonzero Correlation
1
8.3052
0.0040
2
Row Mean Scores Differ
1
8.3052
0.0040
3
General Association
1
8.3052
0.0040
Total Sample Size = 106

For a stratied 2 2 table, the three CMH statistics displayed in Output 2.6.1 test
the same hypothesis. The signicant p-value (0.004) indicates that the association
between treatment and response remains strong after adjusting for gender.

175

176

Chapter 2. The FREQ Procedure

Output 2.6.2. CMH Option: Relative Risks

Clinical Trial for Treatment of Migraine Headaches
Summary Statistics for Treatment by Response
Controlling for Gender
Estimates of the Common Relative Risk (Row1/Row2)
Type of Study
Method
Value
95% Confidence Limits
------------------------------------------------------------------------Case-Control
Mantel-Haenszel
3.3132
1.4456
7.5934
(Odds Ratio)
Logit
3.2941
1.4182
7.6515
Cohort
(Col1 Risk)

Mantel-Haenszel
Logit

2.1636
2.1059

1.2336
1.1951

3.7948
3.7108

Cohort
(Col2 Risk)

Mantel-Haenszel
Logit

0.6420
0.6613

0.4705
0.4852

0.8761
0.9013

Total Sample Size = 106

The CMH option also produces a table of relative risks, as shown in Output 2.6.2.
Because this is a prospective study, the relative risk estimate assesses the effectiveness
of the new drug; the Cohort (Col1 Risk) values are the appropriate estimates for the
rst column, or the risk of improvement. The probability of migraine improvement
with the new drug is just over two times the probability of improvement with the
placebo.
Output 2.6.3. CMH Option: Breslow-Day Test
Clinical Trial for Treatment of Migraine Headaches
Summary Statistics for Treatment by Response
Controlling for Gender
Breslow-Day Test for
Homogeneity of the Odds Ratios
-----------------------------Chi-Square
1.4929
DF
1
Pr > ChiSq
0.2218
Total Sample Size = 106

The large p-value for the Breslow-Day test (0.2218) in Output 2.6.3 indicates no
signicant gender difference in the odds ratios.

Example 2.7. Computing the Cochran-Armitage Trend Test

The data set Pain contains hypothetical data for a clinical trial of a drug therapy
to control pain. The clinical trial investigates whether adverse responses increase
with larger drug doses. Subjects receive either a placebo or one of four drug doses.
An adverse response is recorded as Adverse=Yes; otherwise, it is recorded as
Adverse=No. The number of subjects for each drug dose and response combination is contained in the variable Count.
data pain;
input Dose Adverse $ Count @@;
datalines;
0 No 26
0 Yes 6
1 No 26
1 Yes 7
2 No 23
2 Yes 9
3 No 18
3 Yes 14
4 No 9
4 Yes 23
;

The TABLES statement in the following program produces a two-way table. The
MEASURES option produces measures of association, and the CL option produces
condence limits for these measures. The TREND option tests for a trend across the
ordinal values of the Dose variable with the Cochran-Armitage test. The EXACT
statement produces exact p-values for this test, and the MAXTIME= option terminates the exact computations if they do not complete within 60 seconds. The TEST
statement computes an asymptotic test for Somers D(C|R). These statements produce Output 2.7.1 through Output 2.7.3.
proc freq data=Pain;
weight Count;
tables Dose*Adverse / trend measures cl;
test smdcr;
exact trend / maxtime=60;
title1 Clinical Trial for Treatment of Pain;
run;

177

178

Chapter 2. The FREQ Procedure

Output 2.7.1. Contingency Table

Clinical Trial for Treatment of Pain
The FREQ Procedure
Table of Dose by Adverse
Dose

Adverse

Frequency|
Percent |
Row Pct |
Col Pct |No
|Yes
| Total
---------+--------+--------+
0 |
26 |
6 |
32
| 16.15 |
3.73 | 19.88
| 81.25 | 18.75 |
| 25.49 | 10.17 |
---------+--------+--------+
1 |
26 |
7 |
33
| 16.15 |
4.35 | 20.50
| 78.79 | 21.21 |
| 25.49 | 11.86 |
---------+--------+--------+
2 |
23 |
9 |
32
| 14.29 |
5.59 | 19.88
| 71.88 | 28.13 |
| 22.55 | 15.25 |
---------+--------+--------+
3 |
18 |
14 |
32
| 11.18 |
8.70 | 19.88
| 56.25 | 43.75 |
| 17.65 | 23.73 |
---------+--------+--------+
4 |
9 |
23 |
32
|
5.59 | 14.29 | 19.88
| 28.13 | 71.88 |
|
8.82 | 38.98 |
---------+--------+--------+
Total
102
59
161
63.35
36.65
100.00

The Row Pct values in Output 2.7.1 show the expected increasing trend in the
proportion of adverse effects due to increasing dosage (from 18.75% to 71.88%).

Example 2.7. Computing the Cochran-Armitage Trend Test

Output 2.7.2. Measures of Association

Clinical Trial for Treatment of Pain
Statistics for Table of Dose by Adverse
95%
Statistic
Value
ASE
Confidence Limits
---------------------------------------------------------------------------Gamma
0.5313
0.0935
0.3480
0.7146
Kendalls Tau-b
0.3373
0.0642
0.2114
0.4631
Stuarts Tau-c
0.4111
0.0798
0.2547
0.5675
Somers D C|R
Somers D R|C

0.2569
0.4427

0.0499
0.0837

0.1592
0.2786

0.3547
0.6068

Pearson Correlation
Spearman Correlation

0.3776
0.3771

0.0714
0.0718

0.2378
0.2363

0.5175
0.5178

Lambda Asymmetric C|R

Lambda Asymmetric R|C
Lambda Symmetric

0.2373
0.1250
0.1604

0.0837
0.0662
0.0621

0.0732
0.0000
0.0388

0.4014
0.2547
0.2821

Uncertainty Coefficient C|R

Uncertainty Coefficient R|C
Uncertainty Coefficient Symmetric

0.1261
0.0515
0.0731

0.0467
0.0191
0.0271

0.0346
0.0140
0.0199

0.2175
0.0890
0.1262

0.0499
5.1511
<.0001
<.0001

Sample Size = 161

Output 2.7.2 displays the measures of association produced by the MEASURES

option. Somers D(C|R) measures the association treating the column variable
(Adverse) as the response and the row variable (Dose) as a predictor. Because the
asymptotic 95% condence limits do not contain zero, this indicates a strong positive association. Similarly, the Pearson and Spearman correlation coefcients show
evidence of a strong positive association, as hypothesized.

179

180

Chapter 2. The FREQ Procedure

Output 2.7.3. Trend Test

Clinical Trial for Treatment of Pain
Statistics for Table of Dose by Adverse
Cochran-Armitage Trend Test
------------------------------Statistic (Z)
-4.7918
Asymptotic Test
One-sided Pr < Z
Two-sided Pr > |Z|
Exact Test
One-sided Pr <= Z
Two-sided Pr >= |Z|

<.0001
<.0001

7.237E-07
1.324E-06

Sample Size = 161

The Cochran-Armitage test (Output 2.7.3) supports the trend hypothesis. The small
left-sided p-values for the Cochran-Armitage test indicate that the probability of the
Column 1 level (Adverse=No) decreases as Dose increases or, equivalently, that
the probability of the Column 2 level (Adverse=Yes) increases as Dose increases.
The two-sided p-value tests against either an increasing or decreasing alternative.
This is an appropriate hypothesis when you want to determine whether the drug has
progressive effects on the probability of adverse effects but the direction is unknown.

Example 2.8. Computing Friedmans Chi-Square Statistic

Friedmans test is a nonparametric test for treatment differences in a randomized
complete block design. Each block of the design may be a subject or a homogeneous
group of subjects. If blocks are groups of subjects, the number of subjects in each
block must equal the number of treatments. Treatments are randomly assigned to
subjects within each block. If there is one subject per block, then the subjects are repeatedly measured once under each treatment. The order of treatments is randomized
for each subject.
In this setting, Friedmans test is identical to the ANOVA (row means scores) CMH
statistic when the analysis uses rank scores (SCORES=RANK). The three-way table
uses subject (or subject group) as the stratifying variable, treatment as the row variable, and response as the column variable. PROC FREQ handles ties by assigning
midranks to tied response values. If there are multiple subjects per treatment in each
block, the ANOVA CMH statistic is a generalization of Friedmans test.
The data set Hypnosis contains data from a study investigating whether hypnosis has
the same effect on skin potential (measured in millivolts) for four emotions (Lehmann
1975, p. 264). Eight subjects are asked to display fear, joy, sadness, and calmness
under hypnosis. The data are recorded as one observation per subject for each emotion.

Example 2.8. Computing Friedmans Chi-Square Statistic

data Hypnosis;
length Emotion $ 10;
input Subject Emotion
datalines;
1 fear 23.1 1 joy 22.7
2 fear 57.6 2 joy 53.2
3 fear 10.5 3 joy 9.7
4 fear 23.6 4 joy 19.6
5 fear 11.9 5 joy 13.8
6 fear 54.6 6 joy 47.1
7 fear 21.0 7 joy 13.6
8 fear 20.3 8 joy 23.6
;

$ SkinResponse @@;
1
2
3
4
5
6
7
8

sadness
sadness
sadness
sadness
sadness
sadness
sadness
sadness

22.5
53.7
10.8
21.1
13.7
39.2
13.7
16.3

1
2
3
4
5
6
7
8

calmness
calmness
calmness
calmness
calmness
calmness
calmness
calmness

22.6
53.1
8.3
21.6
13.3
37.0
14.8
14.8

In the following statements, the TABLES statement creates a three-way table stratied by Subject and a two-way table; the variables Emotion and SkinResponse
form the rows and columns of each table. The CMH2 option produces the rst two
Cochran-Mantel-Haenszel statistics, the option SCORES=RANK species that rank
scores are used to compute these statistics, and the NOPRINT option suppresses the
contingency tables. These statements produce Output 2.8.1 and Output 2.8.2.
proc freq data=Hypnosis;
tables Subject*Emotion*SkinResponse
/ cmh2 scores=rank noprint;
run;

Output 2.8.1. CMH Statistics: Stratifying by Subject

The FREQ Procedure
Summary Statistics for Emotion by SkinResponse
Controlling for Subject
Cochran-Mantel-Haenszel Statistics (Based on Rank Scores)
Statistic
Alternative Hypothesis
DF
Value
Prob
--------------------------------------------------------------1
Nonzero Correlation
1
0.2400
0.6242
2
Row Mean Scores Differ
3
6.4500
0.0917
Total Sample Size = 32

Because the CMH statistics in Output 2.8.1 are based on rank scores, the Row Mean
Scores Differ statistic is identical to Friedmans chi-square (Q = 6.45). The p-value
of 0.0917 indicates that differences in skin potential response for different emotions
are signicant at the 10% level but not at the 5% level.

181

182

Chapter 2. The FREQ Procedure

Output 2.8.2. CMH Statistics: No Stratication

The FREQ Procedure
Summary Statistics for Emotion by SkinResponse
Cochran-Mantel-Haenszel Statistics (Based on Rank Scores)
Statistic
Alternative Hypothesis
DF
Value
Prob
--------------------------------------------------------------1
Nonzero Correlation
1
0.0001
0.9933
2
Row Mean Scores Differ
3
0.5678
0.9038
Total Sample Size = 32

When you do not stratify by subject, the Row Mean Scores Differ CMH statistic is
identical to a Kruskal-Wallis test and is not signicant (p=0.9038 in Output 2.8.2).
Thus, adjusting for subject is critical to reducing the background variation due to
subject differences.

Example 2.9. Testing Marginal Homogeneity with

Cochrans Q
When a binary response is measured several times or under different conditions,
Cochrans Q tests that the marginal probability of a positive response is unchanged
across the times or conditions. When there are more than two response categories,
you can use the CATMOD procedure to t a repeated-measures model.
The data set Drugs contains data for a study of three drugs to treat a chronic disease
(Agresti 1990). Forty-six subjects receive drugs A, B, and C. The response to each
drug is either favorable (F) or unfavorable (U).
proc format;
value $ResponseFmt F=Favorable
U=Unfavorable;
data drugs;
input Drug_A $
datalines;
F F F 6
U F F
F F U 16
U F U
F U F 2
U U F
F U U 4
U U U
;

Drug_B $ Drug_C $ Count @@;

2
4
6
6

Example 2.9. Testing Marginal Homogeneity with

Cochrans Q
The following statements create one-way frequency tables of the responses to each
drug. The AGREE option produces Cochrans Q and other measures of agreement
for the three-way table. These statements produce Output 2.9.1 through Output 2.9.3.
proc freq data=Drugs;
weight Count;
tables Drug_A Drug_B Drug_C / nocum;
tables Drug_A*Drug_B*Drug_C / agree noprint;
format Drug_A Drug_B Drug_C $ResponseFmt.;
title Study of Three Drug Treatments for a Chronic Disease;
run;

Output 2.9.1. One-Way Frequency Tables

Study of Three Drug Treatments for a Chronic Disease
The FREQ Procedure
Drug_A
Frequency
Percent
-----------------------------------Favorable
28
60.87
Unfavorable
18
39.13

Drug_B
Frequency
Percent
-----------------------------------Favorable
28
60.87
Unfavorable
18
39.13

Drug_C
Frequency
Percent
-----------------------------------Favorable
16
34.78
Unfavorable
30
65.22

The one-way frequency tables in Output 2.9.1 provide the marginal response for each
drug. For drugs A and B, 61% of the subjects reported a favorable response while
35% of the subjects reported a favorable response to drug C.

183

184

Chapter 2. The FREQ Procedure

Output 2.9.2. Measures of Agreement

Study of Three Drug Treatments for a Chronic Disease
Statistics for Table 1 of Drug_B by Drug_C
Controlling for Drug_A=Favorable
McNemars Test
-----------------------Statistic (S)
10.8889
DF
1
Pr > S
0.0010

Simple Kappa Coefficient

-------------------------------Kappa
-0.0328
ASE
0.1167
95% Lower Conf Limit
-0.2615
95% Upper Conf Limit
0.1960
Sample Size = 28

Statistics for Table 2 of Drug_B by Drug_C

Controlling for Drug_A=Unfavorable
McNemars Test
----------------------Statistic (S)
0.4000
DF
1
Pr > S
0.5271

Simple Kappa Coefficient

-------------------------------Kappa
-0.1538
ASE
0.2230
95% Lower Conf Limit
-0.5909
95% Upper Conf Limit
0.2832
Sample Size = 18

Study of Three Drug Treatments for a Chronic Disease

Summary Statistics for Drug_B by Drug_C
Controlling for Drug_A
Overall Kappa Coefficient
-------------------------------Kappa
-0.0588
ASE
0.1034
95% Lower Conf Limit
-0.2615
95% Upper Conf Limit
0.1439

Test for Equal Kappa

Coefficients
-------------------Chi-Square
0.2314
DF
1
Pr > ChiSq
0.6305

Total Sample Size = 46

References
McNemars test (Output 2.9.2) shows strong discordance between drugs B and C
when the response to drug A is favorable. The small negative value of the simple
kappa indicates no agreement between drug B response and drug C response.
Output 2.9.3. Cochrans Q
Study of Three Drug Treatments for a Chronic Disease
Summary Statistics for Drug_B by Drug_C
Controlling for Drug_A
Cochrans Q, for Drug_A
by Drug_B by Drug_C
----------------------Statistic (Q)
8.4706
DF
2
Pr > Q
0.0145
Total Sample Size = 46

Cochrans Q is statistically signicant (p=0.0144 in Output 2.9.3), which leads to

rejection of the hypothesis that the probability of favorable response is the same for
the three drugs.

References
Agresti, A. (1990), Categorical Data Analysis, New York: John Wiley & Sons, Inc.
Agresti, A. (1992), A Survey of Exact Inference for Contingency Tables, Statistical
Science, 7(1), 131177.
Agresti, A. (1996), An Introduction to Categorical Data Analysis, New York: John
Wiley & Sons, Inc.
Agresti, A., Mehta, C.R. and Patel, N.R. (1990), Exact Inference for Contingency
Tables with Ordered Categories, Journal of the American Statistical
Association, 85, 453458.
Agresti, A., Wackerly, D., and Boyett, J.M. (1979), Exact Conditional Tests
for Cross-Classications: Approximation of Attained Signicance Levels,
Psychometrika, 44, 7583.
Birch, M.W. (1965), The Detection of Partial Association, II: The General Case,
Journal of the Royal Statistical Society, B, 27, 111124.
Bishop, Y., Fienberg, S.E., and Holland, P.W. (1975), Discrete Multivariate Analysis:
Theory and Practice, Cambridge, MA: MIT Press.
Bowker, A.H. (1948), Bowkers Test for Symmetry, Journal of the American
Statistical Association, 43, 572574.
Breslow, N.E. (1996), Statistics in Epidemiology: The Case-Control Study, Journal
of the American Statistical Association, 91, 1426.

185

186

Chapter 2. The FREQ Procedure

Breslow, N.E. and Day, N.E. (1980), Statistical Methods in Cancer Research, Volume
I: The Analysis of Case-Control Studies, IARC Scientic Publications, No. 32,
Lyon, France: International Agency for Research on Cancer.
Breslow, N.E. and Day, N.E. (1987), Statistical Methods in Cancer Research, Volume
II: The Design and Analysis of Cohort Studies, IARC Scientic Publications, No.
82, Lyon, France: International Agency for Research on Cancer.
Bross, I.D.J. (1958), How to Use Ridit Analysis, Biometrics, 14, 1838.
Brown, M.B. and Benedetti, J.K. (1977), Sampling Behavior of Tests for
Correlation in Two-Way Contingency Tables, Journal of the American
Statistical Association, 72, 309315.
Cicchetti, D.V. and Allison, T. (1971), A New Procedure for Assessing Reliability
of Scoring EEG Sleep Recordings, American Journal of EEG Technology, 11,
101109.
Cochran, W.G. (1950), The Comparison of Percentages in Matched Samples,
Biometrika, 37, 256266.
Cochran, W.G. (1954), Some Methods for Strengthening the Common 2 Tests,
Biometrics, 10, 417451.
Collett, D. (1991), Modelling Binary Data, London: Chapman & Hall.
Cohen, J. (1960), A Coefcient of Agreement for Nominal Scales, Educational
and Psychological Measurement, 20, 3746.
Drasgow, F. (1986), Polychoric and Polyserial Correlations in Encyclopedia of
Statistical Sciences, vol. 7, ed. S. Kotz and N. L. Johnson, New York: John
Wiley & Sons, Inc., 6874.
Fienberg, S.E. (1980), The Analysis of Cross-Classied Data, Second Edition,
Cambridge, MA: MIT Press.
Fleiss, J.L. (1981), Statistical Methods for Rates and Proportions, Second Edition,
New York: John Wiley & Sons, Inc.
Fleiss, J.L. and Cohen, J. (1973), The Equivalence of Weighted Kappa and the
Intraclass Correlation Coefcient as Measures of Reliability, Educational and
Psychological Measurement, 33, 613619.
Fleiss, J.L., Cohen, J., and Everitt, B.S. (1969), Large-Sample Standard Errors of
Kappa and Weighted Kappa, Psychological Bulletin, 72, 323327.
Freeman, G.H. and Halton, J.H. (1951), Note on an Exact Treatment of Contingency,
Goodness of Fit and Other Problems of Signicance, Biometrika, 38, 141149.
Gail, M. and Mantel, N. (1977), Counting the Number of r c Contingency
Tables with Fixed Margins, Journal of the American Statistical Association, 72,
859862.
Gart, J.J. (1971), The Comparison of Proportions: A Review of Signicance
Tests, Condence Intervals and Adjustments for Stratication, Review of the
International Statistical Institute, 39(2), 148169.

References
Goodman, L.A. and Kruskal, W.H. (1979), Measures of Association for Cross
Classication, New York: Springer-Verlag.
Greenland, S. and Robins, J.M. (1985), Estimators of the Mantel-Haenszel Variance
Consistent in Both Sparse Data and Large-Strata Limiting Models, Biometrics,
42, 311323.
Haldane, J.B.S. (1955), The Estimation and Signicance of the Logarithm of a Ratio
of Frequencies, Annals of Human Genetics, 20, 309314.
Hollander, M. and Wolfe, D.A. (1973), Nonparametric Statistical Methods, New
York: John Wiley & Sons, Inc.
Jones, M.P., OGorman, T.W., Lemka, J.H., and Woolson, R.F. (1989), A Monte
Carlo Investigation of Homogeneity Tests of the Odds Ratio Under Various
Sample Size Congurations, Biometrics, 45, 171181.
Kendall, M. (1955), Rank Correlation Methods, Second Edition, London: Charles
Grifn and Co.
Kendall, M. and Stuart, A. (1979), The Advanced Theory of Statistics, vol. 2, New
York: Macmillan Publishing Company, Inc.
Kleinbaum, D.G., Kupper, L.L., and Morgenstern, H. (1982), Epidemiologic
Research: Principles and Quantitative Methods, Research Methods Series, New
York: Van Nostrand Reinhold.
Landis, R.J., Heyman, E.R., and Koch, G.G. (1978), Average Partial Association in
Three-way Contingency Tables: A Review and Discussion of Alternative Tests,
International Statistical Review, 46, 237254.
Leemis, L.M. and Trivedi, K.S. (1996), A Comparison of Approximate Interval
Estimators for the Bernoulli Parameter, The American Statistician, 50(1),
6368.
Lehmann, E.L. (1975), Nonparametrics: Statistical Methods Based on Ranks, San
Francisco: Holden-Day, Inc.
Liebetrau, A.M. (1983), Measures of Association, Quantitative Application in the
Social Sciences, vol. 32, Beverly Hills: Sage Publications, Inc.
Mack, G.A. and Skillings, J.H. (1980), A Friedman-Type Rank Test for Main Effects
in a Two-Factor ANOVA, Journal of the American Statistical Association, 75,
947951.
Mantel, N. (1963), Chi-square Tests with One Degree of Freedom: Extensions of the
Mantel-Haenszel Procedure, Journal of the American Statistical Association,
58, 690700.
Mantel, N. and Haenszel, W. (1959), Statistical Aspects of the Analysis of Data from
Retrospective Studies of Disease, Journal of the National Cancer Institute, 22,
719748.
Margolin, B.H. (1988), Test for Trend in Proportions, in Encyclopedia of Statistical
Sciences, vol. 9, ed. S. Kotz and N.L. Johnson, New York: John Wiley & Sons,
Inc., 334336.

187

188

Chapter 2. The FREQ Procedure

McNemar, Q. (1947), Note on the Sampling Error of the Difference between
Correlated Proportions or Percentages, Psychometrika, 12, 153157.
Mehta, C.R. and Patel, N.R. (1983), A Network Algorithm for Performing Fishers
Exact Test in r c Contingency Tables, Journal of the American Statistical
Association, 78, 427434.
Mehta, C.R., Patel, N.R., and Gray, R. (1985), On Computing an Exact Condence
Interval for the Common Odds Ratio in Several 2 2 Contingency Tables,
Journal of the American Statistical Association, 80, 969973.
Mehta, C.R., Patel, N.R., and Senchaudhuri, P. (1991), Exact Stratied Linear Rank
Tests for Binary Data, Computing Science and Statistics: Proceedings of the
23rd Symposium on the Interface, ed. E.M. Keramidas, 200207.
Mehta, C.R., Patel, N.R., and Tsiatis, A.A. (1984), Exact Signicance Testing to
Establish Treatment Equivalence with Ordered Categorical Data, Biometrics,
40, 819825.
Narayanan, A. and Watts, D. (1996), Exact Methods in the NPAR1WAY Procedure,
in Proceedings of the Twenty-First Annual SAS Users Group International
Conference, Cary, NC: SAS Institute Inc., 12901294.
Olsson, U. (1979), Maximum Likelihood Estimation of the Polychoric Correlation
Coefcient, Psychometrika, 12, 443460.
Pirie, W. (1983), Jonckheere Tests for Ordered Alternatives, in Encyclopedia of
Statistical Sciences, vol. 4, ed. S. Kotz and N.L. Johnson, New York: John Wiley
& Sons, Inc., 315318.
Radlow, R. and Alf, E.F. (1975), An Alternate Multinomial Assessment of the
Accuracy of the Chi-Square Test of Goodness of Fit, Journal of the American
Statistical Association, 70, 811813.
Robins, J.M., Breslow, N., and Greenland, S. (1986), Estimators of the MantelHaenszel Variance Consistent in Both Sparse Data and Large-Strata Limiting
Models, Biometrics, 42, 311323.
Snedecor, G.W. and Cochran, W.G. (1989), Statistical Methods, Eighth Edition,
Ames, IA: Iowa State University Press.
Somers, R.H. (1962), A New Asymmetric Measure of Association for Ordinal
Variables, American Sociological Review, 27, 799811.
Stokes, M.E., Davis, C.S., and Koch, G.G. (1995), Categorical Data Analysis Using
the SAS System, Cary, NC: SAS Institute Inc.
Tarone, R.E. (1985), On Heterogeneity Tests Based on Efcient Scores,
Biometrika, 72, 1, 9195.
Theil, H. (1972), Statistical Decomposition Analysis, Amsterdam: North-Holland
Publishing Company.
Thomas, D.G. (1971), Algorithm AS-36. Exact Condence Limits for the Odds
Ratio in a 2 2 Table, Applied Statistics, 20, 105110.

References
Valz, P.D. and Thompson, M.E. (1994), Exact Inference for Kendalls S and
Spearmans Rho with Extensions to Fishers Exact Test in r c Contingency
Tables, Journal of Computational and Graphical Statistics, 3(4), 459472.
van Elteren, P.H. (1960), On the Combination of Independent Two-Sample Tests of
Wilcoxon, Bulletin of the International Statistical Institute, 37, 351361.
Vollset, S.E., Hirji, K.F., and Elashoff, R.M. (1991), Fast Computation of Exact
Condence Limits for the Common Odds Ratio in a Series of 2 2 Tables,
Journal of the American Statistical Association, 86, 404409.
Woolf, B. (1955), On Estimating the Relationship between Blood Group and
Disease, Annals of Human Genetics, 19, 251253.

189

190

Chapter 2. The FREQ Procedure

Chapter 3

The UNIVARIATE Procedure

Chapter Contents
OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
GETTING STARTED . . . . . . . . .
Capabilities of PROC UNIVARIATE
Summarizing a Data Distribution . .
Exploring a Data Distribution . . . .
Modeling a Data Distribution . . . .

.
.
.
.
.

196
196
196
197
200

SYNTAX . . . . . . . . . . . . . .
PROC UNIVARIATE Statement
BY Statement . . . . . . . . . .
CLASS Statement . . . . . . . .
FREQ Statement . . . . . . . . .
HISTOGRAM Statement . . . .
ID Statement . . . . . . . . . . .
INSET Statement . . . . . . . .
OUTPUT Statement . . . . . . .
PROBPLOT Statement . . . . .
QQPLOT Statement . . . . . . .
VAR Statement . . . . . . . . .
WEIGHT Statement . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.

202
203
209
210
212
212
230
230
237
241
253
266
267

DETAILS . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Missing Values . . . . . . . . . . . . . . . . . . . . . . . . .
Rounding . . . . . . . . . . . . . . . . . . . . . . . . . . .
Descriptive Statistics . . . . . . . . . . . . . . . . . . . . .
Calculating the Mode . . . . . . . . . . . . . . . . . . . . .
Calculating Percentiles . . . . . . . . . . . . . . . . . . . .
Tests for Location . . . . . . . . . . . . . . . . . . . . . . .
Condence Limits for Parameters of the Normal Distribution
Robust Estimators . . . . . . . . . . . . . . . . . . . . . . .
Creating Line Printer Plots . . . . . . . . . . . . . . . . . .
Creating High-Resolution Graphics . . . . . . . . . . . . . .
Using the CLASS Statement to Create Comparative Plots . .
Positioning the Inset . . . . . . . . . . . . . . . . . . . . . .
Formulas for Fitted Continuous Distributions . . . . . . . . .
Goodness-of-Fit Tests . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

268
268
269
269
272
273
275
277
278
281
284
284
285
288
292

.
.
.
.
.
.
.
.
.
.
.
.
.

192

Chapter 3. The UNIVARIATE Procedure

Kernel Density Estimates . . . . . . . . . . . . . . . . . .
Construction of Quantile-Quantile and Probability Plots . .
Interpretation of Quantile-Quantile and Probability Plots . .
Distributions for Probability and Q-Q Plots . . . . . . . . .
Estimating Shape Parameters Using Q-Q Plots . . . . . . .
Estimating Location and Scale Parameters Using Q-Q Plots
Estimating Percentiles Using Q-Q Plots . . . . . . . . . . .
Input Data Sets . . . . . . . . . . . . . . . . . . . . . . . .
OUT= Output Data Set in the OUTPUT Statement . . . . .
OUTHISTOGRAM= Output Data Set . . . . . . . . . . .
Tables for Summary Statistics . . . . . . . . . . . . . . . .
ODS Table Names . . . . . . . . . . . . . . . . . . . . . .
ODS Tables for Fitted Distributions . . . . . . . . . . . . .
Computational Resources . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.

297
298
299
300
303
304
305
305
306
308
308
309
310
311

EXAMPLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Example 3.1. Computing Descriptive Statistics for Multiple Variables . . . . 312
Example 3.2. Calculating Modes . . . . . . . . . . . . . . . . . . . . . . . 314
Example 3.3. Identifying Extreme Observations and Extreme Values . . . . 315
Example 3.4. Creating a Frequency Table . . . . . . . . . . . . . . . . . . . 317
Example 3.5. Creating Plots for Line Printer Output . . . . . . . . . . . . . 319
Example 3.6. Analyzing a Data Set With a FREQ Variable . . . . . . . . . . 322
Example 3.7. Saving Summary Statistics in an OUT= Output Data Set . . . 323
Example 3.8. Saving Percentiles in an Output Data Set . . . . . . . . . . . . 325
Example 3.9. Computing Condence Limits for the Mean, Standard Deviation,
and Variance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
Example 3.10. Computing Condence Limits for Quantiles and Percentiles . 328
Example 3.11. Computing Robust Estimates . . . . . . . . . . . . . . . . . 329
Example 3.12. Testing for Location . . . . . . . . . . . . . . . . . . . . . . 331
Example 3.13. Performing a Sign Test Using Paired Data . . . . . . . . . . 332
Example 3.14. Creating a Histogram . . . . . . . . . . . . . . . . . . . . . 333
Example 3.15. Creating a One-Way Comparative Histogram . . . . . . . . . 334
Example 3.16. Creating a Two-Way Comparative Histogram . . . . . . . . . 337
Example 3.17. Adding Insets with Descriptive Statistics . . . . . . . . . . . 338
Example 3.18. Binning a Histogram . . . . . . . . . . . . . . . . . . . . . 340
Example 3.19. Adding a Normal Curve to a Histogram . . . . . . . . . . . . 343
Example 3.20. Adding Fitted Normal Curves to a Comparative Histogram . 345
Example 3.21. Fitting a Beta Curve . . . . . . . . . . . . . . . . . . . . . . 346
Example 3.22. Fitting Lognormal, Weibull, and Gamma Curves . . . . . . . 348
Example 3.23. Computing Kernel Density Estimates . . . . . . . . . . . . . 352
Example 3.24. Fitting a Three-Parameter Lognormal Curve . . . . . . . . . 354
Example 3.25. Annotating a Folded Normal Curve . . . . . . . . . . . . . . 355
Example 3.26. Creating Lognormal Probability Plots . . . . . . . . . . . . . 360
Example 3.27. Creating a Histogram to Display Lognormal Fit . . . . . . . 363
Example 3.28. Creating a Normal Quantile Plot . . . . . . . . . . . . . . . 365
Example 3.29. Adding a Distribution Reference Line . . . . . . . . . . . . 366
Example 3.30. Interpreting a Normal Quantile Plot . . . . . . . . . . . . . . 368
Example 3.31. Estimating Three Parameters from Lognormal Quantile Plots 369

Chapter 3. The UNIVARIATE Procedure

Example 3.32. Estimating Percentiles from Lognormal Quantile Plots . . . . 372
Example 3.33. Estimating Parameters from Lognormal Quantile Plots . . . . 373
Example 3.34. Comparing Weibull Quantile Plots . . . . . . . . . . . . . . 375
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

193

194

Chapter 3. The UNIVARIATE Procedure

Chapter 3

The UNIVARIATE Procedure

Overview
The UNIVARIATE procedure provides the following:
descriptive statistics based on moments (including skewness and kurtosis),
quantiles or percentiles (such as the median), frequency tables, and extreme
values
histograms and comparative histograms. Optionally, these can be tted with
probability density curves for various distributions and with kernel density estimates.
quantile-quantile plots (Q-Q plots) and probability plots. These plots facilitate
the comparison of a data distribution with various theoretical distributions.
goodness-of-t tests for a variety of distributions including the normal
the ability to inset summary statistics on plots produced on a graphics device
the ability to analyze data sets with a frequency variable
the ability to create output data sets containing summary statistics, histogram
intervals, and parameters of tted curves
You can use the PROC UNIVARIATE statement, together with the VAR statement,
to compute summary statistics. See Getting Started on page 196 for introductory
examples. In addition, you can use the following statements to request plots:
the HISTOGRAM statement for creating histograms, the QQPLOT statement
for creating Q-Q plots, and the PROBPLOT statement for creating probability
plots
the CLASS statement together with the HISTOGRAM, QQPLOT, and
PROBPLOT statement for creating comparative histograms, Q-Q plots, and
probability plots
the INSET statement with any of the plot statements for enhancing the plot
with an inset table of summary statistics. The INSET statement is applicable
only to plots produced on graphics devices.

196

Chapter 3. The UNIVARIATE Procedure

Getting Started
The following examples demonstrate how you can use the UNIVARIATE procedure
to analyze the distributions of variables through the use of descriptive statistical measures and graphical displays, such as histograms.

Capabilities of PROC UNIVARIATE

The UNIVARIATE procedure provides a variety of descriptive measures, highresolution graphical displays, and statistical methods, which you can use to summarize, visualize, analyze, and model the statistical distributions of numeric variables.
These tools are appropriate for a broad range of tasks and applications:
Exploring the distributions of the variables in a data set is an important preliminary step in data analysis, data warehousing, and data mining. With the
UNIVARIATE procedure you can use tables and graphical displays, such as
histograms and nonparametric density estimates, to nd key features of distributions, identify outliers and extreme observations, determine the need for data
transformations, and compare distributions.
Modeling the distributions of data and validating distributional assumptions are
basic steps in statistical analysis. You can use the UNIVARIATE procedure to
t parametric distributions (beta, exponential, gamma, lognormal, normal, and
Weibull) and to compute probabilities and percentiles from these models. You
can assess goodness of t with hypothesis tests and with graphical displays
such as probability plots and quantile-quantile plots. You can also use the
UNIVARIATE procedure to validate distributional assumptions for other types
of statistical analysis. When standard assumptions are not met, you can use the
UNIVARIATE procedure to perform nonparametric tests and compute robust
estimates of location and scale.
Summarizing the distribution of the data is often helpful for creating effective
statistical reports and presentations. You can use the UNIVARIATE procedure
to create tables of summary measures, such as means and percentiles, together
with graphical displays, such as histograms and comparative histograms, which
facilitate the interpretation of the report.
The following examples illustrate a few of the tasks that you can carry out with the
UNIVARIATE procedure.

Summarizing a Data Distribution

Figure 3.1 shows a table of basic summary measures and a table of extreme observations for the loan-to-value ratios of 5,840 home mortgages. The ratios are saved
as values of the variable LoanToValueRatio in a data set named HomeLoans. The
following statements request a univariate analysis:

Exploring a Data Distribution

ods select BasicMeasures ExtremeObs;
proc univariate data=homeloans;
var LoanToValueRatio;
run;

The ODS SELECT statement restricts the default output to the tables for basic statistical measures and extreme observations.

Variable:

The UNIVARIATE Procedure

LoanToValueRatio (Loan to Value Ratio)
Basic Statistical Measures

Location
Mean
Median
Mode

Variability

0.292512
0.248050
0.250000

Std Deviation
Variance
Range
Interquartile Range

0.16476
0.02715
1.24780
0.16419

Extreme Observations
-------Lowest------

-----Highest-----

Value

Obs

Value

Obs

0.0651786
0.0690157
0.0699755
0.0702412
0.0704787

1
3
59
84
4

1.13976
1.14209
1.14286
1.17090
1.31298

5776
5791
5801
5799
5811

Figure 3.1. Basic Measures and Extreme Observations

The tables in Figure 3.1 show, in particular, that the average ratio is 0.2925 and the
minimum and maximum ratios are 0.06518 and 1.1398, respectively.

Exploring a Data Distribution

Figure 3.2 shows a histogram of the loan-to-value ratios. The histogram reveals features of the ratio distribution, such as its skewness and the peak at 0.175, which are
not evident from the tables in the previous example. The following statements create
the histogram:
title Home Loan Analysis;
proc univariate data=homeloans noprint;
histogram LoanToValueRatio / cfill=ltgray;
inset n = Number of Homes / position=ne;
run;

The NOPRINT option suppresses the display of summary statistics. The INSET
statement inserts the total number of analyzed home loans in the northeast corner of
the plot.

197

198

Chapter 3. The UNIVARIATE Procedure

Figure 3.2. Histogram for Loan-to-Value Ratio

The data set HomeLoans contains a variable named LoanType that classies the
loans into two types: Gold and Platinum. It is useful to compare the distributions
of LoanToValueRatio for the two types. The following statements request quantiles
for each distribution and a comparative histogram, which are shown in Figure 3.3 and
Figure 3.4.
title Comparison of Loan Types;
ods select Quantiles MyHist;
proc univariate data=HomeLoans;
var LoanToValueRatio;
class LoanType;
histogram LoanToValueRatio / cfill=ltgray
kernel(color=black)
name=MyHist;
inset n=Number of Homes median=Median Ratio (5.3) / position=ne;
label LoanType = Type of Loan;
run;

The ODS SELECT statement restricts the default output to the tables of quantiles.
The CLASS statement species LoanType as a classication variable for the quantile computations and comparative histogram. The KERNEL option adds a smooth
nonparametric estimate of the ratio density to each histogram. The INSET statement
species summary statistics to be displayed directly in the graph.

Exploring a Data Distribution

Comparison of Loan Types

Variable:

The UNIVARIATE Procedure

LoanToValueRatio (Loan to Value Ratio)
LoanType = Gold
Quantiles (Definition 5)
Quantile
100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

Estimate
1.0617647
0.8974576
0.6385908
0.4471369
0.2985099
0.2217033
0.1734568
0.1411130
0.1213079
0.0942167
0.0651786

Comparison of Loan Types

Variable:

The UNIVARIATE Procedure

LoanToValueRatio (Loan to Value Ratio)
LoanType = Platinum
Quantiles (Definition 5)
Quantile

Estimate

100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

1.312981
1.050000
0.691803
0.549273
0.430160
0.366168
0.314452
0.273670
0.253124
0.231114
0.215504

Figure 3.3. Quantiles for Loan-to-Value Ratio

The output in Figure 3.3 shows that the median ratio for Platinum loans (0.366) is
greater than the median ratio for Gold loans (0.222). The comparative histogram in
Figure 3.4 enables you to compare the two distributions more easily.

199

200

Chapter 3. The UNIVARIATE Procedure

Figure 3.4. Comparative Histogram for Loan-to-Value Ratio

The comparative histogram shows that the ratio distributions are similar except for a
shift of about 0.14.
A sample program, univar1.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Modeling a Data Distribution

In addition to summarizing a data distribution as in the preceding example, you can
use PROC UNIVARIATE to statistically model a distribution based on a random
sample of data. The following statements create a data set named Aircraft containing
the measurements of a position deviation for a sample of 30 aircraft components.
data Aircraft;
input Deviation @@;
label Deviation = Position Deviation;
datalines;
-.00653 0.00141 -.00702 -.00734 -.00649 -.00601
-.00631 -.00148 -.00731 -.00764 -.00275 -.00497
-.00741 -.00673 -.00573 -.00629 -.00671 -.00246
-.00222 -.00807 -.00621 -.00785 -.00544 -.00511
-.00138 -.00609 0.00038 -.00758 -.00731 -.00455
;
run;

Modeling a Data Distribution

An initial question in the analysis is whether the measurement distribution is normal.
The following statements request a table of moments, the tests for normality, and a
normal probability plot, which are shown in Figure 3.5 and Figure 3.6:
title Position Deviation Analysis;
ods select Moments TestsForNormality MyPlot;
proc univariate data=Aircraft normaltest;
var Deviation;
probplot Deviation / normal (mu=est sigma=est)
square name=MyPlot;
label Deviation = Position Deviation;
inset mean std / format=6.4;
run;

Position Deviation Analysis

Variable:

The UNIVARIATE Procedure

Deviation (Position Deviation)
Moments

N
Mean
Std Deviation
Skewness
Uncorrected SS
Coeff Variation

30
-0.0053067
0.00254362
1.2562507
0.00103245
-47.932613

Sum Weights
Sum Observations
Variance
Kurtosis
Corrected SS
Std Error Mean

30
-0.1592
6.47002E-6
0.69790426
0.00018763
0.0004644

Tests for Normality

Test

--Statistic---

-----p Value------

Shapiro-Wilk
Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

W
D
W-Sq
A-Sq

Pr
Pr
Pr
Pr

0.845364
0.208921
0.329274
1.784881

<
>
>
>

W
D
W-Sq
A-Sq

0.0005
<0.0100
<0.0050
<0.0050

Figure 3.5. Moments and Tests for Normality

All four goodness-of-t tests in Figure 3.5 reject the hypothesis that the measurements are normally distributed.
Figure 3.6 shows a normal probability plot for the measurements. A linear pattern
of points following the diagonal reference line would indicate that the measurements
are normally distributed. Instead, the curved point pattern suggests that a skewed
distribution, such as the lognormal, is more appropriate than the normal distribution.
A lognormal distribution for Deviation is tted in Example 3.26.
A sample program, univar2.sas, for this example is available in the SAS Sample
Library for Base SAS software.

201

202

Chapter 3. The UNIVARIATE Procedure

Figure 3.6. Normal Probability Plot

Syntax
PROC UNIVARIATE < options > ;
BY variables ;
CLASS variable-1 <(v-options)>< variable-2 <(v-options)>>
< / KEYLEVEL= value1 | ( value1 value2 ) >;
FREQ variable ;
HISTOGRAM < variables >< / options > ;
ID variables ;
INSET keyword-list < / options > ;
OUTPUT < OUT=SAS-data-set >
< keyword1=names. . .keywordk=names >< percentile-options >;
PROBPLOT < variables >< / options > ;
QQPLOT < variables >< / options > ;
VAR variables ;
WEIGHT variable ;
The PROC UNIVARIATE statement invokes the procedure. The VAR statement
species the numeric variables to be analyzed, and it is required if the OUTPUT
statement is used to save summary statistics in an output data set. If you do not use
the VAR statement, all numeric variables in the data set are analyzed.

PROC UNIVARIATE Statement

The plot statements HISTOGRAM, PROBPLOT, and QQPLOT create graphical displays, and the INSET statement enhances these displays by adding a table of summary statistics directly on the graph. You can specify one or more of each of the
plot statements, the INSET statement, and the OUTPUT statement. If you use a VAR
statement, the variables listed in a plot statement must be a subset of the variables
listed in the VAR statement.
You can use a CLASS statement to specify one or two variables that group the data
into classication levels. The analysis is carried out for each combination of levels,
and you can use the CLASS statement with a plot statement to create a comparative
display.
You can specify a BY statement to obtain separate analysis for each BY group. The
FREQ statement species a variable whose values provide the frequency for each
observation. The WEIGHT statement species a variable whose values are used to
weight certain statistics. The ID statement species one or more variables to identify
the extreme observations.

PROC UNIVARIATE Statement

PROC UNIVARIATE < options > ;
The PROC UNIVARIATE statement is required to invoke the UNIVARIATE procedure. You can use the PROC UNIVARIATE statement by itself to request a variety
of statistics for summarizing the data distribution of each analysis variable:

sample moments
basic measures of location and variability
condence intervals for the mean, standard deviation, and variance
tests for location
tests for normality
trimmed and Winsorized means
robust estimates of scale
quantiles and related condence intervals
extreme observations and extreme values
frequency counts for observations
missing values

In addition, you can use options in the PROC UNIVARIATE statement to

specify the input data set to be analyzed

specify a graphics catalog for saving graphical output
specify rounding units for variable values
specify the denition used to calculate percentiles
specify the divisor used to calculate variances and standard deviations
request that plots be produced on line printers and dene special printing characters used for features
suppress tables

203

204

Chapter 3. The UNIVARIATE Procedure

The following are the options that can be used with the PROC UNIVARIATE statement:

ALL

requests all statistics and tables that the FREQ, MODES, NEXTRVAL=5, PLOT,
and CIBASIC options generate. If the analysis variables are not weighted,
this option also requests the statistics and tables generated by the CIPCTLDF,
CIPCTLNORMAL, LOCCOUNT, NORMAL, ROBUSTSCALE, TRIMMED=.25,
and WINSORIZED=.25 options. PROC UNIVARIATE also uses any values
that you specify for ALPHA=, MU0=, NEXTRVAL=, CIBASIC, CIPCTLDF,
CIPCTLNORMAL, TRIMMED=, or WINSORIZED= to produce the output.
ALPHA=

species the level of signicance for 100(1 )% condence intervals. The value
must be between 0 and 1; the default value is 0.05, which results in 95% condence
intervals.
Note that specialized ALPHA= options are available for a number of condence interval options. For example, you can specify CIBASIC( ALPHA=0.10 ) to request a
table of basic condence limits at the 90% level. The default values of these options
are the value of the ALPHA= option in the PROC statement.
ANNOTATE=SAS-data-set
ANNO=SAS-data-set

species an input data set that contains annotate variables as described in

SAS/GRAPH Reference. You can use this data set to add features to your highresolution graphics. PROC UNIVARIATE adds the features in this data set to every
high-resolution graph that is produced in the procedure. PROC UNIVARIATE does
not use the ANNOTATE= data set unless you create a high-resolution graph with
the HISTOGRAM, PROBPLOT, or QQPLOT statement. Use the ANNOTATE=
option in the HISTOGRAM, PROBPLOT, or QQPLOT statement if you want to add
a feature to a specic graph produced by the statement.
CIBASIC <(<TYPE=keyword><ALPHA= >)>

requests condence limits for the mean, standard deviation, and variance based on the
assumption that the data are normally distributed. If you use the CIBASIC option,
you must use the default value of VARDEF=, which is DF.
TYPE=keyword

species the type of condence limit, where keyword is LOWER, UPPER, or

TWOSIDED. The default value is TWOSIDED.
ALPHA=

species the level of signicance for 100(1 )% condence intervals. The

value must be between 0 and 1; the default value is 0.05, which results in
95% condence intervals. The default value is the value of ALPHA= given in
the PROC statement.
CIPCTLDF <(<TYPE=keyword><ALPHA= >)>
CIQUANTDF <(<TYPE=keyword><ALPHA= >)>

requests condence limits for quantiles based on a method that is distribution-free. In

other words, no specic parametric distribution such as the normal is assumed for the

PROC UNIVARIATE Statement

data. PROC UNIVARIATE uses order statistics (ranks) to compute the condence
limits as described by Hahn and Meeker (1991). This option does not apply if you
use a WEIGHT statement.
TYPE=keyword

species the type of condence limit, where keyword is LOWER, UPPER,

SYMMETRIC, or ASYMMETRIC. The default value is SYMMETRIC.
ALPHA=

species the level of signicance for 100(1 )% condence intervals. The

value must be between 0 and 1; the default value is 0.05, which results in
95% condence intervals. The default value is the value of ALPHA= given in
the PROC statement.
CIPCTLNORMAL <(<TYPE=keyword><ALPHA= >)>
CIQUANTNORMAL <(<TYPE=keyword><ALPHA= >)>

requests condence limits for quantiles based on the assumption that the data are
normally distributed. The computational method is described in Section 4.4.1 of
Hahn and Meeker (1991) and uses the noncentral t distribution as given by Odeh and
Owen (1980). This option does not apply if you use a WEIGHT statement
TYPE=keyword

species the type of condence limit, where keyword is LOWER, UPPER, or

TWOSIDED. The default is TWOSIDED.
ALPHA=

species the level of signicance for 100(1 )% condence intervals. The

value must be between 0 and 1; the default value is 0.05, which results in
95% condence intervals. The default value is the value of ALPHA= given in
the PROC statement.
DATA=SAS-data-set

species the input SAS data set to be analyzed. If the DATA= option is omitted, the
procedure uses the most recently created SAS data set.
EXCLNPWGT

excludes observations with nonpositive weight values (zero or negative) from the
analysis. By default, PROC UNIVARIATE treats observations with negative weights
like those with zero weights and counts them in the total number of observations.
This option applies only when you use a WEIGHT statement.
FREQ

requests a frequency table that consists of the variable values, frequencies, cell percentages, and cumulative percentages.
If you specify the WEIGHT statement, PROC UNIVARIATE includes the weighted
count in the table and uses this value to compute the percentages.
GOUT=graphics-catalog

species the SAS catalog that PROC UNIVARIATE uses to save high-resolution
graphics output. If you omit the libref in the name of the graphics-catalog, PROC
UNIVARIATE looks for the catalog in the temporary library called WORK and creates the catalog if it does not exist.

205

206

Chapter 3. The UNIVARIATE Procedure

LOCCOUNT

requests a table that shows the number of observations greater than, not equal to, and
less than the value of MU0=. PROC UNIVARIATE uses these values to construct the
sign test and the signed rank test. This option does not apply if you use a WEIGHT
statement.
MODES|MODE

requests a table of all possible modes. By default, when the data contain multiple
modes, PROC UNIVARIATE displays the lowest mode in the table of basic statistical
measures. When all the values are unique, PROC UNIVARIATE does not produce a
table of modes.
MU0=values
LOCATION=values

species the value of the mean or location parameter (0 ) in the null hypothesis for
tests of location summarized in the table labeled Tests for Location: Mu0=value. If
you specify one value, PROC UNIVARIATE tests the same null hypothesis for all
analysis variables. If you specify multiple values, a VAR statement is required, and
PROC UNIVARIATE tests a different null hypothesis for each analysis variable in
the corresponding order. The default value is 0.
The following statement tests the hypothesis 0 = 0 for the rst variable and the
hypothesis 0 = 0.5 for the second variable.
proc univariate mu0=0 0.5;
NEXTROBS=n

species the number of extreme observations that PROC UNIVARIATE lists in the
table of extreme observations. The table lists the n lowest observations and the n
highest observations. The default value is 5, and n can range between 0 and half the
maximum number of observations. You can specify NEXTROBS=0 to suppress the
table of extreme observations.
NEXTRVAL=n

species the number of extreme values that PROC UNIVARIATE lists in the table of
extreme values. The table lists the n lowest unique values and the n highest unique
values. The default value is 0, and n can range between 0 and half the maximum
number of observations. By default, n = 0 and no table is displayed.
NOBYPLOT

suppresses side-by-side box plots that are created by default when you use the BY
statement and the ALL option or the PLOT option in the PROC statement.
NOPRINT

suppresses all the tables of descriptive statistics that the PROC UNIVARIATE statement creates. NOPRINT does not suppress the tables that the HISTOGRAM statement creates. You can use the NOPRINT option in the HISTOGRAM statement to
suppress the creation of its tables. Use NOPRINT when you want to create an OUT=
output data set only.

PROC UNIVARIATE Statement

NORMAL
NORMALTEST

requests tests for normality that include a series of goodness-of-t tests based on
the empirical distribution function. The table provides test statistics and p-values
for the Shapiro-Wilk test (provided the sample size is less than or equal to 2000),
the Kolmogorov-Smirnov test, the Anderson-Darling test, and the Cramr-von Mises
test. This option does not apply if you use a WEIGHT statement.
PCTLDEF=value
DEF=value

species the denition that PROC UNIVARIATE uses to calculate quantiles. The
default value is 5. Values can be 1, 2, 3, 4, or 5. You cannot use PCTLDEF= when
you compute weighted quantiles. See the section Calculating Percentiles on page
273 for details on quantile denitions.
PLOTS | PLOT

produces a stem-and-leaf plot (or a horizontal bar chart), a box plot, and a normal
probability plot in line printer output. If you use a BY statement, side-by-side box
plots that are labeled Schematic Plots appear after the univariate analysis for the
last BY group.
PLOTSIZE=n

species the approximate number of rows used in line-printer plots requested with the
PLOTS option. If n is larger than the value of the SAS system option PAGESIZE=,
PROC UNIVARIATE uses the value of PAGESIZE=. If n is less than 8, PROC
UNIVARIATE uses eight rows to draw the plots.
ROBUSTSCALE

produces a table with robust estimates of scale. The statistics include the interquartile range, Ginis mean difference, the median absolute deviation about the median
(MAD), and two statistics proposed by Rousseeuw and Croux (1993), Qn , and Sn .
This option does not apply if you use a WEIGHT statement.
ROUND=units

species the units to use to round the analysis variables prior to computing statistics.
If you specify one unit, PROC UNIVARIATE uses this unit to round all analysis
variables. If you specify multiple units, a VAR statement is required, and each unit
rounds the values of the corresponding analysis variable. If ROUND=0, no rounding
occurs. The ROUND= option reduces the number of unique variable values, thereby
reducing memory requirements for the procedure. For example, to make the rounding
unit 1 for the rst analysis variable and 0.5 for the second analysis variable, submit
the statement
proc univariate round=1 0.5;
var yldstren tenstren;
run;

When a variable value is midway between the two nearest rounded points, the value
is rounded to the nearest even multiple of the roundoff value. For example, with a
roundoff value of 1, the variable values of 2.5, 2.2, and 1.5 are rounded to 2;
the values of 0.5, 0.2, and 0.5 are rounded to 0; and the values of 0.6, 1.2, and 1.4
are rounded to 1.

207

208

Chapter 3. The UNIVARIATE Procedure

TRIMMED=values <(<TYPE=keyword><ALPHA= >)>

TRIM=values <(<TYPE=keyword><ALPHA= >)>
requests a table of trimmed means, where value species the number or the proportion of observations that PROC UNIVARIATE trims. If the value is the number n

of trimmed observations, n must be between 0 and half the number of nonmissing

observations. If value is a proportion p between 0 and 1 , the number of observa2
tions that PROC UNIVARIATE trims is the smallest integer that is greater than or
equal to np, where n is the number of observations. To include condence limits
for the mean and the Students t test in the table, you must use the default value of
VARDEF= which is DF. For details concerning the computation of trimmed means,
see the section Trimmed Means on page 279.
TYPE= keyword

species the type of condence limit for the mean, where keyword is LOWER,
UPPER, or TWOSIDED. The default value is TWOSIDED.
ALPHA=

species the level of signicance for 100(1 )% condence intervals. The

value must be between 0 and 1; the default value is 0.05, which results in
95% condence intervals.
This option does not apply if you use a WEIGHT statement.
VARDEF=divisor

species the divisor to use in the calculation of variances and standard deviation. By
default, VARDEF=DF. The following table shows the possible values for divisor and
associated divisors.
Table 3.1. Possible Values for VARDEF=

Value
DF
N
WDF
WEIGHT|WGT

Divisor
Degrees of freedom
Number of observations
Sum of weights minus one
Sum of weights

Formula for Divisor

n1
n
(i wi ) 1
i wi

CSS
The procedure computes the variance as divisor where CSS is the corrected sums of
n
2 . When you weight the analysis variables, CSS =

squares and equals i=1 (xi x)

n
2 where x is the weighted mean.

w
i=1 (wi xi xw )

The default value is DF. To compute the standard error of the mean, condence limits,
and Students t test, use the default value of VARDEF=.
When you use the WEIGHT statement and VARDEF=DF, the variance is an estimate
s2
of s2 where the variance of the ith observation is var(xi ) = wi and wi is the weight
for the ith observation. This yields an estimate of the variance of an observation with
unit weight.
When you use the WEIGHT statement and VARDEF=WGT, the computed variance
2
is asymptotically (for large n) an estimate of s where w is the average weight. This

w
yields an asymptotic estimate of the variance of an observation with average weight.

BY Statement
WINSORIZED=values <(<TYPE=keyword><ALPHA= >)>
WINSOR=values <(<TYPE=keyword><ALPHA=>)>
requests of a table of Winsorized means, where value is the number or the proportion

of observations that PROC UNIVARIATE uses to compute the Winsorized mean. If

the value is the number n of winsorized observations, n must be between 0 and half
the number of nonmissing observations. If value is a proportion p between 0 and 1 ,
2
the number of observations that PROC UNIVARIATE uses is equal to the smallest
integer that is greater than or equal to np, where n is the number of observations. To
include condence limits for the mean and the student t test in the table, you must use
the default value of VARDEF=, which is DF. For details concerning the computation
of Winsorized means, see the section Winsorized Means on page 278.
TYPE=keyword

species the type of condence limit for the mean, where keyword is LOWER,
UPPER, or TWOSIDED. The default is TWOSIDED.
ALPHA=

species the level of signicance for 100(1 )% condence intervals. The

value must be between 0 and 1; the default value is 0.05, which results in
95% condence intervals.
This option does not apply if you use a WEIGHT statement.

BY Statement
BY variables ;
You can specify a BY statement with PROC UNIVARIATE to obtain separate analyses for each BY group. The BY statement species the variables that the procedure
uses to form BY groups. You can specify more than one variable. If you do not
use the NOTSORTED option in the BY statement, the observations in the data set
must either be sorted by all the variables that you specify or they must be indexed
appropriately.
DESCENDING

species that the data set is sorted in descending order by the variable that immediately follows the word DESCENDING in the BY statement.
NOTSORTED

species that observations are not necessarily sorted in alphabetic or numeric order.
The data are grouped in another way, for example, chronological order.
The requirement for ordering or indexing observations according to the values of
BY variables is suspended for BY-group processing when you use the NOTSORTED
option. In fact, the procedure does not use an index if you specify NOTSORTED.
The procedure denes a BY group as a set of contiguous observations that have the
same values for all BY variables. If observations with the same values for the BY
variables are not contiguous, the procedure treats each contiguous set as a separate
BY group.

209

210

Chapter 3. The UNIVARIATE Procedure

CLASS Statement
CLASS variable-1 <(v-options)>< variable-2 <(v-options)>>
< / KEYLEVEL= value1 | ( value1 value2 ) >;
The CLASS statement species one or two variables that the procedure uses to group
the data into classication levels. Variables in a CLASS statement are referred to as
class variables. Class variables can be numeric or character. Class variables can have
oating point values, but they typically have a few discrete values that dene levels of
the variable. You do not have to sort the data by class variables. PROC UNIVARIATE
uses the formatted values of the class variables to determine the classication levels.
You can specify the following v-options enclosed in parentheses after the class variable:
MISSING

species that missing values for the CLASS variable are to be treated as valid classication levels. Special missing values that represent numeric values (the letters A
through Z and the underscore ( ) character) are each considered as a separate value.
If you omit MISSING, PROC UNIVARIATE excludes the observations with a missing class variable value from the analysis. Enclose this option in parentheses after the
class variable.
ORDER=DATA | FORMATTED | FREQ | INTERNAL

species the display order for the class variable values. The default value is
INTERNAL. You can specify the following values with the ORDER=option:
DATA

orders values according to their order in the input data set. When you use
a HISTOGRAM, PROBPLOT, or QQPLOT statement, PROC UNIVARIATE
displays the rows (columns) of the comparative plot from top to bottom (left
to right) in the order that the class variable values rst appear in the input data
set.
FORMATTED

orders values by their ascending formatted values. This order may depend on
your operating environment. When you use a HISTOGRAM, PROBPLOT, or
QQPLOT statement, PROC UNIVARIATE displays the rows (columns) of the
comparative plot from top to bottom (left to right) in increasing order of the
formatted class variable values. For example, suppose a numeric class variable
DAY (with values 1, 2, and 3) has a user-dened format that assigns Wednesday
to the value 1, Thursday to the value 2, and Friday to the value 3. The rows
of the comparative plot will appear in alphabetical order (Friday, Thursday,
Wednesday) from top to bottom.
If there are two or more distinct internal values with the same formatted value,
then PROC UNIVARIATE determines the order by the internal value that occurs rst in the input data set. For numerical variables without an explicit
format, the levels are ordered by their internal values.

CLASS Statement
FREQ

orders values by descending frequency count so that levels with the most observations are listed rst. If two or more values have the same frequency count,
PROC UNIVARIATE uses the formatted values to determine the order.
When you use a HISTOGRAM, PROBPLOT, or QQPLOT statement, PROC
UNIVARIATE displays the rows (columns) of the comparative plot from top
to bottom (left to right) in order of decreasing frequency count for the class
variable values.
INTERNAL

orders values by their unformatted values, which yields the same order as
PROC SORT. This order may depend on your operating environment.
When you use a HISTOGRAM, PROBPLOT, or QQPLOT statement, PROC
UNIVARIATE displays the rows (columns) of the comparative plot from top
to bottom (left to right) in increasing order of the internal (unformatted) values
of the class variable. The rst class variable is used to label the rows of the
comparative plots (top to bottom). The second class variable is used to label
the columns of the comparative plots (left to right). For example, suppose a
numeric class variable DAY (with values 1, 2, and 3) has a user-dened format
that assigns Wednesday to the value 1, Thursday to the value 2, and Friday to
the value 3. The rows of the comparative plot will appear in day-of-the-week
order (Wednesday, Thursday, Friday) from top to bottom.
You can specify the following option after the slash (/) in the CLASS statement.
KEYLEVEL=value1 | ( value1 value2 )

species the key cell in a comparative plot. PROC UNIVARIATE rst determines
the bin size and midpoints for the key cell, and then extends the midpoint list to
accommodate the data ranges for the remaining cells. Thus, the choice of the key
cell determines the uniform horizontal axis that PROC UNIVARIATE uses for all
cells. If you specify only one class variable and use a HISTOGRAM statement,
KEYLEVEL= value identies the key cell as the level for which variable is equal to
value. By default, PROC UNIVARIATE sorts the levels in the order that is determined by the ORDER= option. Then, the key cell is the rst occurrence of a level in
this order. The cells display in order from top to bottom or left to right. Consequently,
the key cell appears at the top (or left). When you specify a different key cell with
the KEYLEVEL= option, this cell appears at the top (or left).
Likewise, with the PROBPLOT and QQPLOT statements, the key cell determines
uniform axis scaling. If you specify two class variables, use KEYLEVEL= value1
value2 to identify the key cell as the level for which variable-n is equal to value-n.
By default, PROC UNIVARIATE sorts the levels of the rst CLASS variable in the
order that is determined by its ORDER= option and, within each of these levels, it
sorts the levels of the second CLASS variable in the order that is determined by its
ORDER= option. Then, the default key cell is the rst occurrence of a combination
of levels for the two variables in this order. The cells display in the order of the rst
CLASS variable from top to bottom and in the order of the second CLASS variable
from left to right. Consequently, the default key cell appears at the upper left corner.

211

212

Chapter 3. The UNIVARIATE Procedure

When you specify a different key cell with the KEYLEVEL= option, this cell appears
at the upper left corner.
The length of the KEYLEVEL= value cannot exceed 16 characters and you must
specify a formatted value.
The KEYLEVEL= option does not apply unless you specify a HISTOGRAM,
PROBPLOT, or QQPLOT statement.

FREQ Statement
FREQ variable ;
The FREQ statement species a numeric variable whose value represents the frequency of the observation. If you use the FREQ statement, the procedure assumes
that each observation represents n observations, where n is the value of variable. If
the variable is not an integer, the SAS System truncates it. If the variable is less
than 1 or is missing, the procedure excludes that observation from the analysis. See
Example 3.6.
Note: The FREQ statement affects the degrees of freedom, but the WEIGHT statement does not.

HISTOGRAM Statement
HISTOGRAM < variables >< / options >;
The HISTOGRAM statement creates histograms and optionally superimposes estimated parametric and nonparametric probability density curves. You cannot use the
WEIGHT statement with the HISTOGRAM statement. You can use any number of
HISTOGRAM statements after a PROC UNIVARIATE statement. The components
of the HISTOGRAM statement are described as follows.
variables

are the variables for which histograms are to be created. If you specify a VAR statement, the variables must also be listed in the VAR statement. Otherwise, the variables
can be any numeric variables in the input data set. If you do not specify variables in
a VAR statement or in the HISTOGRAM statement, then by default, a histogram is
created for each numeric variable in the DATA= data set. If you use a VAR statement
and do not specify any variables in the HISTOGRAM statement, then by default, a
histogram is created for each variable listed in the VAR statement.
For example, suppose a data set named Steel contains exactly two numeric variables
named Length and Width. The following statements create two histograms, one for
Length and one for Width:
proc univariate data=Steel;
histogram;
run;

HISTOGRAM Statement
Likewise, the following statements create histograms for Length and Width:
proc univariate data=Steel;
var Length Width;
histogram;
run;

The following statements create a histogram for Length only:

proc univariate data=Steel;
var Length Width;
histogram Length;
run;
options

add features to the histogram. Specify all options after the slash (/) in the
HISTOGRAM statement. Options can be one of the following:
primary options for tted parametric distributions and kernel density estimates
secondary options for tted parametric distributions and kernel density estimates
general options for graphics and output data sets
For example, in the following statements, the NORMAL option displays a tted normal curve on the histogram, the MIDPOINTS= option species midpoints for the
histogram, and the CTEXT= option species the color of the text:
proc univariate data=Steel;
histogram Length / normal
midpoints = 5.6 5.8 6.0 6.2 6.4
ctext
= blue;
run;

Table 3.2 through Table 3.12 list the HISTOGRAM options by function. For complete
descriptions, see the the section Dictionary of Options on page 217.

Parametric Density Estimation Options

Table 3.2 lists primary options that display a parametric density estimate on the histogram.

213

214

Chapter 3. The UNIVARIATE Procedure

Table 3.2. Primary Options for Parametric Fitted Distributions
BETA(beta-options)
Fits beta distribution with threshold

parameter , scale parameter , and

shape parameters and
EXPONENTIAL(exponential-options)

Fits exponential distribution with

threshold parameter and scale parameter

GAMMA(gamma-options)

Fits gamma distribution with threshold parameter , scale parameter ,

and shape parameter

LOGNORMAL(lognormal-options)

Fits lognormal distribution with

threshold parameter , scale parameter , and shape parameter

NORMAL(normal-options)

Fits normal distribution with mean

and standard deviation

WEIBULL(Weibull-options)

Fits Weibull distribution with threshold parameter , scale parameter ,

and shape parameter c

Table 3.3 through Table 3.9 list secondary options that specify parameters for tted
parametric distributions and that control the display of tted curves. Specify these
secondary options in parentheses after the primary distribution option. For example,
you can t a normal curve by specifying the NORMAL option as follows:
proc univariate;
histogram / normal(color=red mu=10 sigma=0.5);
run;

The COLOR= normal-option draws the curve in red, and the MU= and SIGMA=
normal-options specify the parameters = 10 and = 0.5 for the curve. Note
that the sample mean and sample standard deviation are used to estimate and ,
respectively, when the MU= and SIGMA= normal-options are not specied.
Table 3.3. Secondary Options Used with All Parametric Distribution Options
COLOR=color
Species color of density curve

FILL

Fills area under density curve

L=linetype

Species line type of curve

MIDPERCENTS

Prints table of midpoints of histogram intervals

NOPRINT

Suppresses tables summarizing curve

PERCENTS=value-list

Lists percents for which quantiles calculated from data and

quantiles estimated from curve are tabulated

W=n

Species width of density curve

HISTOGRAM Statement
Table 3.4. Secondary Beta-Options
ALPHA=value
Species rst shape parameter for beta curve

BETA=value

Species second shape parameter for beta curve

SIGMA=value | EST

Species scale parameter for beta curve

THETA=value | EST

Species lower threshold parameter for beta curve

Table 3.5. Secondary Exponential-Options

Species standard deviation for normal curve

Table 3.9. Secondary Weibull-Options

C=value
Species shape parameter c for Weibull curve

SIGMA=value

Species scale parameter for Weibull curve

THETA=value | EST

Species threshold parameter for Weibull curve

Nonparametric Density Estimation Options

Use the option KERNEL(kernel-options) to compute kernel density estimates.
Specify the following secondary options in parentheses after the KERNEL option
to control features of density estimates requested with the KERNEL option.
Table 3.10. Kernel-Options
C=value-list | MISE

Species standardized bandwidth parameter c

COLOR=color

Species color of the kernel density curve

FILL

Fills area under kernel density curve

K=NORMAL |
QUADRATIC |
TRIANGULAR
L=linetype

Species type of kernel function

Species line type used for kernel density curve

LOWER=

Species lower bound for kernel density curve

UPPER=

Species upper bound for kernel density curve

W=n

Species line width for kernel density curve

215

216

Chapter 3. The UNIVARIATE Procedure

General Options
Table 3.11 summarizes options for enhancing histograms, and Table 3.12 summarizes
options for requesting output data sets.
Table 3.11. General Graphics Options

Option
ANNOKEY
ANNOTATE=
BARWIDTH=
CAXIS=
CBARLINE=
CFILL=
CFRAME=
CFRAMESIDE=
CFRAMETOP=
CGRID=
CHREF=
CPROP=
CTEXT=
CTEXTSIDE=
CTEXTTOP=
CVREF=
DESCRIPTION=
ENDPOINTS=
FONT=
FORCEHIST
GRID
FRONTREF
HEIGHT=
HMINOR=
HOFFSET=
HREF=
HREFLABELS=
HREFLABPOS=
INFONT=
INHEIGHT=
INTERTILE=
LGRID=
LHREF=
LVREF=
MAXNBIN=
MAXSIGMAS=
MIDPOINTS=
NAME=
NCOLS=
NOBARS
NOFRAME

Description
Applies annotation requested in ANNOTATE= data set to key cell only
Species annotate data set
Species width for the bars
Species color for axis
Species color for outlines of histogram bars
Species color for lling under curve
Species color for frame
Species color for lling frame for row labels
Species color for lling frame for column labels
Species color for grid lines
Species color for HREF= lines
Species color for proportion of frequency bar
Species color for text
Species color for row labels of comparative histograms
Species color for column labels of comparative histograms
Species color for VREF= lines
Species description for plot in graphics catalog
Lists endpoints for histogram intervals
Species software font for text
Forces creation of histogram
Creates a grid
Draws reference lines in front of histogram bars
Species height of text used outside framed areas
Species number of horizontal minor tick marks
Species offset for horizontal axis
Species reference lines perpendicular to the horizontal axis
Species labels for HREF= lines
Species vertical position of labels for HREF= lines
Species software font for text inside framed areas
Species height of text inside framed areas
Species distance between tiles
Species a line type for grid lines
Species line style for HREF= lines
Species line style for VREF= lines
Species maximum number of bins to display
Limits the number of bins that display to within a specied number of
standard deviations above and below mean of data in key cell
Lists midpoints for histogram intervals
Species name for plot in graphics catalog
Species number of columns in comparative histogram
Suppresses histogram bars
Suppresses frame around plotting area

HISTOGRAM Statement
Table 3.11. (continued)

Option
NOHLABEL
NOPLOT
NOVLABEL
NOVTICK
NROWS=
PFILL=
RTINCLUDE
TURNVLABELS
VAXIS=
VAXISLABEL=
VMINOR=
VOFFSET=
VREF=
VREFLABELS=
VREFLABPOS=
VSCALE=
WAXIS=
WBARLINE=
WGRID=

Description
Suppresses label for horizontal axis
Suppresses plot
Suppresses label for vertical axis
Suppresses tick marks and tick mark labels for vertical axis
Species number of rows in comparative histogram
Species pattern for lling under curve
Includes right endpoint in interval
Turn and vertically string out characters in labels for vertical axis
Species AXIS statement or values for vertical axis
Species label for vertical axis
Species number of vertical minor tick marks
Species length of offset at upper end of vertical axis
Species reference lines perpendicular to the vertical axis
Species labels for VREF= lines
Species horizontal position of labels for VREF= lines
Species scale for vertical axis
Species line thickness for axes and frame
Species line thickness for bar outlines
Species line thickness for grid

Table 3.12. Options for Requesting Output Data Sets

Option
MIDPERCENTS
OUTHISTOGRAM=

Description
Creates table of histogram intervals
Species information on histogram intervals

Dictionary of Options
The following entries provide detailed descriptions of options in the HISTOGRAM
statement.
ALPHA=value

species the shape parameter for tted curves requested with the BETA and
GAMMA options. Enclose the ALPHA= option in parentheses after the BETA or
GAMMA options. By default, the procedure calculates a maximum likelihood estimate for . You can specify A= as an alias for ALPHA= if you use it as a beta-option.
You can specify SHAPE= as an alias for ALPHA= if you use it as a gamma-option.
ANNOKEY

applies the annotation requested with the ANNOTATE= option to the key cell only.
By default, the procedure applies annotation to all of the cells. This option is not
available unless you use the CLASS statement. You can use the KEYLEVEL= option
in the CLASS statement to specify the key cell.

217

218

Chapter 3. The UNIVARIATE Procedure

ANNOTATE=SAS-data-set
ANNO=SAS-data-set

species an input data set containing annotate variables as described in SAS/GRAPH

Software: Reference. The ANNOTATE= data set you specify in the HISTOGRAM
statement is used for all plots created by the statement. You can also specify an
ANNOTATE= data set in the PROC UNIVARIATE statement to enhance all plots
created by the procedure.
BARWIDTH=value

species the width of the histogram bars in screen percent units.

BETA <(beta-options)>

displays a tted beta density curve on the histogram. The BETA option can occur
only once in a HISTOGRAM statement. The beta distribution is bounded below by
the parameter and above by the value + . Use the THETA= and SIGMA= betaoptions to specify these parameters. By default, THETA=0 and SIGMA=1. You can
specify THETA=EST and SIGMA=EST to request maximum likelihood estimates
for and . See Example 3.21.
Note: Three- and four-parameter maximum likelihood estimation may not always
converge. The beta distribution has two shape parameters, and . If these parameters are known, you can specify their values with the ALPHA= and BETA=
beta-options. By default, the procedure computes maximum likelihood estimates for
and . Table 3.3 (page 214) and Table 3.4 (page 215) list options you can specify
with the BETA option.
BETA=value
B=value

species the second shape parameter for beta density curves requested with the
BETA option. Enclose the BETA= option in parentheses after the BETA option. By
default, the procedure calculates a maximum likelihood estimate for .
C=value

species the shape parameter c for Weibull density curves requested with the
WEIBULL option. Enclose the C= Weibull-option in parentheses after the WEIBULL
option. If you do not specify a value for c, the procedure calculates a maximum likelihood estimate. You can specify the SHAPE= Weibull-option as an alias for the C=
Weibull-option.
C=value-list | MISE

species the standardized bandwidth parameter c for kernel density estimates requested with the KERNEL option. Enclose the C= kernel-option in parentheses after
the KERNEL option. You can specify up to ve values to request multiple estimates.
You can also specify the C=MISE option, which produces the estimate with a bandwidth that minimizes the approximate mean integrated square error (MISE).
You can also use the C= kernel-option with the K= kernel-option, which species the
kernel function, to compute multiple estimates. If you specify more kernel functions
than bandwidths, the last bandwidth in the list is repeated for the remaining estimates.
Likewise, if you specify more bandwidths than kernel functions, the last kernel func-

HISTOGRAM Statement
tion is repeated for the remaining estimates. If you do not specify a value for c, the
bandwidth that minimizes the approximate MISE is used for all the estimates.
CAXIS=color
CAXES=color
CA=color

species the color for the axes and tick marks. This option overrides any COLOR=
specications in an AXIS statement. The default value is the rst color in the device
color list.
CBARLINE=color

species the color for the outline of the histogram bars. This option overrides the C=
option in the SYMBOL1 statement. The default value is the rst color in the device
color list.
CFILL=color

species the color to ll the bars of the histogram (or the area under a tted density
curve if you also specify the FILL option). See the entries for the FILL and PFILL=
options for additional details. Refer to SAS/GRAPH Software: Reference for a list of
colors. By default, bars and curve areas are not lled.
CFRAME=color

species the color for the area that is enclosed by the axes and frame. The area is not
lled by default.
CFRAMESIDE=color

species the color to ll the frame area for the row labels that display along the left
side of the comparative histogram. This color also lls the frame area for the label
of the corresponding class variable (if you associate a label with the variable). By
default, these areas are not lled. This option is not available unless you use the
CLASS statement.
CFRAMETOP=color

species the color to ll the frame area for the column labels that display across the
top of the comparative histogram. This color also lls the frame area for the label
of the corresponding class variable (if you associate a label with the variable). By
default, these areas are not lled. This option is not available unless you use the
CLASS statement.
CGRID=color

species the color for grid lines when a grid displays on the histogram. The default
color is the rst color in the device color list. This option also produces a grid.
CHREF=color
CH=color

species the color for horizontal axis reference lines requested by the HREF= option.
The default is the rst color in the device color list.
COLOR=color

species the color of the density curve. Enclose the COLOR= option in parentheses
after the distribution option or the KERNEL option. If you use the COLOR= option
with the KERNEL option, you can specify a list of up to ve colors in parentheses

219

220

Chapter 3. The UNIVARIATE Procedure

for multiple kernel density estimates. If there are more estimates than colors, the last
color specied is used for the remaining estimates.

CPROP=color | EMPTY

species the color for a horizontal bar whose length (relative to the width of the tile)
indicates the proportion of the total frequency that is represented by the corresponding
cell in a comparative histogram. By default, no bars are displayed. This option is not
available unless you use the CLASS statement. You can specify the keyword EMPTY
to display empty bars. See Example 3.20.
CTEXT=color
CT=color

species the color for tick mark values and axis labels. The default is the color
specied for the CTEXT= option in the GOPTIONS statement. In the absence of a
GOPTIONS statement, the default color is the rst color in the device color list.
CTEXTSIDE=color

species the color for the row labels that display along the left side of the comparative
histogram. By default, the color specied by the CTEXT= option is used. If you omit
the CTEXT= option, the color specied in the GOPTIONS statement is used. If you
omit the GOPTIONS statement, the the rst color in the device color list is used.
This option is not available unless you use the CLASS statement. You can specify
the CFRAMESIDE= option to change the background color for the row labels.
CTEXTTOP=color

species the color for the column labels that display along the left side of the comparative histogram. By default, the color specied by the CTEXT= option is used.
If you omit the CTEXT= option, the color specied in the GOPTIONS statement is
used. If you omit the GOPTIONS statement, the the rst color in the device color list
is used. This option is not available unless you specify the CLASS statement. You
can use the CFRAMETOP= option to change the background color for the column
labels.
CVREF=color
CV=color

species the color for lines requested with the VREF= option. The default is the rst
color in the device color list.
DESCRIPTION=string
DES=string

species a description, up to 40 characters long, that appears in the PROC GREPLAY

master menu. The default value is the variable name.
ENDPOINTS <=values | KEY | UNIFORM>

uses the endpoints as the tick mark values for the horizontal axis and determines how
to compute the bin width of the histogram bars, where values species values for both
the left and right endpoint of each histogram interval. The width of the histogram bars
is the difference between consecutive endpoints. The procedure uses the same values
for all variables.
The range of endpoints must cover the range of the data. For example, if you specify

HISTOGRAM Statement
endpoints=2 to 10 by 2

then all of the observations must fall in the intervals [2,4) [4,6) [6,8) [8,10]. You also
must use evenly spaced endpoints which you list in increasing order.
KEY

determines the endpoints for the data in the key cell. The initial
number of endpoints is based on the number of observations in the
key cell using the method of Terrell and Scott (1985). The procedure extends the endpoint list for the key cell in either direction as
necessary until it spans the data in the remaining cells.

UNIFORM

determines the endpoints by using all the observations as if there

were no cells. In other words, the number of endpoints is based
on the total sample size by using the method of Terrell and Scott
(1985).

Neither KEY nor UNIFORM apply unless you use the CLASS statement.
If you omit ENDPOINTS, the procedure uses the midpoints. If you specify
ENDPOINTS, the procedure computes the endpoints by using an algorithm (Terrell
and Scott 1985) that is primarily applicable to continuous data that are approximately
normally distributed.
If you specify both MIDPOINTS= and ENDPOINTS, the procedure issues a warning
message and uses the endpoints.
If you specify RTINCLUDE, the procedure includes the right endpoint of each histogram interval in that interval instead of including the left endpoint.
If you use a CLASS statement and specify ENDPOINTS, the procedure uses
ENDPOINTS=KEY as the default. However if the key cell is empty, then the procedure uses ENDPOINTS=UNIFORM.
EXPONENTIAL <(exponential-options)>
EXP <(exponential-options)>

displays a tted exponential density curve on the histogram. The EXPONENTIAL

option can occur only once in a HISTOGRAM statement. The parameter must be
less than or equal to the minimum data value. Use the THETA= exponential-option
to specify . By default, THETA=0. You can specify THETA=EST to request the
maximum likelihood estimate for . Use the SIGMA= exponential-option to specify . By default, the procedure computes a maximum likelihood estimate for .
Table 3.3 (page 214) and Table 3.5 (page 215) list options you can specify with the
EXPONENTIAL option.
FILL

lls areas under the tted density curve or the kernel density estimate with colors
and patterns. The FILL option can occur with only one tted curve. Enclose the
FILL option in parentheses after a density curve option or the KERNEL option. The
CFILL= and PFILL= options specify the color and pattern for the area under the
curve. For a list of available colors and patterns, see SAS/GRAPH Reference.

221

222

Chapter 3. The UNIVARIATE Procedure

FONT=font

species a software font for reference line and axis labels. You can also specify fonts
for axis labels in an AXIS statement. The FONT= font takes precedence over the
FTEXT= font specied in the GOPTIONS statement. Hardware characters are used
by default.
FORCEHIST

forces the creation of a histogram if there is only one unique observation. By default,
a histogram is not created if the standard deviation of the data is zero.
FRONTREF

draws reference lines requested with the HREF= and VREF= options in front of the
histogram bars. By default, reference lines are drawn behind the histogram bars and
can be obscured by them.
GAMMA <(gamma-options)>

displays a tted gamma density curve on the histogram. The GAMMA option can
occur only once in a HISTOGRAM statement. The parameter must be less than
the minimum data value. Use the THETA= gamma-option to specify . By default, THETA=0. You can specify THETA=EST to request the maximum likelihood
estimate for . Use the ALPHA= and the SIGMA= gamma-options to specify the
shape parameter and the scale parameter . By default, PROC UNIVARIATE
computes maximum likelihood estimates for and . The procedure calculates the
maximum likelihood estimate of iteratively using the Newton-Raphson approximation. Table 3.3 (page 214) and Table 3.6 (page 215) list options you can specify with
the GAMMA option. See Example 3.22.
GRID

displays a grid on the histogram. Grid lines are horizontal lines that are positioned at
major tick marks on the vertical axis.
HEIGHT=value

species the height, in percentage screen units, of text for axis labels, tick mark
labels, and legends. This option takes precedence over the HTEXT= option in the
GOPTIONS statement.
HMINOR=n
HM=n

species the number of minor tick marks between each major tick mark on the horizontal axis. Minor tick marks are not labeled. By default, HMINOR=0.
HOFFSET=value

species the offset, in percentage screen units, at both ends of the horizontal axis.
You can use HOFFSET=0 to eliminate the default offset.
HREF=values

draws reference lines that are perpendicular to the horizontal axis at the values
that you specify. If a reference line is almost completely obscured, then use the
FRONTREF option to draw the reference lines in front of the histogram bars. Also
see the CHREF=, HREFCHAR=, and LHREF= options.

HISTOGRAM Statement
HREFLABELS=label1 . . . labeln
HREFLABEL=label1 . . . labeln
HREFLAB=label1 . . . labeln

species labels for the lines requested by the HREF= option. The number of labels
must equal the number of lines. Enclose each label in quotes. Labels can have up to
16 characters.
HREFLABPOS=1 | 2 | 3

species the vertical position of HREFLABELS= labels.

If you specify
HREFLABPOS=1, the labels are positioned along the top of the histogram. If
you specify HREFLABPOS=2, the labels are staggered from top to bottom of the
histogram. If you specify HREFLABPOS=3, the labels are positioned along the
bottom of the histogram. By default, HREFLABPOS=1.
INFONT=font

species a software font to use for text inside the framed areas of the histogram.
The INFONT= option takes precedence over the FTEXT= option in the GOPTIONS
statement. For a list of fonts, see SAS/GRAPH Reference.
INHEIGHT=value

species the height, in percentage screen units, of text used inside the framed areas
of the histogram. By default, the height specied by the HEIGHT= option is used.
If you do not specify the HEIGHT= option, the height specied with the HTEXT=
option in the GOPTIONS statement is used.
INTERTILE=value

species the distance, in horizontal percentage screen units, between the framed areas, which are called tiles. By default, INTERTILE=0.75 percentage screen units.
This option is not available unless you use the CLASS statement. You can specify
INTERTILE=0 to create contiguous tiles.
K=NORMAL | QUADRATIC | TRIANGULAR

species the kernel function (normal, quadratic, or triangular) used to compute a kernel density estimate. You can specify up to ve values to request multiple estimates.
You must enclose this option in parentheses after the KERNEL option. You can also
use the K= kernel-option with the C= kernel-option, which species standardized
bandwidths. If you specify more kernel functions than bandwidths, the procedure
repeats the last bandwidth in the list for the remaining estimates. Likewise, if you
specify more bandwidths than kernel functions, the procedure repeats the last kernel
function for the remaining estimates. By default, K=NORMAL.
KERNEL<(kernel-options)>

superimposes up to ve kernel density estimates on the histogram. By default, the

procedure uses the AMISE method to compute kernel density estimates. To request
multiple kernel density estimates on the same histogram, specify a list of values for
either the C= kernel-option or K= kernel-option. Table 3.10 (page 215) lists options
you can specify with the KERNEL option. See Example 3.23.

223

224

Chapter 3. The UNIVARIATE Procedure

L=linetype

species the line type used for tted density curves. Enclose the L= option in parentheses after the distribution option or the KERNEL option. If you use the L= option
with the KERNEL option, you can specify a list of up to ve line types for multiple kernel density estimates. See the entries for the C= and K= options for details
on specifying multiple kernel density estimates. By default, L=1, which produces a
solid line.
LGRID=linetype

species the line type for the grid when a grid displays on the histogram. By default,
LGRID=1, which produces a solid line. This option also creates a grid.
LHREF=linetype
LH=linetype

species the line type for the reference lines that you request with the HREF= option.
By default, LHREF=2, which produces a dashed line.
LOGNORMAL<(lognormal-options)>

displays a tted lognormal density curve on the histogram. The LOGNORMAL option can occur only once in a HISTOGRAM statement. The parameter must be less
than the minimum data value. Use the THETA= lognormal-option to specify . By
default, THETA=0. You can specify THETA=EST to request the maximum likelihood estimate for . Use the SIGMA= and ZETA= lognormal-options to specify
and . By default, the procedure computes maximum likelihood estimates for and
. Table 3.3 (page 214) and Table 3.7 (page 215) list options you can specify with the
LOGNORMAL option. See Example 3.22 and Example 3.24.
LOWER=value-list

species lower bounds for kernel density estimates requested with the KERNEL option. Enclose the LOWER= option in parentheses after the KERNEL option. You can
specify up to ve lower bounds for multiple kernel density estimates. If you specify
more kernel estimates than lower bounds, the last lower bound is repeated for the
remaining estimates. The default is a missing value, indicating no lower bounds for
tted kernel density curves.
LVREF=linetype
LV=linetype

species the line type for lines requested with the VREF= option. By default,
LVREF=2, which produces a dashed line.
MAXNBIN=n

species the maximum number of bins displayed in the comparative histogram. This
option is useful when the scales or ranges of the data distributions differ greatly from
cell to cell. By default, the bin size and midpoints are determined for the key cell, and
then the midpoint list is extended to accommodate the data ranges for the remaining
cells. However, if the cell scales differ considerably, the resulting number of bins
may be so great that each cell histogram is scaled into a narrow region. By using
MAXNBIN= to limit the number of bins, you can narrow the window about the data
distribution in the key cell. This option is not available unless you specify the CLASS
statement. The MAXNBIN= option is an alternative to the MAXSIGMAS= option.

HISTOGRAM Statement
MAXSIGMAS=value

limits the number of bins displayed in the comparative histogram to a range of value
standard deviations (of the data in the key cell) above and below the mean of the data
in the key cell. This option is useful when the scales or ranges of the data distributions
differ greatly from cell to cell. By default, the bin size and midpoints are determined
for the key cell, and then the midpoint list is extended to accommodate the data ranges
for the remaining cells. However, if the cell scales differ considerably, the resulting
number of bins may be so great that each cell histogram is scaled into a narrow region.
By using MAXSIGMAS= to limit the number of bins, you can narrow the window
that surrounds the data distribution in the key cell. This option is not available unless
you specify the CLASS statement.
MIDPERCENTS

requests a table listing the midpoints and percentage of observations in each histogram interval. If you specify MIDPERCENTS in parentheses after a density estimate option, the procedure displays a table that lists the midpoints, the observed
percentage of observations, and the estimated percentage of the population in each
interval (estimated from the tted distribution). See Example 3.18.
MIDPOINTS=values | KEY | UNIFORM

species how to determine the midpoints for the histogram intervals, where values
determines the width of the histogram bars as the difference between consecutive
midpoints. The procedure uses the same values for all variables.
The range of midpoints, extended at each end by half of the bar width, must cover the
range of the data. For example, if you specify
midpoints=2 to 10 by 0.5

then all of the observations should fall between 1.75 and 10.25. You must use evenly
spaced midpoints listed in increasing order.
KEY

determines the midpoints for the data in the key cell. The initial
number of midpoints is based on the number of observations in the
key cell that use the method of Terrell and Scott (1985). The procedure extends the midpoint list for the key cell in either direction
as necessary until it spans the data in the remaining cells.

UNIFORM

determines the midpoints by using all the observations as if there

were no cells. In other words, the number of midpoints is based
on the total sample size by using the method of Terrell and Scott
(1985).

Neither KEY nor UNIFORM apply unless you use the CLASS statement. By default,
if you use a CLASS statement, MIDPOINTS=KEY; however, if the key cell is empty
then MIDPOINTS=UNIFORM. Otherwise, the procedure computes the midpoints by
using an algorithm (Terrell and Scott 1985) that is primarily applicable to continuous
data that are approximately normally distributed.

225

226

Chapter 3. The UNIVARIATE Procedure

MU=value

species the parameter for normal density curves requested with the NORMAL
option. Enclose the MU= option in parentheses after the NORMAL option. By
default, the procedure uses the sample mean for .
NAME=string

species a name for the plot, up to eight characters long, that appears in the PROC
GREPLAY master menu. The default value is UNIVAR.
NCOLS=n
NCOL=n

species the number of columns in a comparative histogram. By default, NCOLS=1

if you specify only one class variable, and NCOLS=2 if you specify two class variables. This option is not available unless you use the CLASS statement. If you specify
two class variables, you can use the NCOLS= option with the NROWS= option.
NOBARS

suppresses drawing of histogram bars, which is useful for viewing tted curves only.
NOFRAME

suppresses the frame around the subplot area.

NOHLABEL

suppresses the label for the horizontal axis. You can use this option to reduce clutter.
NOPLOT
NOCHART

suppresses the creation of a plot. Use this option when you only want to tabulate
summary statistics for a tted density or create an OUTHISTOGRAM= data set.
NOPRINT

suppresses tables summarizing the tted curve. Enclose the NOPRINT option in
parentheses following the distribution option.
NORMAL<(normal-options)>

displays a tted normal density curve on the histogram. The NORMAL option can
occur only once in a HISTOGRAM statement. Use the MU= and SIGMA= normaloptions to specify and . By default, the procedure uses the sample mean and
sample standard deviation for and . Table 3.3 (page 214) and Table 3.8 (page 215)
list options you can specify with the NORMAL option. See Example 3.19.
NOVLABEL

suppresses the label for the vertical axis. You can use this option to reduce clutter.
NOVTICK

suppresses the tick marks and tick mark labels for the vertical axis. This option also
suppresses the label for the vertical axis.
NROWS=n
NROW=n

species the number of rows in a comparative histogram. By default, NROWS=2.

This option is not available unless you use the CLASS statement. If you specify two
class variables, you can use the NCOLS= option with the NROWS= option.

HISTOGRAM Statement

227

OUTHISTOGRAM=SAS-data-set
OUTHIST=SAS-data-set

creates a SAS data set that contains information about histogram intervals.
Specically, the data set contains the midpoints of the histogram intervals, the
observed percentage of observations in each interval, and the estimated percentage
of observations in each interval (estimated from each of the specied tted curves).
PERCENTS=values
PERCENT=values

species a list of percents for which quantiles calculated from the data and quantiles
estimated from the tted curve are tabulated. The percents must be between 0 and
100. Enclose the PERCENTS= option in parentheses after the curve option. The
default percents are 1, 5, 10, 25, 50, 75, 90, 95, and 99.
PFILL=pattern

species a pattern used to ll the bars of the histograms (or the areas under a tted
curve if you also specify the FILL option). See the entries for the CFILL= and FILL
options for additional details. Refer to SAS/GRAPH Software: Reference for a list of
pattern values. By default, the bars and curve areas are not lled.
RTINCLUDE

includes the right endpoint of each histogram interval in that interval. By default, the
left endpoint is included in the histogram interval.
SCALE=value

is an alias for the SIGMA= option for curves requested by the BETA,
EXPONENTIAL, GAMMA, and WEIBULL options and an alias for the ZETA=
option for curves requested by the LOGNORMAL option.
SHAPE=value

is an alias for the ALPHA= option for curves requested with the GAMMA option, an
alias for the SIGMA= option for curves requested with the LOGNORMAL option,
and an alias for the C= option for curves requested with the WEIBULL option.
SIGMA=value | EST

species the parameter for the tted density curve when you request the BETA,
EXPONENTIAL, GAMMA, LOGNORMAL, NORMAL, and WEIBULL options.
See Table 3.13 for a summary of how to use the SIGMA= option. You must enclose
this option in parentheses after the density curve option. As a beta-option, you can
specify SIGMA=EST to request a maximum likelihood estimate for .
Table 3.13. Uses of the SIGMA= Option

Distribution Keyword
BETA
EXPONENTIAL
GAMMA
WEIBULL
LOGNORMAL
NORMAL
WEIBULL

SIGMA= Species
Scale parameter
Scale parameter
Scale parameter
Scale parameter
Shape parameter
Scale parameter
Scale parameter

Default Value
1
Maximum likelihood estimate
Maximum likelihood estimate
Maximum likelihood estimate
Maximum likelihood estimate
Standard deviation
Maximum likelihood estimate

Alias
SCALE=
SCALE=
SCALE=
SCALE=
SCALE=
SHAPE=
SCALE=

228

Chapter 3. The UNIVARIATE Procedure

THETA=value | EST

species the lower threshold parameter for curves requested with the BETA,
EXPONENTIAL, GAMMA, LOGNORMAL, and WEIBULL options. Enclose the
THETA= option in parentheses after the curve option. By default, THETA=0. If you
specify THETA=EST, an estimate is computed for .
THRESHOLD= value

is an alias for the THETA= option. See the preceding entry for the THETA= option.
TURNVLABELS
TURNVLABEL

turns the characters in the vertical axis labels so that they display vertically. This
happens by default when you use a hardware font.
UPPER=value-list

species upper bounds for kernel density estimates requested with the KERNEL option. Enclose the UPPER= option in parentheses after the KERNEL option. You can
specify up to ve upper bounds for multiple kernel density estimates. If you specify
more kernel estimates than upper bounds, the last upper bound is repeated for the
remaining estimates. The default is a missing value, indicating no upper bounds for
tted kernel density curves.
VAXIS=name|value-list

species the name of an AXIS statement describing the vertical axis. Alternatively,
you can specify a value-list for the vertical axis.
VAXISLABEL=label

species a label for the vertical axis. Labels can have up to 40 characters.
VMINOR=n
VM=n

species the number of minor tick marks between each major tick mark on the vertical
axis. Minor tick marks are not labeled. The default is zero.
VOFFSET=value

species the offset, in percentage screen units, at the upper end of the vertical axis.
VREF=value-list

draws reference lines perpendicular to the vertical axis at the values specied. Also
see the CVREF=, LVREF=, and VREFCHAR= options. If a reference line is almost
completely obscured, then use the FRONTREF option to draw the reference lines in
front of the histogram bars.
VREFLABELS=label1. . . labeln
VREFLABEL=label1. . . labeln
VREFLAB=label1. . . labeln

species labels for the lines requested by the VREF= option. The number of labels
must equal the number of lines. Enclose each label in quotes. Labels can have up to
16 characters.

HISTOGRAM Statement
VREFLABPOS=n

species the horizontal position of VREFLABELS= labels. If you specify

VREFLABPOS=1, the labels are positioned at the left of the histogram. If you
specify VREFLABPOS=2, the labels are positioned at the right of the histogram. By
default, VREFLABPOS=1.
VSCALE=COUNT | PERCENT | PROPORTION

species the scale of the vertical axis for a histogram. The value COUNT requests
the data be scaled in units of the number of observations per data unit. The value
PERCENT requests the data be scaled in units of percent of observations per data
unit. The value PROPORTION requests the data be scaled in units of proportion of
observations per data unit. The default is PERCENT.
W=n

species the width, in pixels, of the tted density curve or the kernel density estimate
curve. By default, W=1. You must enclose this option in parentheses after the density
curve option or the KERNEL option. As a kernel-option, you can specify a list of up
to ve W= values.
WAXIS=n

species the line thickness, in pixels, for the axes and frame. By default, WAXIS=1.
WBARLINE=n

species the width of bar outlines. By default, WBARLINE=1.

WEIBULL<(Weibull-options)>

displays a tted Weibull density curve on the histogram. The WEIBULL option can
occur only once in a HISTOGRAM statement. The parameter must be less than
the minimum data value. Use the THETA= Weibull-option to specify . By default,
THETA=0. You can specify THETA=EST to request the maximum likelihood estimate for . Use the C= and SIGMA= Weibull-options to specify the shape parameter
c and the scale parameter . By default, the procedure computes the maximum likelihood estimates for c and . Table 3.3 (page 214) and Table 3.9 (page 215) list option
you can specify with the WEIBULL option. See Example 3.22.
PROC UNIVARIATE calculates the maximum likelihood estimate of a iteratively by
using the Newton-Raphson approximation. See also the C=, SIGMA=, and THETA=
Weibull-options.
WGRID=n

species the line thickness for the grid.

ZETA= value

species a value for the scale parameter for lognormal density curves requested
with the LOGNORMAL option. Enclose the ZETA= lognormal-option in parentheses
after the LOGNORMAL option. By default, the procedure calculates a maximum
likelihood estimate for . You can specify the SCALE= option as an alias for the
ZETA= option.

229

230

Chapter 3. The UNIVARIATE Procedure

ID Statement
ID variables ;
The ID statement species one or more variables to include in the table of extreme
observations. The corresponding values of the ID variables appear beside the n
largest and n smallest observations, where n is the value of NEXTROBS= option.
See Example 3.3.

INSET Statement
INSET keyword-list < / options > ;
The INSET statement places a box or table of summary statistics, called an inset,
directly in a high-resolution graph created with the HISTOGRAM, PROBPLOT, or
QQPLOT statement.
The INSET statement must follow the HISTOGRAM, PROBPLOT, or QQPLOT
statement that creates the plot that you want to augment. The inset appears in all
the graphs that the preceding plot statement produces.
You can use multiple INSET statements after a plot statement to add multiple insets
to a plot. See Example 3.17.
In an INSET statement, you specify one or more keywords that identify the information to display in the inset. The information is displayed in the order that you request
the keywords. Keywords can be any of the following:
statistical keywords
primary keywords
secondary keywords
The available statistical keywords are:
Table 3.14. Descriptive Statistic Keywords

CSS
CV
KURTOSIS
MAX
MEAN
MIN
MODE
N
NMISS
NOBS
RANGE
SKEWNESS
STD
STDMEAN
SUM
SUMWGT
USS
VAR

Corrected sum of squares

Coefcient of variation
Kurtosis
Largest value
Sample mean
Smallest value
Most frequent value
Sample size
Number of missing values
Number of observations
Range
Skewness
Standard deviation
Standard error of the mean
Sum of the observations
Sum of the weights
Uncorrected sum of squares
Variance

INSET Statement
Table 3.15. Percentile Statistic Keywords

P1
P5
P10
Q1
MEDIAN
Q3
P90
P95
P99
QRANGE

1st percentile
5th percentile
10th percentile
Lower quartile (25th percentile)
Median (50th percentile)
Upper quartile (75th percentile)
90th percentile
95th percentile
99th percentile
Interquartile range (Q3 - Q1)

Table 3.16. Robust Statistics Keywords

GINI
MAD
QN
SN
STD GINI
STD MAD
STD QN
STD QRANGE
STD SN

Ginis mean difference

Median absolute difference about the
median
Qn , alternative to MAD
Sn , alternative to MAD
Ginis standard deviation
MAD standard deviation
Qn standard deviation
Interquartile range standard deviation
Sn standard deviation

Table 3.17. Hypothesis Testing Keywords

MSIGN
NORMALTEST
PNORMAL
SIGNRANK
PROBM
PROBN
PROBS
PROBT
T

Sign statistic
Test statistic for normality
Probability value for the test of normality
Signed rank statistic
Probability of greater absolute value for
the sign statistic
Probability value for the test of normality
Probability value for the signed rank test
Probability value for the Students t test
Statistics for Students t test

A primary keyword enables you to specify secondary keywords in parentheses immediately after the primary keyword. Primary keywords are BETA, EXPONENTIAL,
GAMMA, LOGNORMAL, NORMAL, WEIBULL, WEIBULL2, KERNEL, and
KERNELn. If you specify a primary keyword but omit a secondary keyword, the
inset displays a colored line and the distribution name as a key for the density curve.
By default, PROC UNIVARIATE identies inset statistics with appropriate labels and
prints numeric values using appropriate formats. To customize the label, specify the
keyword followed by an equal sign (=) and the desired label in quotes. To customize
the format, specify a numeric format in parentheses after the keyword. Labels can
have up to 24 characters.

231

232

Chapter 3. The UNIVARIATE Procedure

If you specify both a label and a format for a statistic, the label must appear before
the format. For example,
inset n=Sample Size std=Std Dev (5.2);

requests customized labels for two statistics and displays the standard deviation with
a eld width of 5 and two decimal places.
The following tables list primary keywords:
Table 3.18. Parametric Density Primary Keywords

Keyword
BETA

Distribution
Beta

Plot Statement Availability

All plot statements

Keyword
KERNEL
KERNELn

Description
Displays statistics for all kernel estimates
Displays statistics for only the nth kernel density estimate
n = 1, 2, 3, 4, or 5

Table 3.20 through Table 3.28 list the secondary keywords available with primary
keywords in Table 3.18 and Table 3.19.
Table 3.20. Secondary Keywords Available with the BETA Keyword

Secondary Keyword
ALPHA
BETA
SIGMA
THETA
MEAN
STD

Alias
SHAPE1
SHAPE2
SCALE
THRESHOLD

Description
First shape parameter
Second shape parameter
Scale parameter
Lower threshold parameter
Mean of the tted distribution
Standard deviation of the tted distribution

Table 3.21. Secondary Keywords Available with the EXP Keyword

Secondary Keyword
SIGMA
THETA
MEAN
STD

Alias
SCALE
THRESHOLD

Description
Scale parameter
Threshold parameter
Mean of the tted distribution
Standard deviation of the tted distribution

INSET Statement
Table 3.22. Secondary Keywords Available with the GAMMA Keyword

234

Chapter 3. The UNIVARIATE Procedure

Table 3.28. Goodness-of-Fit Statistics for Fitted Curves

Secondary Keyword
AD
ADPVAL
CVM
CVMPVAL
KSD
KSDPVAL

Description
Anderson-Darling EDF test statistic
Anderson-Darling EDF test p-value
Cramr-von Mises EDF test statistic
Cramr-von Mises EDF test p-value
Kolmogorov-Smirnov EDF test statistic
Kolmogorov-Smirnov EDF test p-value

The inset statistics listed in Table 3.18 through Table 3.28 are not available unless
you request a plot statement and options that calculate these statistics. For example,
proc univariate data=score;
histogram final / normal;
inset mean std normal(ad adpval);
run;

The MEAN and STD keywords display the sample mean and standard deviation of
FINAL. The NORMAL keyword with the secondary keywords AD and ADPVAL
display the Anderson-Darling goodness-of-t test statistic and p-value. The statistics that are specied with the NORMAL keyword are available only because the
NORMAL option is requested in the HISTOGRAM statement.
The KERNEL or KERNELn keyword is available only if you request a kernel density
estimate in a HISTOGRAM statement. The WEIBULL2 keyword is available only
if you request a two-parameter Weibull distribution in the PROBPLOT or QQPLOT
statement.

Summary of Options
The following table lists INSET statement options, which are specied after the slash
(/) in the INSET statement. For complete descriptions, see the section Dictionary of
Options on page 235.
Table 3.29. INSET Options

CFILL=color | BLANK

Species color of inset background

CFILLH=color

Species color of header background

CFRAME=color

Species color of frame

CHEADER=color

Species color of header text

CSHADOW=color

Species color of drop shadow

CTEXT=color

Species color of inset text

DATA

Species data units for POSITION=(x, y) coordinates

DATA=SAS-data-set

Species data set for statistics in the inset table

FONT=font

Species font of text

FORMAT=format

Species format of values in inset

HEADER=quoted string

Species header text

HEIGHT=value

Species height of inset text

NOFRAME

Suppresses frame around inset

INSET Statement

235

Table 3.29. (continued)

POSITION=position

Species position of inset

REFPOINT=BR | BL | TR | TL

Species reference point of inset positioned with

POSITION=(x, y) coordinates

Dictionary of Options
The following entries provide detailed descriptions of options for the INSET statement.
To specify the same format for all the statistics in the INSET statement, use the
FORMAT= option.
To create a completely customized inset, use a DATA= data set. The data set contains
the label and the value that you want to display in the inset.
If you specify multiple kernel density estimates, you can request inset statistics for
all the estimates with the KERNEL keyword. Alternatively, you can display inset
statistics for individual curves with the KERNELn keyword, where n is the curve
number between 1 and 5.
CFILL=color | BLANK

species the color of the background. If you omit the CFILLH= option the header
background is included. By default, the background is empty, which causes items
that overlap the inset (such as curves or histogram bars) to show through the inset.
If you specify a value for CFILL= option, then overlapping items no longer show
through the inset. Use CFILL=BLANK to leave the background uncolored and to
prevent items from showing through the inset.
CFILLH=color

species the color of the header background. The default value is the CFILL= color.
CFRAME=color

species the color of the frame. The default value is the same color as the axis of the
plot.
CHEADER=color

species the color of the header text. The default value is the CTEXT= color.
CSHADOW=color

species the color of the drop shadow. By default, if a CSHADOW= option is not
specied, a drop shadow is not displayed.
CTEXT=color

species the color of the text. The default value is the same color as the other text on
the plot.
DATA

species that data coordinates are to be used in positioning the inset with the
POSITION= option. The DATA option is available only when you specify
POSITION=(x,y). You must place DATA immediately after the coordinates (x,y).

236

Chapter 3. The UNIVARIATE Procedure

DATA=SAS-data-set

requests that PROC UNIVARIATE display customized statistics from a SAS data set
in the inset table. The data set must contain two variables:
LABEL
VALUE

a character variable whose values provide labels for inset entries.

a variable that is either character or numeric and whose values provide values for inset entries.

The label and value from each observation in the data set occupy one line in the inset.
The position of the DATA= keyword in the keyword list determines the position of its
lines in the inset.
FONT=font

species the font of the text. By default, if you locate the inset in the interior of the
plot then the font is SIMPLEX. If you locate the inset in the exterior of the plot then
the font is the same as the other text on the plot.
FORMAT=format

species a format for all the values in the inset. If you specify a format for a particular
statistic, then this format overrides FORMAT= format. For more information about
SAS formats, see SAS Language Reference: Dictionary
HEADER=string

species the header text. The string cannot exceed 40 characters. By default, no
header line appears in the inset. If all the keywords that you list in the INSET statement are secondary keywords that correspond to a tted curve on a histogram, PROC
UNIVARIATE displays a default header that indicates the distribution and identies
the curve.
HEIGHT=value

species the height of the text.

NOFRAME

suppresses the frame drawn around the text.

POSITION=position
POS=position

determines the position of the inset. The position is a compass point keyword, a
margin keyword, or a pair of coordinates (x,y). You can specify coordinates in axis
percent units or axis data units. The default value is NW, which positions the inset
in the upper left (northwest) corner of the display. See the section Positioning the
Inset on page 285.
REFPOINT=BR | BL | TR | TL

species the reference point for an inset that PROC UNIVARIATE positions by a pair
of coordinates with the POSITION= option. The REFPOINT= option species which
corner of the inset frame that you want to position at coordinates (x,y). The keywords
are BL, BR, TL, and TR, which correspond to bottom left, bottom right, top left, and
top right. The default value is BL. You must use REFPOINT= with POSITION=(x,y)
coordinates.

OUTPUT Statement

OUTPUT Statement
OUTPUT < OUT=SAS-data-set >
< keyword1=names. . .keywordk=names >< percentile-options >;
The OUTPUT statement saves statistics and BY variables in an output data set. When
you use a BY statement, each observation in the OUT= data set corresponds to one
of the BY groups. Otherwise, the OUT= data set contains only one observation.
You can use any number of OUTPUT statements in the UNIVARIATE procedure.
Each OUTPUT statement creates a new data set containing the statistics specied in
that statement. You must use the VAR statement with the OUTPUT statement. The
OUTPUT statement must contain a specication of the form keyword=names or the
PCTLPTS= and PCTLPRE= specications. See Example 3.7 and Example 3.8.
OUT=SAS-data-set

identies the output data set. If SAS-data-set does not exist, PROC UNIVARIATE
creates it. If you omit OUT=, the data set is named DATAn, where n is the smallest
integer that makes the name unique. The default SAS-data-set is DATAn.
keyword=name

species the statistics to include in the output data set and gives names to the new
variables that contain the statistics. Specify a keyword for each desired statistic, an
equal sign, and the names of the variables to contain the statistic. In the output
data set, the rst variable listed after a keyword in the OUTPUT statement contains
the statistic for the rst variable listed in the VAR statement; the second variable
contains the statistic for the second variable in the VAR statement, and so on. If the
list of names following the equal sign is shorter than the list of variables in the VAR
statement, the procedure uses the names in the order in which the variables are listed
in the VAR statement. The available keywords are listed in the following tables:
Table 3.30. Descriptive Statistic Keywords

CSS
CV
KURTOSIS
MAX
MEAN
MIN
MODE
N
NMISS
NOBS
RANGE
SKEWNESS
STD
STDMEAN
SUM
SUMWGT
USS
VAR

Corrected sum of squares

237

238

Chapter 3. The UNIVARIATE Procedure

Table 3.31. Quantile Statistic Keywords

P1
P5
P10
Q1
MEDIAN
Q3
P90
P95
P99
QRANGE

Table 3.32. Robust Statistics Keywords

GINI
MAD
QN
SN
STD GINI
STD MAD
STD QN
STD QRANGE
STD SN

Ginis mean difference

Table 3.33. Hypothesis Testing Keywords

MSIGN
NORMALTEST
SIGNRANK
PROBM
PROBN
PROBS
PROBT
T

Sign statistic
Test statistic for normality
Signed rank statistic
Probability of a greater absolute value for
the sign statistic
Probability value for the test of normality
Probability value for the signed rank test
Probability value for the Students t test
Statistic for the Students t test

To store the same statistic for several analysis variables, specify a list of names. The
order of the names corresponds to the order of the analysis variables in the VAR
statement. PROC UNIVARIATE uses the rst name to create a variable that contains
the statistic for the rst analysis variable, the next name to create a variable that
contains the statistic for the second analysis variable, and so on. If you do not want
to output statistics for all the analysis variables, specify fewer names than the number
of analysis variables.
The UNIVARIATE procedure automatically computes the 1st, 5th, 10th, 25th, 50th,
75th, 90th, 95th, and 99th percentiles for the data. These can be saved in an output
data set using keyword=names specications. For additional percentiles, you can use
the following percentile-options:

OUTPUT Statement
PCTLPTS=percentiles

species one or more percentiles that are not automatically computed by the
UNIVARIATE procedure. The PCTLPRE= and PCTLPTS= options must be used
together. You can specify percentiles with the expression start TO stop BY increment
where start is a starting number, stop is an ending number, and increment is a number
to increment by. The PCTLPTS= option generates additional percentiles and outputs
them to a data set; these additional percentiles are not printed.
To compute the 50th, 95th, 97.5th, and 100th percentiles, submit the statement
output pctlpre=P_ pctlpts=50,95 to 100 by 2.5;

You can use PCTLPTS= to output percentiles that are not in the list of quantile statistics. PROC UNIVARIATE computes the requested percentiles based on the method
that you specify with the PCTLDEF= option in the PROC UNIVARIATE statement.
You must use PCTLPRE=, and optionally PCTLNAME=, to specify variable names
for the percentiles. For example, the following statements create an output data set
that is named Pctls that contains the 20th and 40th percentiles of the analysis variables PreTest and PostTest:
proc univariate data=Score;
var PreTest PostTest;
output out=Pctls pctlpts=20 40 pctlpre=PreTest_ PostTest_
pctlname=P20 P40;
run;

PROC UNIVARIATE saves the 20th and 40th percentiles for PreTest and PostTest in
the variables PreTest P20, PostTest P20, PreTest P40, and PostTest P40.
PCTLPRE=prexes

species one or more prexes to create the variable names for the variables that contain the PCTLPTS= percentiles. To save the same percentiles for more than one
analysis variable, specify a list of prexes. The order of the prexes corresponds
to the order of the analysis variables in the VAR statement. The PCTLPRE= and
PCTLPTS= options must be used together.
The procedure generates new variable names using the prex and the percentile values. If the specied percentile is an integer, the variable name is simply the prex
followed by the value. If the specied value is not an integer, an underscore replaces
the decimal point in the variable name, and decimal values are truncated to one decimal place. For example, the following statements create the variables PWID20,
PWID33 3, PWID66 6, and PWID80 for the 20th, 33.33rd, 66.67th, and 80th percentiles of Width, respectively:
proc univariate noprint;
var Width;
output pctlpts=20 33.33 66.67 80 pctlpre=pwid;
run;

239

240

Chapter 3. The UNIVARIATE Procedure

If you request percentiles for more than one variable, you should list prexes in the
same order in which the variables appear in the VAR statement. If combining the
prex and percentile value results in a name longer than 32 characters, the prex is
truncated so that the variable name is 32 characters.

PCTLNAME=sufxes

species one or more sufxes to create the names for the variables that contain the
PCTLPTS= percentiles. PROC UNIVARIATE creates a variable name by combining the PCTLPRE= value and sufx-name. Because the sufx names are associated
with the percentiles that are requested, list the sufx names in the same order as the
PCTLPTS= percentiles. If you specify n sufxes with the PCTLNAME= option and
m percentile values with the PCTLPTS= option, where m > n, the sufxes are used
to name the rst n percentiles, and the default names are used for the remaining mn
percentiles. For example, consider the following statements:
proc univariate;
var Length Width Height;
output pctlpts = 20 40
pctlpre = pl pw ph
pctlname = twenty;
run;

The value TWENTY in the PCTLNAME= option is used for only the rst percentile
in the PCTLPTS= list. This sufx is appended to the values in the PCTLPRE= option
to generate the new variable names PLTWENTY, PWTWENTY, and PHTWENTY,
which contain the 20th percentiles for Length, Width, and Height, respectively.
Since a second PCTLNAME= sufx is not specied, variable names for the 40th
percentiles for Length, Width, and Height are generated using the prexes and percentile values. Thus, the output data set contains the variables PLTWENTY, PL40,
PWTWENTY, PW40, PHTWENTY, and PH40.
You must specify PCTLPRE= to supply prex names for the variables that contain
the PCTLPTS= percentiles.
If the number of PCTLNAME= values is fewer than the number of percentiles, or
if you omit PCTLNAME=, PROC UNIVARIATE uses the percentile as the sufx to
create the name of the variable that contains the percentile. For an integer percentile,
PROC UNIVARIATE uses the percentile. Otherwise, PROC UNIVARIATE truncates
decimal values of percentiles to two decimal places and replaces the decimal point
with an underscore.
If either the prex and sufx name combination or the prex and percentile name
combination is longer than 32 characters, PROC UNIVARIATE truncates the prex
name so that the variable name is 32 characters.

PROBPLOT Statement

PROBPLOT Statement
PROBPLOT < variables >< / options >;
The PROBPLOT statement creates a probability plot, which compares ordered variable values with the percentiles of a specied theoretical distribution. If the data
distribution matches the theoretical distribution, the points on the plot form a linear pattern. Consequently, you can use a probability plot to determine how well a
theoretical distribution models a set of measurements.
Probability plots are similar to Q-Q plots, which you can create with the QQPLOT
statement. Probability plots are preferable for graphical estimation of percentiles,
whereas Q-Q plots are preferable for graphical estimation of distribution parameters.
You can use any number of PROBPLOT statements in the UNIVARIATE procedure.
The components of the PROBPLOT statement are described as follows.
variables

are the variables for which to create probability plots. If you specify a VAR statement,
the variables must also be listed in the VAR statement. Otherwise, the variables can
be any numeric variables in the input data set. If you do not specify a list of variables,
then by default the procedure creates a probability plot for each variable listed in the
VAR statement, or for each numeric variable in the DATA= data set if you do not
specify a VAR statement. For example, each of the following PROBPLOT statements
produces two probability plots, one for Length and one for Width:
proc univariate data=Measures;
var Length Width;
probplot;
proc univariate data=Measures;
probplot Length Width;
run;
options

specify the theoretical distribution for the plot or add features to the plot. If you
specify more than one variable, the options apply equally to each variable. Specify
all options after the slash (/) in the PROBPLOT statement. You can specify only
one option naming a distribution in each PROBPLOT statement, but you can specify
any number of other options. The distributions available are the beta, exponential,
gamma, lognormal, normal, two-parameter Weibull, and three-parameter Weibull.
By default, the procedure produces a plot for the normal distribution.
In the following example, the NORMAL option requests a normal probability plot
for each variable, while the MU= and SIGMA= normal-options request a distribution
reference line corresponding to the normal distribution with = 10 and = 0.3.
The SQUARE option displays the plot in a square frame, and the CTEXT= option
species the text color.
proc univariate data=Measures;
probplot Length1 Length2 / normal(mu=10 sigma=0.3)
square ctext=blue;
run;

241

242

Chapter 3. The UNIVARIATE Procedure

Table 3.34 through Table 3.43 list the PROBPLOT options by function. For complete
descriptions, see the section Dictionary of Options on page 245. Options can be
any of the following:
primary options
secondary options
general options

Distribution Options
Table 3.34 lists options for requesting a theoretical distribution.
Table 3.34. Primary Options for Theoretical Distributions
BETA(beta-options)
Species beta probability plot for

shape parameters and specied with mandatory ALPHA= and

BETA= beta-options
EXPONENTIAL(exponential-options)

Species exponential probability plot

GAMMA(gamma-options)

Species gamma probability plot for

shape parameter specied with
mandatory ALPHA= gamma-option

LOGNORMAL(lognormal-options)

Species lognormal probability plot

for shape parameter specied
with mandatory SIGMA= lognormaloption

NORMAL(normal-options)

Species normal probability plot

WEIBULL(Weibull-options)

Species three-parameter Weibull

probability plot for shape parameter c specied with mandatory C=
Weibull-option

WEIBULL2(Weibull2-options)

Species two-parameter
probability plot

Weibull

Table 3.35 through Table 3.42 list secondary options that specify distribution parameters and control the display of a distribution reference line. Specify these options in
parentheses after the distribution keyword. For example, you can request a normal
probability plot with a distribution reference line by specifying the NORMAL option
as follows:
proc univariate;
probplot Length / normal(mu=10 sigma=0.3 color=red);
run;

PROBPLOT Statement

243

Table 3.35. Secondary Reference Line Options Used with All Distributions
COLOR=color
Species color of distribution reference line

L=linetype

Species line type of distribution reference line

W=n

Species width of distribution reference line

Table 3.36. Secondary Beta-Options

ALPHA=value-list | EST
Species mandatory shape parameter

BETA=value-list | EST

Species mandatory shape parameter

SIGMA=value | EST

Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.37. Secondary Exponential-Options

SIGMA=value | EST
Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.38. Secondary Gamma-Options

ALPHA=value-list | EST
Species mandatory shape parameter

SIGMA=value | EST

Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.39. Secondary Lognormal-Options

SIGMA=value
Species mandatory shape parameter

SLOPE=value | EST

Species slope of distribution reference line

THETA=value|EST

Species 0 for distribution reference line

ZETA=value

Species 0 for distribution reference line (slope is

exp(0 ))

Table 3.40. Secondary Normal-Options

MU=value | EST
Species 0 for distribution reference line

SIGMA=value | EST

Species 0 for distribution reference line

Table 3.41. Secondary Weibull-Options

C=value-list | EST
Species mandatory shape parameter c

SIGMA=value | EST

Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.42. Secondary Weibull2-Options

C=value | EST
Species c0 for distribution reference line (slope is 1/c0 )

SIGMA=value | EST

Species 0 for distribution reference line (intercept is

log(0 ))

SLOPE=value | EST

Species slope of distribution reference line

THETA=value

Species known lower threshold 0

244

Chapter 3. The UNIVARIATE Procedure

General Graphics Options

Table 3.43 summarizes general options for enhancing probability plots.
Table 3.43. General Graphics Options

Option
ANNOKEY
ANNOTATE=
CAXIS=
CFRAME=
CFRAMESIDE=
CFRAMETOP=
CGRID=
CHREF=
CTEXT=
CVREF=
DESCRIPTION=
FONT=
GRID
HEIGHT=
HMINOR=
HREF=
HREFLABELS=
INFONT=
INHEIGHT=
INTERTILE=
LGRID=
LHREF=
LVREF=
NADJ=
NAME=
NCOLS=
NOFRAME
NOHLABEL
NOVLABEL
NOVTICK
NROWS=
PCTLMINOR
PCTLORDER=
RANKADJ=
SQUARE
VAXISLABEL=
VMINOR=
VREF=
VREFLABELS=
VREFLABPOS=
WAXIS=

Description
Applies annotation requested in ANNOTATE= data set to key cell only
Species annotate data set
Species color for axis
Species color for frame
Species color for lling frame for row labels
Species color for lling frame for column labels
Species color for grid lines
Species color for HREF= lines
Species color for text
Species color for VREF= lines
Species description for plot in graphics catalog
Species software font for text
Creates a grid
Species height of text used outside framed areas
Species number of horizontal minor tick marks
Species reference lines perpendicular to the horizontal axis
Species labels for HREF= lines
Species software font for text inside framed areas
Species height of text inside framed areas
Species distance between tiles
Species a line type for grid lines
Species line style for HREF= lines
Species line style for VREF= lines
Adjusts sample size when computing percentiles
Species name for plot in graphics catalog
Species number of columns in comparative probability plot
Suppresses frame around plotting area
Suppresses label for horizontal axis
Suppresses label for vertical axis
Suppresses tick marks and tick mark labels for vertical axis
Species number of rows in comparative probability plot
Requests minor tick marks for percentile axis
Species tick mark labels for percentile axis
Adjusts ranks when computing percentiles
Displays plot in square format
Species label for vertical axis
Species number of vertical minor tick marks
Species reference lines perpendicular to the vertical axis
Species labels for VREF= lines
Species horizontal position of labels for VREF= lines
Species line thickness for axes and frame

PROBPLOT Statement
Dictionary of Options
The following entries provide detailed descriptions of options in the PROBPLOT
statement.
ALPHA=value | EST

species the mandatory shape parameter for probability plots requested with the
BETA and GAMMA options. Enclose the ALPHA= option in parentheses after the
BETA or GAMMA options. If you specify ALPHA=EST, a maximum likelihood
estimate is computed for .
ANNOKEY

applies the annotation requested with the ANNOTATE= option to the key cell only.
By default, the procedure applies annotation to all of the cells. This option is not
available unless you use the CLASS statement. Specify the KEYLEVEL= option in
the CLASS statement to specify the key cell.
ANNOTATE=SAS-data-set
ANNO=SAS-data-set

species an input data set containing annotate variables as described in SAS/GRAPH

creates a beta probability plot for each combination of the required shape parameters
and specied by the required ALPHA= and BETA= beta-options. If you specify
ALPHA=EST and BETA=EST, the procedure creates a plot based on maximum likelihood estimates for and . You can specify the SCALE= beta-option as an alias
for the SIGMA= beta-option and the THRESHOLD= beta-option as an alias for the
THETA= beta-option. To create a plot that is based on maximum likelihood estimates
for and , specify ALPHA=EST and BETA=EST.
To obtain graphical estimates of and , specify lists of values in the ALPHA= and
BETA= beta-options, and select the combination of and that most nearly linearizes the point pattern. To assess the point pattern, you can add a diagonal distribution reference line corresponding to lower threshold parameter 0 and scale parameter
0 with the THETA= and SIGMA= beta-options. Alternatively, you can add a line
that corresponds to estimated values of 0 and 0 with the beta-options THETA=EST
and SIGMA=EST. Agreement between the reference line and the point pattern indicates that the beta distribution with parameters , , 0 , and 0 is a good t.
BETA=value | EST
B=value | EST

species the mandatory shape parameter for probability plots requested with the
BETA option. Enclose the BETA= option in parentheses after the BETA option. If
you specify BETA=EST, a maximum likelihood estimate is computed for .

245

246

Chapter 3. The UNIVARIATE Procedure

C=value | EST

species the shape parameter c for probability plots requested with the WEIBULL
and WEIBULL2 options. Enclose this option in parentheses after the WEIBULL
or WEIBULL2 option. C= is a required Weibull-option in the WEIBULL option; in
this situation, it accepts a list of values, or if you specify C=EST, a maximum likelihood estimate is computed for c. You can optionally specify C=value or C=EST
as a Weibull2-option with the WEIBULL2 option to request a distribution reference line; in this situation, you must also specify Weibull2-option SIGMA=value or
SIGMA=EST.
CAXIS=color
CAXES=color

species the color for the axes. This option overrides any COLOR= specications in
an AXIS statement. The default value is the rst color in the device color list.
CFRAME=color

species the color for the area that is enclosed by the axes and frame. The area is not
lled by default.
CFRAMESIDE=color

species the color to ll the frame area for the row labels that display along the left
side of a comparative probability plot. This color also lls the frame area for the
label of the corresponding class variable (if you associate a label with the variable).
By default, these areas are not lled. This option is not available unless you use the
CLASS statement.
CFRAMETOP=color

species the color to ll the frame area for the column labels that display across the
top of a comparative probability plot. This color also lls the frame area for the
label of the corresponding class variable (if you associate a label with the variable).
By default, these areas are not lled. This option does not apply unless you use the
CLASS statement.
CGRID=color

species the color for grid lines when a grid displays on the plot. The default color is
the rst color in the device color list. This option also produces a grid.
CHREF=color
CH=color

species the color for horizontal axis reference lines requested by the HREF= option.
The default color is the rst color in the device color list.
COLOR=color

species the color of the diagonal distribution reference line. The default color is the
rst color in the device color list. Enclose the COLOR= option in parentheses after a
distribution option keyword.

PROBPLOT Statement
CTEXT=color

species the color for tick mark values and axis labels. The default color is the color
that you specify for the CTEXT= option in the GOPTIONS statement. If you omit
the GOPTIONS statement, the default is the rst color in the device color list.
CVREF=color
CV=color

species the color for the reference lines requested by the VREF= option. The default
color is the rst color in the device color list.
DESCRIPTION=string
DES=string

species a description, up to 40 characters long, that appears in the PROC GREPLAY

master menu. The default string is the variable name.
EXPONENTIAL<(exponential-options)>
EXP<(exponential-options)>

creates an exponential probability plot. To assess the point pattern, add a diagonal distribution reference line corresponding to 0 and 0 with the THETA= and SIGMA=
exponential-options. Alternatively, you can add a line corresponding to estimated
values of the threshold parameter 0 and the scale parameter with the exponentialoptions THETA=EST and SIGMA=EST. Agreement between the reference line and
the point pattern indicates that the exponential distribution with parameters 0 and
0 is a good t. You can specify the SCALE= exponential-option as an alias for the
SIGMA= exponential-option and the THRESHOLD= exponential-option as an alias
for the THETA= exponential-option.
FONT=font

species a software font for the reference lines and axis labels. You can also specify
fonts for axis labels in an AXIS statement. The FONT= font takes precedence over
the FTEXT= font specied in the GOPTIONS statement. Hardware characters are
used by default.
GAMMA(ALPHA=value | EST <gamma-options>)

creates a gamma probability plot for each value of the shape parameter given by
the mandatory ALPHA= gamma-option. If you specify ALPHA=EST, the procedure
creates a plot based on a maximum likelihood estimate for . To obtain a graphical
estimate of , specify a list of values for the ALPHA= gamma-option, and select the
value that most nearly linearizes the point pattern. To assess the point pattern, add
a diagonal distribution reference line corresponding to 0 and 0 with the THETA=
and SIGMA= gamma-options. Alternatively, you can add a line corresponding to
estimated values of the threshold parameter 0 and the scale parameter with the
gamma-options THETA=EST and SIGMA=EST. Agreement between the reference
line and the point pattern indicates that the gamma distribution with parameters , 0
and 0 is a good t. You can specify the SCALE= gamma-option as an alias for the
SIGMA= gamma-option and the THRESHOLD= gamma-option as an alias for the
THETA= gamma-option.

247

248

Chapter 3. The UNIVARIATE Procedure

GRID

displays a grid. Grid lines are reference lines that are perpendicular to the percentile
axis at major tick marks.
HEIGHT=value

species the height, in percentage screen units, of text for axis labels, tick mark
labels, and legends. This option takes precedence over the HTEXT= option in the
GOPTIONS statement.
HMINOR=n
HM=n

species the number of minor tick marks between each major tick mark on the horizontal axis. Minor tick marks are not labeled. By default, HMINOR=0.
HREF=values

draws reference lines that are perpendicular to the horizontal axis at the values you
specify.
HREFLABELS=label1 . . . labeln
HREFLABEL=label1 . . . labeln
HREFLAB=label1 . . . labeln

species labels for the reference lines requested by the HREF= option. The number of
labels must equal the number of reference lines. Labels can have up to 16 characters.
HREFLABPOS=n

species the vertical position of HREFLABELS= labels.

If you specify
HREFLABPOS=1, the labels are positioned along the top of the plot. If you
specify HREFLABPOS=2, the labels are staggered from top to bottom of the plot.
If you specify HREFLABPOS=3, the labels are positioned along the bottom of the
plot. By default, HREFLABPOS=1.
INFONT=font

species a software font to use for text inside the framed areas of the plot. The
INFONT= option takes precedence over the FTEXT= option in the GOPTIONS statement. For a list of fonts, see SAS/GRAPH Reference.
INHEIGHT=value

species the height, in percentage screen units, of text used inside the framed areas
of the plot. By default, the height specied by the HEIGHT= option is used. If you
do not specify the HEIGHT= option, the height specied with the HTEXT= option
in the GOPTIONS statement is used.
INTERTILE=value

species the distance, in horizontal percentage screen units, between the framed areas, which are called tiles. By default, the tiles are contiguous. This option is not
available unless you use the CLASS statement.
L=linetype

species the line type for a diagonal distribution reference line. Enclose the L= option
in parentheses after a distribution option. By default, L=1, which produces a solid
line.

PROBPLOT Statement
LGRID=linetype

species the line type for the grid requested by the GRID= option. By default,
LGRID=1, which produces a solid line.
LHREF=linetype
LH=linetype

species the line type for the reference lines that you request with the HREF= option.
By default, LHREF=2, which produces a dashed line.
LOGNORMAL(SIGMA=value | EST <lognormal-options>)
LNORM(SIGMA=value | EST <lognormal-options>)

creates a lognormal probability plot for each value of the shape parameter given by
the mandatory SIGMA= lognormal-option. If you specify SIGMA=EST, the procedure creates a plot based on a maximum likelihood estimate for . To obtain a graphical estimate of , specify a list of values for the SIGMA= lognormal-option, and select
the value that most nearly linearizes the point pattern. To assess the point pattern, add
a diagonal distribution reference line corresponding to 0 and 0 with the THETA=
and ZETA= lognormal-options. Alternatively, you can add a line corresponding to
estimated values of the threshold parameter 0 and the scale parameter 0 with the
lognormal-options THETA=EST and ZETA=EST. Agreement between the reference
line and the point pattern indicates that the lognormal distribution with parameters
, 0 and 0 is a good t. You can specify the THRESHOLD= lognormal-option as
an alias for the THETA= lognormal-option and the SCALE= lognormal-option as an
alias for the ZETA= lognormal-option. See Example 3.26.
LVREF=linetype

species the line type for the reference lines requested with the VREF= option. By
default, LVREF=2, which produces a dashed line.
MU=value | EST

species the mean 0 for a normal probability plot requested with the NORMAL
option. Enclose the MU= normal-option in parentheses after the NORMAL option.
The MU= normal-option must be specied with the SIGMA= normal-option, and
they request a distribution reference line. You can specify MU=EST to request a
distribution reference line with 0 equal to the sample mean.
NADJ=value

species the adjustment value added to the sample size in the calculation of theoretical percentiles. By default, NADJ= 1 . Refer to Chambers et al. (1983).
4
NAME=string

species a name for the plot, up to eight characters long, that appears in the PROC
GREPLAY master menu. The default value is UNIVAR.
NCOLS=n
NCOL=n

species the number of columns in a comparative probability plot. By default,

NCOLS=1 if you specify only one class variable, and NCOLS=2 if you specify two
class variables. This option is not available unless you use the CLASS statement. If
you specify two class variables, you can use the NCOLS= option with the NROWS=
option.

249

250

Chapter 3. The UNIVARIATE Procedure

NOFRAME

suppresses the frame around the subplot area.

NOHLABEL

suppresses the label for the horizontal axis. You can use this option to reduce clutter.
NORMAL<(normal-options)>

creates a normal probability plot. This is the default if you omit a distribution option.
To assess the point pattern, you can add a diagonal distribution reference line corresponding to 0 and 0 with the MU= and SIGMA= normal-options. Alternatively,
you can add a line corresponding to estimated values of 0 and 0 with the normaloptions MU=EST and SIGMA=EST; the estimates of the mean 0 and the standard
deviation 0 are the sample mean and sample standard deviation. Agreement between the reference line and the point pattern indicates that the normal distribution
with parameters 0 and 0 is a good t.
NOVLABEL

suppresses the label for the vertical axis. You can use this option to reduce clutter.
NOVTICK

suppresses the tick marks and tick mark labels for the vertical axis. This option also
suppresses the label for the vertical axis.
NROWS=n
NROW=n

species the number of rows in a comparative probability plot. By default,

NROWS=2. This option is not available unless you use the CLASS statement. If
you specify two class variables, you can use the NCOLS= option with the NROWS=
option.
PCTLMINOR

requests minor tick marks for the percentile axis. The HMINOR option overrides the
minor tick marks requested by the PCTLMINOR option.
PCTLORDER=values

species the tick marks that are labeled on the theoretical percentile axis. Since
the values are percentiles, the labels must be between 0 and 100, exclusive. The
values must be listed in increasing order and must cover the plotted percentile range.
Otherwise, the default values of 1, 5, 10, 25, 50, 75, 90, 95, and 99 are used.
RANKADJ=value

species the adjustment value added to the ranks in the calculation of theoretical
3
percentiles. By default, RANKADJ= 8 , as recommended by Blom (1958). Refer to
Chambers et al. (1983) for additional information.
SCALE=value | EST

is an alias for the SIGMA= option for plots requested by the BETA, EXPONENTIAL,
GAMMA, and WEIBULL options and for the ZETA= option when you request the
LOGNORMAL option. See the entries for the SIGMA= and ZETA= options.

PROBPLOT Statement

251

SHAPE=value | EST

is an alias for the ALPHA= option with the GAMMA option, for the SIGMA= option with the LOGNORMAL option, and for the C= option with the WEIBULL and
WEIBULL2 options. See the entries for the ALPHA=, SIGMA=, and C= options.
SIGMA=value | EST

species the parameter , where > 0. Alternatively, you can specify SIGMA=EST
to request a maximum likelihood estimate for 0 . The interpretation and use of the
SIGMA= option depend on the distribution option with which it is used. See Table
3.44 for a summary of how to use the SIGMA= option. You must enclose this option
in parentheses after the distribution option.
Table 3.44. Uses of the SIGMA= Option

Distribution Option

Use of the SIGMA= Option

BETA
EXPONENTIAL
GAMMA
WEIBULL
LOGNORMAL

THETA=0 and SIGMA=0 request a distribution reference

line corresponding to 0 and 0 .

NORMAL

MU=0 and SIGMA=0 request a distribution reference line

corresponding to 0 and 0 . SIGMA=EST requests a line with
0 equal to the sample standard deviation.

WEIBULL2

SIGMA=0 and C=c0 request a distribution reference line corresponding to 0 and c0 .

SIGMA=1 . . . n requests n probability plots with shape parameters 1 . . . n . The SIGMA= option must be specied.

SLOPE=value | EST

species the slope for a distribution reference line requested with the LOGNORMAL
and WEIBULL2 options. Enclose the SLOPE= option in parentheses after the distribution option. When you use the SLOPE= lognormal-option with the LOGNORMAL
option, you must also specify a threshold parameter value 0 with the THETA=
lognormal-option to request the line. The SLOPE= lognormal-option is an alternative
to the ZETA= lognormal-option for specifying 0 , since the slope is equal to exp(0 ).
When you use the SLOPE= Weibull2-option with the WEIBULL2 option, you must
also specify a scale parameter value 0 with the SIGMA= Weibull2-option to request
the line. The SLOPE= Weibull2-option is an alternative to the C= Weibull2-option for
specifying c0 , since the slope is equal to c1 .
0
For example, the rst and second PROBPLOT statements produce the same probability plots and the third and fourth PROBPLOT statements produce the same probability plots:
proc univariate data=Measures;
probplot Width / lognormal(sigma=2 theta=0 zeta=0);
probplot Width / lognormal(sigma=2 theta=0 slope=1);
probplot Width / weibull2(sigma=2 theta=0 c=.25);
probplot Width / weibull2(sigma=2 theta=0 slope=4);
run;

252

Chapter 3. The UNIVARIATE Procedure

SQUARE

displays the probability plot in a square frame. By default, the plot is in a rectangular
frame.
THETA=value | EST

species the lower threshold parameter for plots requested with the BETA,
EXPONENTIAL, GAMMA, LOGNORMAL, WEIBULL, and WEIBULL2 options.
Enclose the THETA= option in parentheses after a distribution option. When used
with the WEIBULL2 option, the THETA= option species the known lower threshold 0 , for which the default is 0. When used with the other distribution options, the
THETA= option species 0 for a distribution reference line; alternatively in this situation, you can specify THETA=EST to request a maximum likelihood estimate for
0 . To request the line, you must also specify a scale parameter.
THRESHOLD=value | EST

is an alias for the THETA= option.

VAXISLABEL=label

species a label for the vertical axis. Labels can have up to 40 characters.
VMINOR=n
VM=n

species the number of minor tick marks between each major tick mark on the vertical
axis. Minor tick marks are not labeled. The default is zero.
VREF=values

draws reference lines perpendicular to the vertical axis at the values specied. Also
see the CVREF=, LVREF=, and VREFCHAR= options.
VREFLABELS=label1. . . labeln
VREFLABEL=label1. . . labeln
VREFLAB=label1. . . labeln

species labels for the reference lines requested by the VREF= option. The number
of labels must equal the number of reference lines. Enclose each label in quotes.
Labels can have up to 16 characters.
VREFLABPOS=n

species the horizontal position of VREFLABELS= labels. If you specify

VREFLABPOS=1, the labels are positioned at the left of the histogram. If you
specify VREFLABPOS=2, the labels are positioned at the right of the histogram. By
default, VREFLABPOS=1.
W=n

species the width, in pixels, for a diagonal distribution line. Enclose the W= option
in parentheses after the distribution option. By default, W=1.
WAXIS=n

species the line thickness, in pixels, for the axes and frame. By default, WAXIS=1.

QQPLOT Statement
WEIBULL(C=value | EST <Weibull-options>)
WEIB(C=value | EST <Weibull-options>)

creates a three-parameter Weibull probability plot for each value of the required shape
parameter c specied by the mandatory C= Weibull-option. To create a plot that is
based on a maximum likelihood estimate for c, specify C=EST. To obtain a graphical
estimate of c, specify a list of values in the C= Weibull-option, and select the value
that most nearly linearizes the point pattern. To assess the point pattern, add a diagonal distribution reference line corresponding to 0 and 0 with the THETA= and
SIGMA= Weibull-options. Alternatively, you can add a line corresponding to estimated values of 0 and 0 with the Weibull-options THETA=EST and SIGMA=EST.
Agreement between the reference line and the point pattern indicates that the Weibull
distribution with parameters c, 0 , and 0 is a good t. You can specify the SCALE=
Weibull-option as an alias for the SIGMA= Weibull-option and the THRESHOLD=
Weibull-option as an alias for the THETA= Weibull-option.
WEIBULL2<(Weibull2-options)>
W2<(Weibull2-options)>

creates a two-parameter Weibull probability plot. You should use the WEIBULL2 option when your data have a known lower threshold 0 , which is 0 by default. To specify the threshold value 0 , use the THETA= Weibull2-option. By default, THETA=0.
An advantage of the two-parameter Weibull plot over the three-parameter Weibull
plot is that the parameters c and can be estimated from the slope and intercept of
the point pattern. A disadvantage is that the two-parameter Weibull distribution applies only in situations where the threshold parameter is known. To obtain a graphical
estimate of 0 , specify a list of values for the THETA= Weibull2-option, and select
the value that most nearly linearizes the point pattern. To assess the point pattern, add
a diagonal distribution reference line corresponding to 0 and c0 with the SIGMA=
and C= Weibull2-options. Alternatively, you can add a distribution reference line corresponding to estimated values of 0 and c0 with the Weibull2-options SIGMA=EST
and C=EST. Agreement between the reference line and the point pattern indicates that
the Weibull distribution with parameters c0 , 0 , and 0 is a good t. You can specify
the SCALE= Weibull2-option as an alias for the SIGMA= Weibull2-option and the
SHAPE= Weibull2-option as an alias for the C= Weibull2-option.
ZETA=value | EST

species a value for the scale parameter for the lognormal probability plots requested with the LOGNORMAL option. Enclose the ZETA= lognormal-option in
parentheses after the LOGNORMAL option. To request a distribution reference line
with intercept 0 and slope exp(0 ), specify the THETA=0 and ZETA=0 .

QQPLOT Statement
QQPLOT < variables >< / options >;
The QQPLOT statement creates quantile-quantile plots (Q-Q plots) using highresolution graphics and compares ordered variable values with quantiles of a specied
theoretical distribution. If the data distribution matches the theoretical distribution,
the points on the plot form a linear pattern. Thus, you can use a Q-Q plot to determine
how well a theoretical distribution models a set of measurements.

253

254

Chapter 3. The UNIVARIATE Procedure

Q-Q plots are similar to probability plots, which you can create with the PROBPLOT
statement. Q-Q plots are preferable for graphical estimation of distribution parameters, whereas probability plots are preferable for graphical estimation of percentiles.
You can use any number of QQPLOT statements in the UNIVARIATE procedure.
The components of the QQPLOT statement are described as follows.

variables

are the variables for which to create Q-Q plots. If you specify a VAR statement, the
variables must also be listed in the VAR statement. Otherwise, the variables can be
any numeric variables in the input data set. If you do not specify a list of variables,
then by default the procedure creates a Q-Q plot for each variable listed in the VAR
statement, or for each numeric variable in the DATA= data set if you do not specify
a VAR statement. For example, each of the following QQPLOT statements produces
two Q-Q plots, one for Length and one for Width:
proc univariate data=Measures;
var Length Width;
qqplot;
proc univariate data=Measures;
qqplot Length Width;
run;
options

specify the theoretical distribution for the plot or add features to the plot. If you specify more than one variable, the options apply equally to each variable. Specify all
options after the slash (/) in the QQPLOT statement. You can specify only one option
naming the distribution in each QQPLOT statement, but you can specify any number of other options. The distributions available are the beta, exponential, gamma,
lognormal, normal, two-parameter Weibull, and three-parameter Weibull. By default,
the procedure produces a plot for the normal distribution.
In the following example, the NORMAL option requests a normal Q-Q plot for each
variable. The MU= and SIGMA= normal-options request a distribution reference line
with intercept 10 and slope 0.3 for each plot, corresponding to a normal distribution
with mean = 10 and standard deviation = 0.3. The SQUARE option displays
the plot in a square frame, and the CTEXT= option species the text color.
proc univariate data=measures;
qqplot length1 length2 / normal(mu=10 sigma=0.3)
square ctext=blue;
run;

Table 3.45 through Table 3.54 list the QQPLOT options by function. For complete
descriptions, see the section Dictionary of Options on page 258.
Options can be any of the following:

primary options
secondary options
general options

QQPLOT Statement
Distribution Options
Table 3.45 lists primary options for requesting a theoretical distribution.
Table 3.45. Primary Options for Theoretical Distributions
BETA(beta-options)
Species beta Q-Q plot for shape

parameters and specied with

mandatory ALPHA= and BETA=
beta-options

EXPONENTIAL(exponential-options)

Species exponential Q-Q plot

GAMMA(gamma-options)

Species gamma Q-Q plot for shape

parameter specied with mandatory ALPHA= gamma-option

LOGNORMAL(lognormal-options)

Species lognormal Q-Q plot for

shape parameter specied with
mandatory SIGMA= lognormaloption

NORMAL(normal-options)

Species normal Q-Q plot

WEIBULL(Weibull-options)

Species three-parameter Weibull QQ plot for shape parameter c specied

with mandatory C= Weibull-option

WEIBULL2(Weibull2-options)

Species two-parameter Weibull QQ plot

Table 3.46 through Table 3.53 list secondary options that specify distribution parameters and control the display of a distribution reference line. Specify these options in
parentheses after the distribution keyword. For example, you can request a normal
Q-Q plot with a distribution reference line by specifying the NORMAL option as
follows:
proc univariate;
qqplot Length / normal(mu=10 sigma=0.3 color=red);
run;

The MU= and SIGMA= normal-options display a distribution reference line that
corresponds to the normal distribution with mean 0 = 10 and standard deviation
0 = 0.3, and the COLOR= normal-option species the color for the line.
Table 3.46. Secondary Reference Line Options Used with All Distributions
COLOR=color
Species color of distribution reference line

L=linetype

Species line type of distribution reference line

W=n

Species width of distribution reference line

255

256

Chapter 3. The UNIVARIATE Procedure

Table 3.47. Secondary Beta-Options
ALPHA=value-list | EST
Species mandatory shape parameter

BETA=value-list | EST

Species mandatory shape parameter

SIGMA=value | EST

Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.48. Secondary Exponential-Options

SIGMA=value | EST
Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.49. Secondary Gamma-Options

ALPHA=value-list | EST
Species mandatory shape parameter

SIGMA=value | EST

Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.50. Secondary Lognormal-Options

SIGMA=value-list | EST
Species mandatory shape parameter

SLOPE=value | EST

Species slope of distribution reference line

THETA=value|EST

Species 0 for distribution reference line

ZETA=value

Species 0 for distribution reference line (slope is

exp(0 ))

Table 3.51. Secondary Normal-Options

MU=value | EST
Species 0 for distribution reference line

SIGMA=value | EST

Species 0 for distribution reference line

Table 3.52. Secondary Weibull-Options

C=value-list | EST
Species mandatory shape parameter c

SIGMA=value | EST

Species 0 for distribution reference line

THETA=value | EST

Species 0 for distribution reference line

Table 3.53. Secondary Weibull2-Options

C=value | EST
Species c0 for distribution reference line (slope is 1/c0 )

SIGMA=value | EST

Species 0 for distribution reference line (intercept is

log(0 ))

SLOPE=value | EST

Species slope of distribution reference line

THETA=value

Species known lower threshold 0

General Options
Table 3.54 summarizes general options for enhancing Q-Q plots.
Table 3.54. General Graphics Options

Option
ANNOKEY
ANNOTATE=

Description
Applies annotation requested in ANNOTATE= data set to key cell only
Species annotate data set

QQPLOT Statement
Table 3.54. (continued)

Option
CAXIS=
CFRAME=
CFRAMESIDE=
CFRAMETOP=
CGRID=
CHREF=
CTEXT=
CVREF=
DESCRIPTION=
FONT=
GRID
HEIGHT=
HMINOR=
HREF=
HREFLABELS=
HREFLABPOS=
INFONT=
INHEIGHT=
INTERTILE=
LGRID=
LHREF=
LVREF=
NADJ=
NAME=
NCOLS=
NOFRAME
NOHLABEL
NOVLABEL
NOVTICK
NROWS=
PCTLAXIS
PCTLMINOR
PCTLSCALE
RANKADJ=
SQUARE
VAXISLABEL=
VMINOR=
VREF=
VREFLABELS=
VREFLABPOS=
WAXIS=

Description
Species color for axis
Species color for frame
Species color for lling frame for row labels
Species color for lling frame for column labels
Species color for grid lines
Species color for HREF= lines
Species color for text
Species color for VREF= lines
Species description for plot in graphics catalog
Species software font for text
Creates a grid
Species height of text used outside framed areas
Species number of horizontal minor tick marks
Species reference lines perpendicular to the horizontal axis
Species labels for HREF= lines
Species vertical position of labels for HREF= lines
Species software font for text inside framed areas
Species height of text inside framed areas
Species distance between tiles
Species a line type for grid lines
Species line style for HREF= lines
Species line style for VREF= lines
Adjusts sample size when computing percentiles
Species name for plot in graphics catalog
Species number of columns in comparative Q-Q plot
Suppresses frame around plotting area
Suppresses label for horizontal axis
Suppresses label for vertical axis
Suppresses tick marks and tick mark labels for vertical axis
Species number of rows in comparative Q-Q plot
Displays a nonlinear percentile axis
Requests minor tick marks for percentile axis
Replaces theoretical quantiles with percentiles
Adjusts ranks when computing percentiles
Displays plot in square format
Species label for vertical axis
Species number of vertical minor tick marks
Species reference lines perpendicular to the vertical axis
Species labels for VREF= lines
Species horizontal position of labels for VREF= lines
Species line thickness for axes and frame

257

258

Chapter 3. The UNIVARIATE Procedure

Dictionary of Options
The following entries provide detailed descriptions of options in the QQPLOT statement.
ALPHA=value | EST

species the mandatory shape parameter for quantile plots requested with the BETA
and GAMMA options. Enclose the ALPHA= option in parentheses after the BETA
or GAMMA options. If you specify ALPHA=EST, a maximum likelihood estimate
is computed for .
ANNOKEY

applies the annotation requested with the ANNOTATE= option to the key cell only.
By default, the procedure applies annotation to all of the cells. This option is not
available unless you use the CLASS statement. Specify the KEYLEVEL= option in
the CLASS statement to specify the key cell.
ANNOTATE=SAS-data-set
ANNO=SAS-data-set

species an input data set containing annotate variables as described in SAS/GRAPH

creates a beta quantile plot for each combination of the required shape parameters
and specied by the required ALPHA= and BETA= beta-options. If you specify
ALPHA=EST and BETA=EST, the procedure creates a plot based on maximum likelihood estimates for and . You can specify the SCALE= beta-option as an alias
for the SIGMA= beta-option and the THRESHOLD= beta-option as an alias for the
THETA= beta-option. To create a plot that is based on maximum likelihood estimates
for and , specify ALPHA=EST and BETA=EST.
To obtain graphical estimates of and , specify lists of values in the ALPHA= and
BETA= beta-options, and select the combination of and that most nearly linearizes the point pattern. To assess the point pattern, you can add a diagonal distribution reference line corresponding to lower threshold parameter 0 and scale parameter
0 with the THETA= and SIGMA= beta-options. Alternatively, you can add a line
that corresponds to estimated values of 0 and 0 with the beta-options THETA=EST
and SIGMA=EST. Agreement between the reference line and the point pattern indicates that the beta distribution with parameters , , 0 , and 0 is a good t.
BETA=value | EST
B=value | EST

species the mandatory shape parameter for quantile plots requested with the BETA
option. Enclose the BETA= option in parentheses after the BETA option. If you
specify BETA=EST, a maximum likelihood estimate is computed for .

QQPLOT Statement
C=value | EST

species the shape parameter c for quantile plots requested with the WEIBULL and
WEIBULL2 options. Enclose this option in parentheses after the WEIBULL or
WEIBULL2 option. C= is a required Weibull-option in the WEIBULL option; in
this situation, it accepts a list of values, or if you specify C=EST, a maximum likelihood estimate is computed for c. You can optionally specify C=value or C=EST
as a Weibull2-option with the WEIBULL2 option to request a distribution reference line; in this situation, you must also specify Weibull2-option SIGMA=value or
SIGMA=EST.
CAXIS=color
CAXES=color

species the color for the axes. This option overrides any COLOR= specications in
an AXIS statement. The default value is the rst color in the device color list.
CFRAME=color

species the color for the area that is enclosed by the axes and frame. The area is not
lled by default.
CFRAMESIDE=color

species the color to ll the frame area for the row labels that display along the left
side of a comparative quantile plot. This color also lls the frame area for the label
of the corresponding class variable (if you associate a label with the variable). By
default, these areas are not lled. This option is not available unless you use the
CLASS statement.
CFRAMETOP=color

species the color to ll the frame area for the column labels that display across the
top of a comparative quantile plot. This color also lls the frame area for the label
of the corresponding class variable (if you associate a label with the variable). By
default, these areas are not lled. This option does not apply unless you use the
CLASS statement.
CGRID=color

species the color for grid lines when a grid displays on the plot. The default color is
the rst color in the device color list. This option also produces a grid.
CHREF=color
CH=color

species the color for horizontal axis reference lines requested by the HREF= option.
The default color is the rst color in the device color list.
COLOR=color

259

260

Chapter 3. The UNIVARIATE Procedure

CVREF=color
CV=color

species the color for the reference lines requested by the VREF= option. The default
color is the rst color in the device color list.
DESCRIPTION=string
DES=string

species a description, up to 40 characters long, that appears in the PROC GREPLAY

master menu. The default string is the variable name.
EXPONENTIAL<(exponential-options)>
EXP<(exponential-options)>

creates an exponential quantile plot. To assess the point pattern, add a diagonal distribution reference line corresponding to 0 and 0 with the THETA= and SIGMA=
exponential-options. Alternatively, you can add a line corresponding to estimated
values of the threshold parameter 0 and the scale parameter with the exponentialoptions THETA=EST and SIGMA=EST. Agreement between the reference line and
the point pattern indicates that the exponential distribution with parameters 0 and
0 is a good t. You can specify the SCALE= exponential-option as an alias for the
SIGMA= exponential-option and the THRESHOLD= exponential-option as an alias
for the THETA= exponential-option.
FONT=font

creates a gamma quantile plot for each value of the shape parameter given by the
mandatory ALPHA= gamma-option. If you specify ALPHA=EST, the procedure
creates a plot based on a maximum likelihood estimate for . To obtain a graphical
estimate of , specify a list of values for the ALPHA= gamma-option, and select the
value that most nearly linearizes the point pattern. To assess the point pattern, add
a diagonal distribution reference line corresponding to 0 and 0 with the THETA=
and SIGMA= gamma-options. Alternatively, you can add a line corresponding to
estimated values of the threshold parameter 0 and the scale parameter with the
gamma-options THETA=EST and SIGMA=EST. Agreement between the reference
line and the point pattern indicates that the gamma distribution with parameters , 0
and 0 is a good t. You can specify the SCALE= gamma-option as an alias for the
SIGMA= gamma-option and the THRESHOLD= gamma-option as an alias for the
THETA= gamma-option.
GRID

displays a grid of horizontal lines positioned at major tick marks on the vertical axis.
HEIGHT=value

species the height, in percentage screen units, of text for axis labels, tick mark
labels, and legends. This option takes precedence over the HTEXT= option in the
GOPTIONS statement.

QQPLOT Statement
HMINOR=n
HM=n

species the number of minor tick marks between each major tick mark on the horizontal axis. Minor tick marks are not labeled. By default, HMINOR=0.
HREF=values

draws reference lines that are perpendicular to the horizontal axis at specied values.
When you use the PCTLAXIS option, HREF= values must be in quantile units.
HREFLABELS=label1 . . . labeln
HREFLABEL=label1 . . . labeln
HREFLAB=label1 . . . labeln

species labels for the reference lines requested by the HREF= option. The number of
labels must equal the number of reference lines. Labels can have up to 16 characters.
HREFLABPOS=n

species the vertical position of HREFLABELS= labels.

species the line type for a diagonal distribution reference line. Enclose the L= option
in parentheses after a distribution option. By default, L=1, which produces a solid
line.
LGRID=linetype

species the line type for the grid requested by the GRID option. By default,
LGRID=1, which produces a solid line. The LGRID= option also produces a grid.
LHREF=linetype
LH=linetype

species the line type for the reference lines that you request with the HREF= option.
By default, LHREF=2, which produces a dashed line.

261

262

Chapter 3. The UNIVARIATE Procedure

LOGNORMAL(SIGMA=value | EST <lognormal-options>)

LNORM(SIGMA=value | EST <lognormal-options>)

creates a lognormal quantile plot for each value of the shape parameter given by the
mandatory SIGMA= lognormal-option. If you specify SIGMA=EST, the procedure
creates a plot based on a maximum likelihood estimate for . To obtain a graphical
estimate of , specify a list of values for the SIGMA= lognormal-option, and select
the value that most nearly linearizes the point pattern. To assess the point pattern, add
a diagonal distribution reference line corresponding to 0 and 0 with the THETA=
and ZETA= lognormal-options. Alternatively, you can add a line corresponding to
estimated values of the threshold parameter 0 and the scale parameter 0 with the
lognormal-options THETA=EST and ZETA=EST. Agreement between the reference
line and the point pattern indicates that the lognormal distribution with parameters
, 0 and 0 is a good t. You can specify the THRESHOLD= lognormal-option as
an alias for the THETA= lognormal-option and the SCALE= lognormal-option as an
alias for the ZETA= lognormal-option. See Example 3.31 through Example 3.33.
LVREF=linetype

species the line type for the reference lines requested with the VREF= option. By
default, LVREF=2, which produces a dashed line.
MU=value | EST

species the mean 0 for a normal quantile plot requested with the NORMAL option. Enclose the MU= normal-option in parentheses after the NORMAL option. The
MU= normal-option must be specied with the SIGMA= normal-option, and they request a distribution reference line. You can specify MU=EST to request a distribution
reference line with 0 equal to the sample mean.
NADJ=value

species the adjustment value added to the sample size in the calculation of theoreti1
cal percentiles. By default, NADJ= 4 . Refer to Chambers et al. (1983) for additional
information.
NAME=string

species a name for the plot, up to eight characters long, that appears in the PROC
GREPLAY master menu. The default value is UNIVAR.
NCOLS=n
NCOL=n

species the number of columns in a comparative quantile plot. By default,

suppresses the frame around the subplot area. If you specify the PCTLAXIS option,
then you cannot specify the NOFRAME option.
NOHLABEL

suppresses the label for the horizontal axis. You can use this option to reduce clutter.

QQPLOT Statement
NORMAL<(normal-options)>

creates a normal quantile plot. This is the default if you omit a distribution option.
To assess the point pattern, you can add a diagonal distribution reference line corresponding to 0 and 0 with the MU= and SIGMA= normal-options. Alternatively,
you can add a line corresponding to estimated values of 0 and 0 with the normaloptions MU=EST and SIGMA=EST; the estimates of the mean 0 and the standard
deviation 0 are the sample mean and sample standard deviation. Agreement between the reference line and the point pattern indicates that the normal distribution
with parameters 0 and 0 is a good t. See Example 3.28 and Example 3.30.
NOVLABEL

suppresses the label for the vertical axis. You can use this option to reduce clutter.
NOVTICK

suppresses the tick marks and tick mark labels for the vertical axis. This option also
suppresses the label for the vertical axis.
NROWS=n
NROW=n

species the number of rows in a comparative quantile plot. By default, NROWS=2.

This option is not available unless you use the CLASS statement. If you specify two
class variables, you can use the NCOLS= option with the NROWS= option.
PCTLAXIS<(axis-options)>

adds a nonlinear percentile axis along the frame of the Q-Q plot opposite the theoretical quantile axis. The added axis is identical to the axis for probability plots produced
with the PROBPLOT statement. When using the PCTLAXIS option, you must specify HREF= values in quantile units, and you cannot use the NOFRAME option. You
can specify the following axis-options:
Table 3.55. Axis Options

GRID

Draws vertical grid lines at major percentiles

GRIDCHAR=character

Species grid line plotting character on line printer

LABEL=string

Species label for percentile axis

LGRID=linetype

Species line type for grid

PCTLMINOR

requests minor tick marks for the percentile axis when you specify PCTLAXIS. The
HMINOR option overrides the PCTLMINOR option.
PCTLSCALE

requests scale labels for the theoretical quantile axis in percentile units, resulting in
a nonlinear axis scale. Tick marks are drawn uniformly across the axis based on the
quantile scale. In all other respects, the plot remains the same, and you must specify
HREF= values in quantile units. For a true nonlinear axis, use the PCTLAXIS option
or use the PROBPLOT statement.
RANKADJ=value

263

264

Chapter 3. The UNIVARIATE Procedure

SCALE=value | EST

is an alias for the SIGMA= option for plots requested by the BETA, EXPONENTIAL,
GAMMA, WEIBULL, and WEIBULL2 options and for the ZETA= option with the
LOGNORMAL option. See the entries for the SIGMA= and ZETA= options.
SHAPE=value | EST

species the parameter , where > 0. Alternatively, you can specify SIGMA=EST
to request a maximum likelihood estimate for 0 . The interpretation and use of the
SIGMA= option depend on the distribution option with which it is used, as summarized in Table 3.56. Enclose this option in parentheses after the distribution option.
Table 3.56. Uses of the SIGMA= Option

Distribution Option

Use of the SIGMA= Option

BETA
EXPONENTIAL
GAMMA
WEIBULL
LOGNORMAL

THETA=0 and SIGMA=0 request a distribution reference

line corresponding to 0 and 0 .

NORMAL

MU=0 and SIGMA=0 request a distribution reference line

corresponding to 0 and 0 . SIGMA=EST requests a line with
0 equal to the sample standard deviation.

WEIBULL2

SIGMA=0 and C=c0 request a distribution reference line corresponding to 0 and c0 .

SIGMA=1 . . . n requests n quantile plots with shape parameters 1 . . . n . The SIGMA= option must be specied.

SLOPE=value | EST

species the slope for a distribution reference line requested with the LOGNORMAL
and WEIBULL2 options. Enclose the SLOPE= option in parentheses after the distribution option. When you use the SLOPE= lognormal-option with the LOGNORMAL
option, you must also specify a threshold parameter value 0 with the THETA=
lognormal-option to request the line. The SLOPE= lognormal-option is an alternative
to the ZETA= lognormal-option for specifying 0 , since the slope is equal to exp(0 ).
When you use the SLOPE= Weibull2-option with the WEIBULL2 option, you must
also specify a scale parameter value 0 with the SIGMA= Weibull2-option to request
the line. The SLOPE= Weibull2-option is an alternative to the C= Weibull2-option for
specifying c0 , since the slope is equal to c1 .
0
For example, the rst and second QQPLOT statements produce the same quantile
plots and the third and fourth QQPLOT statements produce the same quantile plots:
proc univariate
qqplot Width
qqplot Width
qqplot Width
qqplot Width

data=Measures;
/ lognormal(sigma=2 theta=0 zeta=0);
/ lognormal(sigma=2 theta=0 slope=1);
/ weibull2(sigma=2 theta=0 c=.25);
/ weibull2(sigma=2 theta=0 slope=4);

QQPLOT Statement
SQUARE

displays the quantile plot in a square frame. By default, the frame is rectangular.
THETA=value | EST

is an alias for the THETA= option.

VAXISLABEL=label

species a label for the vertical axis. Labels can have up to 40 characters.
VMINOR=n
VM=n

species the number of minor tick marks between each major tick mark on the vertical
axis. Minor tick marks are not labeled. The default is zero.
VREF=values

species the horizontal position of VREFLABELS= labels. If you specify

VREFLABPOS=1, the labels are positioned at the left of the histogram. If you
specify VREFLABPOS=2, the labels are positioned at the right of the histogram. By
default, VREFLABPOS=1.
W=n

species the width, in pixels, for a diagonal distribution line. Enclose the W= option
in parentheses after the distribution option. By default, W=1.
WAXIS=n

species the line thickness, in pixels, for the axes and frame. By default, WAXIS=1.

265

266

Chapter 3. The UNIVARIATE Procedure

WEIBULL(C=value | EST <Weibull-options>)

WEIB(C=value | EST <Weibull-options>)

creates a three-parameter Weibull quantile plot for each value of the required shape
parameter c specied by the mandatory C= Weibull-option. To create a plot that is
based on a maximum likelihood estimate for c, specify C=EST. To obtain a graphical
estimate of c, specify a list of values in the C= Weibull-option, and select the value
that most nearly linearizes the point pattern. To assess the point pattern, add a diagonal distribution reference line corresponding to 0 and 0 with the THETA= and
SIGMA= Weibull-options. Alternatively, you can add a line corresponding to estimated values of 0 and 0 with the Weibull-options THETA=EST and SIGMA=EST.
Agreement between the reference line and the point pattern indicates that the Weibull
distribution with parameters c, 0 , and 0 is a good t. You can specify the SCALE=
Weibull-option as an alias for the SIGMA= Weibull-option and the THRESHOLD=
Weibull-option as an alias for the THETA= Weibull-option. See Example 3.34.
WEIBULL2<(Weibull2-options)>
W2<(Weibull2-options)>

creates a two-parameter Weibull quantile plot. You should use the WEIBULL2 option
when your data have a known lower threshold 0 , which is 0 by default. To specify
the threshold value 0 , use the THETA= Weibull2-option. By default, THETA=0. An
advantage of the two-parameter Weibull plot over the three-parameter Weibull plot
is that the parameters c and can be estimated from the slope and intercept of the
point pattern. A disadvantage is that the two-parameter Weibull distribution applies
only in situations where the threshold parameter is known. To obtain a graphical estimate of 0 , specify a list of values for the THETA= Weibull2-option, and select the
value that most nearly linearizes the point pattern. To assess the point pattern, add
a diagonal distribution reference line corresponding to 0 and c0 with the SIGMA=
and C= Weibull2-options. Alternatively, you can add a distribution reference line corresponding to estimated values of 0 and c0 with the Weibull2-options SIGMA=EST
and C=EST. Agreement between the reference line and the point pattern indicates that
the Weibull distribution with parameters c0 , 0 , and 0 is a good t. You can specify
the SCALE= Weibull2-option as an alias for the SIGMA= Weibull2-option and the
SHAPE= Weibull2-option as an alias for the C= Weibull2-option. See Example 3.34.
ZETA=value | EST

species a value for the scale parameter for the lognormal quantile plots requested
with the LOGNORMAL option. Enclose the ZETA= lognormal-option in parentheses after the LOGNORMAL option. To request a distribution reference line with
intercept 0 and slope exp(0 ), specify the THETA=0 and ZETA=0 .

VAR Statement
VAR variables ;
The VAR statement species the analysis variables and their order in the results. By
default, if you omit the VAR statement, PROC UNIVARIATE analyzes all numeric
variables that are not listed in the other statements.

WEIGHT Statement
Using the Output Statement with the VAR Statement
You must provide a VAR statement when you use an OUTPUT statement. To store
the same statistic for several analysis variables in the OUT= data set, you specify a
list of names in the OUTPUT statement. PROC UNIVARIATE makes a one-to-one
correspondence between the order of the analysis variables in the VAR statement and
the list of names that follow a statistic keyword.

WEIGHT Statement
WEIGHT variable ;
The WEIGHT statement species numeric weights for analysis variables in the statistical calculations. The UNIVARIATE procedure uses the values wi of the WEIGHT
variable to modify the computation of a number of summary statistics by assuming
that the variance of the ith value xi of the analysis variable is equal to 2 /wi , where
is an unknown parameter. The values of the WEIGHT variable do not have to be integers and are typically positive. By default, observations with nonpositive or missing
values of the WEIGHT variable are handled as follows:
If the value is zero, the observation is counted in the total number of observations.
If the value is negative, it is converted to zero, and the observation is counted
in the total number of observations.
If the value is missing, the observation is excluded from the analysis.
To exclude observations that contain negative and zero weights from the analysis, use
EXCLNPWGT. Note that most SAS/STAT procedures, such as PROC GLM, exclude
negative and zero weights by default. The weight variable does not change how
the procedure determines the range, mode, extreme values, extreme observations, or
number of missing values. When you specify a WEIGHT statement, the procedure
also computes a weighted standard error and a weighted version of Students t test.
The Students t test is the only test of location that PROC UNIVARIATE computes
when you weight the analysis variables.
When you specify a WEIGHT variable, the procedure uses its values, wi , to compute
weighted versions of the statistics provided in the Moments table. For example, the
procedure computes a weighted mean xw and a weighted variance s2 as
w
i wi xi

xw =

i wi

and
s2 =
w

1
d

wi (xi xw )2
i

In Release 6.12 and earlier releases, observations were used in the analysis if and only if the
WEIGHT variable value was greater than zero.

In Release 6.12 and earlier releases, weighted skewness and kurtosis were not computed.

267

268

Chapter 3. The UNIVARIATE Procedure

where xi is the ith variable value. The divisor d is controlled by the VARDEF= option
in the PROC UNIVARIATE statement.
The WEIGHT statement does not affect the determination of the mode, extreme values, extreme observations, or the number of missing values of the analysis variables.
However, the weights wi are used to compute weighted percentiles. The WEIGHT
variable has no effect on graphical displays produced with the plot statements.
The CIPCTLDF, CIPCTLNORMAL, LOCCOUNT, NORMAL, ROBUSTSCALE,
TRIMMED=, and WINSORIZED= options are not available with the WEIGHT statement.
To compute weighted skewness or kurtosis, use VARDEF=DF or VARDEF=N in the
PROC statement.
You cannot specify the HISTOGRAM, PROBPLOT, or QQPLOT statements with the
WEIGHT statement.
When you use the WEIGHT statement, consider which value of the VARDEF= option
is appropriate. See VARDEF= and the calculation of weighted statistics in for more
information.

Details
Missing Values
PROC UNIVARIATE excludes missing values for an analysis variable before calculating statistics. Each analysis variable is treated individually; a missing value for an
observation in one variable does not affect the calculations for other variables. The
statements handle missing values as follows:
If a BY or an ID variable value is missing, PROC UNIVARIATE treats it like
any other BY or ID variable value. The missing values form a separate BY
group.
If the FREQ variable value is missing or nonpositive, PROC UNIVARIATE
excludes the observation from the analysis.
If the WEIGHT variable value is missing, PROC UNIVARIATE excludes the
observation from the analysis.
PROC UNIVARIATE tabulates the number of missing values and reports this information in the ODS table named Missing Values; see the section ODS Table Names
on page 309. Before the number of missing values is tabulated, PROC UNIVARIATE
excludes observations when
you use the FREQ statement and the frequencies are nonpositive.
you use the WEIGHT statement and the weights are missing or nonpositive
(you must specify the EXCLNPWGT option).

In Release 6.12 and earlier releases, the weights did not affect the computation of percentiles and
the procedure did not exclude the observations with missing weights from the count of observations.

Descriptive Statistics

Rounding
When you specify ROUND=u, PROC UNIVARIATE rounds a variable by using the
rounding unit to divide the number line into intervals with midpoints of the form
ui, where u is the nonnegative rounding unit and i is an integer. The interval width
is u. Any variable value that falls in an interval is rounded to the midpoint of that
interval. A variable value that is midway between two midpoints, and is therefore on
the boundary of two intervals, rounds to the even midpoint. Even midpoints occur
when i is an even integer (0, 2, 4, . . .).
When ROUND=1 and the analysis variable values are between 2.5 and 2.5, the
intervals are as follows:
Table 3.57. Intervals for Rounding When ROUND=1

i
2
1
0
1
2

Interval
[2.5,1.5]
[1.5,0.5]
[0.5,0.5]
[0.5,1.5]
[1.5,2.5]

Midpoint
2
1
0
1
2

Left endpt rounds to

2
2
0
0
2

Right endpt rounds to

2
0
0
2
2

When ROUND=.5 and the analysis variable values are between 1.25 and 1.25, the
intervals are as follows:
Table 3.58. Intervals for Rounding When ROUND=0.5

i
2
1
0
1
2

Interval
[1.25,0.75]
[0.75,0.25]
[0.25,0.25]
[0.25,0.75]
[0.75,1.25]

Midpoint
1.0
0.5
0.0
0.5
1.0

Left endpt rounds to

1
1
0
0
1

Right endpt rounds to

1
0
0
1
1

As the rounding unit increases, the interval width also increases. This reduces
the number of unique values and decreases the amount of memory that PROC
UNIVARIATE needs.

Descriptive Statistics
This section provides computational details for the descriptive statistics that are computed with the PROC UNIVARIATE statement. These statistics can also be saved in
the OUT= data set by specifying the keywords listed in Table 3.30 on page 237 in the
OUTPUT statement.
Standard algorithms (Fisher 1973) are used to compute the moment statistics. The
computational methods used by the UNIVARIATE procedure are consistent with
those used by other SAS procedures for calculating descriptive statistics.
The following sections give specic details on a number of statistics calculated by
the UNIVARIATE procedure.

269

270

Chapter 3. The UNIVARIATE Procedure

Mean
The sample mean is calculated as
n
i=1 wi xi
n
i=1 wi

xw =

where n is the number of nonmissing values for a variable, xi is the ith value of the
variable, and wi is the weight associated with the ith value of the variable. If there is
no WEIGHT variable, the formula reduces to
1
n

xi
i=1

Sum
The sum is calculated as n wi xi , where n is the number of nonmissing values
i=1
for a variable, xi is the ith value of the variable, and wi is the weight associated with
the ith value of the variable. If there is no WEIGHT variable, the formula reduces to
n
i=1 xi .

Sum of the Weights

n
The sum of the weights is calculated as
i=1 wi , where n is the number of nonmissing values for a variable and wi is the weight associated with the ith value of the
variable. If there is no WEIGHT variable, the sum of the weights is n.

Variance
The variance is calculated as
1
d

wi (xi xw )2

i=1

where n is the number of nonmissing values for a variable, xi is the ith value of the
variable, xw is the weighted mean, wi is the weight associated with the ith value of

the variable, and d is the divisor controlled by the VARDEF= option in the PROC
UNIVARIATE statement:

n1
n
d=
( i wi ) 1

i wi

if VARDEF=DF (default)
if VARDEF=N
if VARDEF=WDF
if VARDEF=WEIGHT|WGT

If there is no WEIGHT variable, the formula reduces to

1
d

(xi x)2

i=1

Descriptive Statistics
Standard Deviation
The standard deviation is calculated as
1
d

sw =

wi (xi xw )2

i=1

where n is the number of nonmissing values for a variable, xi is the ith value of the
variable, xw is the weighted mean, wi is the weight associated with the ith value of

the variable, and d is the divisor controlled by the VARDEF= option in the PROC
UNIVARIATE statement. If there is no WEIGHT variable, the formula reduces to

1
d

(xi x)2

i=1

Skewness
The sample skewness, which measures the tendency of the deviations to be larger in
one direction than in the other, is calculated as follows depending on the VARDEF=
option:
Table 3.59. Formulas for Skewness

VARDEF
DF (default)

Formula
n
(n 1)(n 2)
1
n

n
3/2

wi
i=1

WDF

3/2

wi
i=1

xi xw

missing

WEIGHT|WGT

missing

where n is the number of nonmissing values for a variable, xi is the ith value of the
variable, xw is the sample average, s is the sample standard deviation, and wi is the

weight associated with the ith value of the variable. If VARDEF=DF, then n must be
greater than 2. If there is no WEIGHT variable, then wi = 1 for all i = 1, . . . , n.
The sample skewness can be positive or negative; it measures the asymmetry of the

3
data distribution and estimates the theoretical skewness 1 = 3 2 2 , where 2
and 3 are the second and third central moments. Observations that are normally
distributed should have a skewness near zero.

271

272

Chapter 3. The UNIVARIATE Procedure

Kurtosis
The sample kurtosis, which measures the heaviness of tails, is calculated as follows
depending on the VARDEF= option:
Table 3.60. Formulas for Kurtosis

VARDEF
DF (default)

Formula
n(n + 1)
(n 1)(n 2)(n 3)
1
n

n
2
wi
i=1

WDF

2
wi
i=1

xi xw

3(n 1)2
(n 2)(n 3)

missing

WEIGHT|WGT

xi xw

missing

where n is the number of nonmissing values for a variable, xi is the ith value of the
variable, xw is the sample average, sw is the sample standard deviation, and wi is the

weight associated with the ith value of the variable. If VARDEF=DF, then n must be
greater than 3. If there is no WEIGHT variable, then wi = 1 for all i = 1, . . . , n.
The sample kurtosis measures the heaviness of the tails of the data distribution. It

estimates the adjusted theoretical kurtosis denoted as 2 3, where 2 = 242 , and

4 is the fourth central moment. Observations that are normally distributed should
have a kurtosis near zero.

Coefcient of Variation (CV)

The coefcient of variation is calculated as
CV =

100 sw
xw

Calculating the Mode

The mode is the value that occurs most often in the data. PROC UNIVARIATE counts
repetitions of the values of the analysis variables or, if you specify the ROUND=
option, the rounded values. If a tie occurs for the most frequent value, the procedure reports the lowest mode in the table labeled Basic Statistical Measures in the
statistical output. To list all possible modes, use the MODES option in the PROC
UNIVARIATE statement. When no repetitions occur in the data (as with truly continuous data), the procedure does not report the mode. The WEIGHT statement has
no effect on the mode. See Example 3.2.

Calculating Percentiles

Calculating Percentiles
The UNIVARIATE procedure automatically computes the 1st, 5th, 10th, 25th, 50th,
75th, 90th, 95th, and 99th percentiles (quantiles), as well as the minimum and maximum of each analysis variable. To compute percentiles other than these default
percentiles, use the PCTLPTS= and PCTLPRE= options in the OUTPUT statement.
You can specify one of ve denitions for computing the percentiles with the
PCTLDEF= option. Let n be the number of nonmissing values for a variable, and
let x1 , x2 , . . . , xn represent the ordered values of the variable. Let the tth percentile
t
be y, set p = 100 , and let
np = j + g
(n + 1)p = j + g

when PCTLDEF=1, 2, 3, or 5
when PCTLDEF=4

where j is the integer part of np, and g is the fractional part of np. Then the
PCTLDEF= option denes the tth percentile, y, as described in the following table:
Table 3.61. Percentile Denitions

PCTLDEF
1

Description
Weighted average at xnp

Formula
y = (1 g)xj + gxj+1
where x0 is taken to be x1
y
y
y
y

Observation numbered
closest to np

Empirical distribution function

Weighted average aimed

at x(n+1)p

Empirical distribution function

with averaging

1
= xj
if g < 2
1
= xj
if g = 2 and j is even
1
= xj+1 if g = 2 and j is odd
1
= xj+1 if g > 2
y = xj
if g = 0
y = xj+1 if g > 0

y = (1 g)xj + gxj+1
where xn+1 is taken to be xn
1
y = 2 (xj + xj+1 ) if g = 0
y = xj+1
if g > 0

Weighted Percentiles
When you use a WEIGHT statement, the percentiles are computed differently. The
100pth weighted percentile y is computed from the empirical distribution function
with averaging

1
2 (xi

+ xi+1 ) if
xi+1
if

i
j=1 wj
i
j=1 wj

= pW
< pW <

i+1
j=1 wj

where wi is the weight associated with xi , and where W =

the weights.

n
i=1 wi

is the sum of

Note that the PCTLDEF= option is not applicable when a WEIGHT statement is
used. However, in this case, if all the weights are identical, the weighted percentiles
are the same as the percentiles that would be computed without a WEIGHT statement
and with PCTLDEF=5.

273

274

Chapter 3. The UNIVARIATE Procedure

Condence Limits for Percentiles

You can use the CIPCTLNORMAL option to request condence limits for percentiles, assuming the data are normally distributed. These limits are described
in Section 4.4.1 of Hahn and Meeker (1991). When 0 < p < 1 , the two-sided
2
100(1 )% condence limits for the 100pth percentile are

lower limit = X g ( ; 1 p, n)s

upper limit = X g (1 ; p, n)s

2
where n is the sample size. When 1 p < 1, the two-sided 100(1 )% condence
2
limits for the 100pth percentile are

lower limit = X + g ( ; 1 p, n)s

upper limit = X + g (1 ; p, n)s

2
One-sided 100(1 )% condence bounds are computed by replacing by in the
2
appropriate preceding equation. The factor g (, p, n) is related to the noncentral t
distribution and is described in Owen and Hua (1977) and Odeh and Owen (1980).
See Example 3.10.
You can use the CIPCTLDF option to request distribution-free condence limits for
percentiles. In particular, it is not necessary to assume that the data are normally
distributed. These limits are described in Section 5.2 of Hahn and Meeker (1991).
The two-sided 100(1 )% condence limits for the 100pth percentile are
lower limit = X(l)
upper limit = X(u)
where X(j) is the jth order statistic when the data values are arranged in increasing
order:
X(1) X(2) . . . X(n)
The lower rank l and upper rank u are integers that are symmetric (or nearly symmetric) around [np] + 1 where [np] is the integer part of np, and where n is the sample
size. Furthermore, l and u are chosen so that X(l) and X(u) are as close to X[n+1]p as
possible while satisfying the coverage probability requirement
Q(u 1; n, p) Q(l 1; n, p) 1
where Q(k; n, p) is the cumulative binomial probability
k

Q(k; n, p) =
i=0

n i
p (1 p)ni
i

Tests for Location

In some cases, the coverage requirement cannot be met, particularly when n is
small and p is near 0 or 1. To relax the requirement of symmetry, you can specify
CIPCTLDF(TYPE = ASYMMETRIC). This option requests symmetric limits when
the coverage requirement can be met, and asymmetric limits otherwise.
If you specify CIPCTLDF(TYPE = LOWER), a one-sided 100(1 )% lower condence bound is computed as X(l) , where l is the largest integer that satises the
inequality
1 Q(l 1; n, p) 1
with 0 < l n. Likewise, if you specify CIPCTLDF(TYPE = UPPER), a one-sided
100(1 )% lower condence bound is computed as X(u) , where u is the largest
integer that satises the inequality
Q(u 1; n, p) 1

where 0 < u n

Note that condence limits for percentiles are not computed when a WEIGHT statement is specied. See Example 3.10.

Tests for Location

PROC UNIVARIATE provides three tests for location: Students t test, the sign test,
and the Wilcoxon signed rank test. All three tests produce a test statistic for the
null hypothesis that the mean or median is equal to a given value 0 against the
two-sided alternative that the mean or median is not equal to 0 . By default, PROC
UNIVARIATE sets the value of 0 to zero. You can use the MU0= option in the
PROC UNIVARIATE statement to specify the value of 0 . Students t test is appropriate when the data are from an approximately normal population; otherwise, use
nonparametric tests such as the sign test or the signed rank test. For large sample
situations, the t test is asymptotically equivalent to a z test. If you use the WEIGHT
statement, PROC UNIVARIATE computes only one weighted test for location, the t
test. You must use the default value for the VARDEF= option in the PROC statement
(VARDEF=DF). See Example 3.12.
You can also use these tests to compare means or medians of paired data. Data are
said to be paired when subjects or units are matched in pairs according to one or more
variables, such as pairs of subjects with the same age and gender. Paired data also
occur when each subject or unit is measured at two times or under two conditions. To
compare the means or medians of the two times, create an analysis variable that is the
difference between the two measures. The test that the mean or the median difference
of the variables equals zero is equivalent to the test that the means or medians of the
two original variables are equal. Note that you can also carry out these tests using
the PAIRED statement in the TTEST procedure; refer to Chapter 77, The TTEST
Procedure, in SAS/STAT Users Guide. Also see Example 3.13.

275

276

Chapter 3. The UNIVARIATE Procedure

Students t Test
PROC UNIVARIATE calculates the t statistic as
t=

x 0

s/ n

where x is the sample mean, n is the number of nonmissing values for a variable, and

s is the sample standard deviation. The null hypothesis is that the population mean
equals 0 . When the data values are approximately normally distributed, the probability under the null hypothesis of a t statistic that is as extreme, or more extreme,
than the observed value (the p-value) is obtained from the t distribution with n 1
degrees of freedom. For large n, the t statistic is asymptotically equivalent to a z test.
When you use the WEIGHT statement and the default value of VARDEF=, which is
DF, the t statistic is calculated as
tw =

xw 0

n
sw /
i=1 wi

where xw is the weighted mean, sw is the weighted standard deviation, and wi is

the weight for ith observation. The tw statistic is treated as having a Students t
distribution with n 1 degrees of freedom. If you specify the EXCLNPWGT option
in the PROC statement, n is the number of nonmissing observations when the value
of the WEIGHT variable is positive. By default, n is the number of nonmissing
observations for the WEIGHT variable.

Sign Test
PROC UNIVARIATE calculates the sign test statistic as
M = (n+ n )/2
where n+ is the number of values that are greater than 0 , and n is the number
of values that are less than 0 . Values equal to 0 are discarded. Under the null
hypothesis that the population median is equal to 0 , the p-value for the observed
statistic Mobs is
min(n+ ,n )

Pr(|Mobs | |M |) = 0.5(nt 1)
j=0

nt
j

where nt = n+ + n is the number of xi values not equal to 0 .

Note: If n+ and n are equal, the p-value is equal to one.

Condence Limits for Parameters of the Normal Distribution

Wilcoxon Signed Rank Test
The signed rank statistic S is computed as
+
ri

S=
i:xi >0

nt (nt + 1)
4

+
where ri is the rank of |xi 0 | after discarding values of xi = 0 , and nt is the
number of xi values not equal to 0 . Average ranks are used for tied values.

If n 20, the signicance of S is computed from the exact distribution of S, where

the distribution is a convolution of scaled binomial distributions. When n > 20, the
signicance of S is computed by treating

n1
nV S 2

as a Students t variate with n 1 degrees of freedom. V is computed as

1
1
n(n + 1)(2n + 1)
24
48

V =

ti (ti + 1)(ti 1)

where the sum is over groups tied in absolute value and where ti is the number of
values in the ith group (Iman 1974; Conover 1999). The null hypothesis tested is that
the mean (or median) is zero, assuming that the distribution is symmetric. Refer to
Lehmann (1998).

Condence Limits for Parameters of the Normal Distribution

The two-sided 100(1 )% condence interval for the mean has upper and lower
limits
s
x t1 ;n1

n
2
where s2 =

1
n1

(xi x)2 and t1 ;n1 is the (1 ) percentile of the t distri

2
2

bution with n 1 degrees of freedom. The one-sided upper 100(1 )% condence

limit is computed as x+ n t1;n1 and the one-sided lower 100(1)% condence
s
s
limit is computed as x n t1;n1 . See Example 3.9.

The two-sided 100(1 )% condence interval for the standard deviation has lower
and upper limits
s

n1
2
1 ;n1
2

and s

n1
2

2 ;n1

277

278

Chapter 3. The UNIVARIATE Procedure

respectively, where 2

1 2 ;n1

and 2

2 ;n1

are the (1 ) and

percentiles of the

chi-square distribution with n 1 degrees of freedom. A one-sided 100(1 )%

condence limit has lower and upper limits
s

n1
2
1;n1

and s

n1
2
;n1

respectively. The 100(1 )% condence interval for the variance has upper and
lower limits equal to the squares of the corresponding upper and lower limits for the
standard deviation. When you use the WEIGHT statement and specify VARDEF=DF
in the PROC statement, the 100(1 )% condence interval for the weighted mean
is
sw

xw t1

n
i=1 wi

where xw is the weighted mean, sw is the weighted standard deviation, wi is the

weight for ith observation, and t1 is the (1 ) percentile for the t distribution
2
with n 1 degrees of freedom.

Robust Estimators
A statistical method is robust if it is insensitive to moderate or even large departures
from the assumptions that justify the method. PROC UNIVARIATE provides several
methods for robust estimation of location and scale. See Example 3.11.

Winsorized Means
The Winsorized mean is a robust estimator of the location that is relatively insensitive
to outliers. The k-times Winsorized mean is calculated as

xwk =

1
n

nk1

x(i) + (k + 1)x(nk)

(k + 1)x(k+1) +
i=k+2

where n is the number of observations, and x(i) is the ith order statistic when the
observations are arranged in increasing order:
x(1) x(2) . . . x(n)
The Winsorized mean is computed as the ordinary mean after the k smallest observations are replaced by the (k +1)st smallest observation, and the k largest observations
are replaced by the (k + 1)st largest observation.
For data from a symmetric distribution, the Winsorized mean is an unbiased estimate
of the population mean. However, the Winsorized mean does not have a normal
distribution even if the data are from a normal population.

Robust Estimators

279

The Winsorized sum of squared deviations is dened as

nk1

s2
wk

(x(i) xwk )2 + (k + 1)(x(nk) xwk )2

= (k + 1)(x(k+1) xwk ) +

i=k+2

The Winsorized t statistic is given by

twk =

xwk 0

SE(wk )
x

where 0 denotes the location under the null hypothesis, and the standard error of the
Winsorized mean is
SE(wk ) =
x

n 2k 1

swk
n(n 1)

When the data are from a symmetric distribution, the distribution of twk is approximated by a Students t distribution with n 2k 1 degrees of freedom (Tukey and
McLaughlin 1963; Dixon and Tukey 1968).
The Winsorized 100(1 )% condence interval for the location parameter has
2
upper and lower limits
xwk t1 ;n2k1 SE(wk )

x
2

where t1 ;n2k1 is the (1 )100th percentile of the Students t distribution with

2
2

n 2k 1 degrees of freedom.

Trimmed Means
Like the Winsorized mean, the trimmed mean is a robust estimator of the location
that is relatively insensitive to outliers. The k-times trimmed mean is calculated as

xtk =

1
n 2k

x(i)
i=k+1

where n is the number of observations, and x(i) is the ith order statistic when the
observations are arranged in increasing order:
x(1) x(2) . . . x(n)
The trimmed mean is computed after the k smallest and k largest observations are
deleted from the sample. In other words, the observations are trimmed at each end.

280

Chapter 3. The UNIVARIATE Procedure

For a symmetric distribution, the symmetrically trimmed mean is an unbiased estimate of the population mean. However, the trimmed mean does not have a normal
distribution even if the data are from a normal population.
A robust estimate of the variance of the trimmed mean ttk can be based on
the Winsorized sum of squared deviations s2 , which is dened in the section
wk
Winsorized Means on page 278; refer to Tukey and McLaughlin (1963). This can
be used to compute a trimmed t test which is based on the test statistic

ttk =

(tk 0 )
x
SE(tk )
x

where the standard error of the trimmed mean is

SE(tk ) =
x

swk
(n 2k)(n 2k 1)

When the data are from a symmetric distribution, the distribution of ttk is approximated by a Students t distribution with n 2k 1 degrees of freedom (Tukey and
McLaughlin 1963; Dixon and Tukey 1968).
The trimmed 100(1 )% condence interval for the location parameter has upper
and lower limits
xtk t1 ;n2k1 SE(tk )

x
2

where t1 ;n2k1 is the (1 )100th percentile of the Students t distribution with

2
2

n 2k 1 degrees of freedom.

Robust Estimates of Scale

The sample standard deviation, which is the most commonly used estimator of scale,
is sensitive to outliers. Robust scale estimators, on the other hand, remain bounded
when a single data value is replaced by an arbitrarily large or small value. The
UNIVARIATE procedure computes several robust measures of scale, including the
interquartile range, Ginis mean difference G, the median absolute deviation about
the median (MAD), Qn , and Sn . In addition, the procedure computes estimates of
the normal standard deviation derived from each of these measures.
The interquartile range (IQR) is simply the difference between the upper and lower
quartiles. For a normal population, can be estimated as IQR/1.34898.
Ginis mean difference is computed as
G=

1
n
2

|xi xj |
i<j

Creating Line Printer Plots

For a normal population, the expected value of G is 2/ . Thus G /2 is a robust

estimator of when the data are from a normal sample. For the normal distribution,
this estimator has high efciency relative to the usual sample standard deviation, and
it is also less sensitive to the presence of outliers.
A very robust scale estimator is the MAD, the median absolute deviation from the
median (Hampel 1974), which is computed as
MAD = medi (|xi medj (xj )|)
where the inner median, medj (xj ), is the median of the n observations, and the outer
median (taken over i) is the median of the n absolute values of the deviations about
the inner median. For a normal population, 1.4826MAD is an estimator of .
The MAD has low efciency for normal distributions, and it may not always be appropriate for symmetric distributions. Rousseeuw and Croux (1993) proposed two
statistics as alternatives to the MAD. The rst is
Sn = 1.1926medi (medj (|xi xj |))
where the outer median (taken over i) is the median of the n medians of |xi xj |,
j = 1, 2, . . . , n. To reduce small-sample bias, csn Sn is used to estimate , where csn
is a correction factor; refer to Croux and Rousseeuw (1992).
The second statistic proposed by Rousseeuw and Croux (1993) is
Qn = 2.219{|xi xj |; i < j}(k)
where
k=

n
2

+1
2

n
distances between
2
the data points. The bias-corrected statistic cqn Qn is used to estimate , where cqn is
a correction factor; refer to Croux and Rousseeuw (1992).
In other words, Qn is 2.219 times the kth order statistic of the

Creating Line Printer Plots

The PLOTS option in the PROC UNIVARIATE statement provides up to four diagnostic line printer plots to examine the data distribution. These plots are the stemand-leaf plot or horizontal bar chart, the box plot, the normal probability plot, and the
side-by-side box plots. If you specify the WEIGHT statement, PROC UNIVARIATE
provides a weighted histogram, a weighted box plot based on the weighted quantiles,
and a weighted normal probability plot.

281

282

Chapter 3. The UNIVARIATE Procedure

Note that these plots are a legacy feature of the UNIVARIATE procedure in earlier
versions of SAS. They predate the addition of the HISTOGRAM, PROBPLOT, and
QQPLOT statements, which provide high-resolution graphics displays. Also note
that line printer plots requested with the PLOTS option are mainly intended for use
with the ODS LISTING destination. See Example 3.5.

Stem-and-Leaf Plot
The rst plot in the output is either a stem-and-leaf plot (Tukey 1977) or a horizontal
bar chart. If any single interval contains more than 49 observations, the horizontal bar
chart appears. Otherwise, the stem-and-leaf plot appears. The stem-and-leaf plot is
like a horizontal bar chart in that both plots provide a method to visualize the overall
distribution of the data. The stem-and-leaf plot provides more detail because each
point in the plot represents an individual data value.
To change the number of stems that the plot displays, use PLOTSIZE= to increase or
decrease the number of rows. Instructions that appear below the plot explain how to
determine the values of the variable. If no instructions appear, you multiply Stem.Leaf
by 1 to determine the values of the variable. For example, if the stem value is 10 and
the leaf value is 1, then the variable value is approximately 10.1. For the stem-andleaf plot, the procedure rounds a variable value to the nearest leaf. If the variable
value is exactly halfway between two leaves, the value rounds to the nearest leaf with
an even integer value. For example, a variable value of 3.15 has a stem value of 3 and
a leaf value of 2.

Box Plot
The box plot, also known as a schematic box plot, appears beside the stem-and-leaf
plot. Both plots use the same vertical scale. The box plot provides a visual summary
of the data and identies outliers. The bottom and top edges of the box correspond to
the sample 25th (Q1) and 75th (Q3) percentiles. The box length is one interquartile
range (Q3 - Q1). The center horizontal line with asterisk endpoints corresponds to
the sample median. The central plus sign (+) corresponds to the sample mean. If the
mean and median are equal, the plus sign falls on the line inside the box. The vertical
lines that project out from the box, called whiskers, extend as far as the data extend,
up to a distance of 1.5 interquartile ranges. Values farther away are potential outliers.
The procedure identies the extreme values with a zero or an asterisk (*). If zero
appears, the value is between 1.5 and 3 interquartile ranges from the top or bottom
edge of the box. If an asterisk appears, the value is more extreme.
Note: To produce box plots using high-resolution graphics, use the BOXPLOT procedure in SAS/STAT software; refer to Chapter 18, The BOXPLOT Procedure, in
SAS/STAT Users Guide.

Normal Probability Plot

The normal probability plot plots the empirical quantiles against the quantiles of a
standard normal distribution. Asterisks (*) indicate the data values. The plus signs
(+) provide a straight reference line that is drawn by using the sample mean and
standard deviation. If the data are from a normal distribution, the asterisks tend to fall

Creating Line Printer Plots

along the reference line. The vertical coordinate is the data value, and the horizontal
coordinate is 1 (vi ) where
vi

ri 3
8
1
n+ 4

1 () = inverse of the standard normal distribution function

ri
= rank of the ith data value when ordered from smallest to largest
n
= number of nonmissing observations
For a weighted normal probability plot, the ith ordered observation is plotted against
1 (vi ) where
vi

3
(1 8i )
1
(1+ 4n )

i
j=1 w(j)
n
i=1 wi

w(j) = weight associated with the jth ordered observation

When each observation has an identical weight, wj = w, the formula for vi reduces
to the expression for vi in the unweighted normal probability plot:
vi =

3
i 8
1
n+ 4

When the value of VARDEF= is WDF or WEIGHT, a reference line with intercept

and slope is added to the plot. When the value of VARDEF= is DF or N, the slope

n
w

is w where w = i=1 i is the average weight.

When each observation has an identical weight and the value of VARDEF= is DF, N,
or WEIGHT, the reference line reduces to the usual reference line with intercept

and slope in the unweighted normal probability plot.

If the data are normally distributed with mean , standard deviation , and each observation has an identical weight w, then the points on the plot should lie approximately
on a straight line. The intercept for this line is . The slope is when VARDEF= is

WDF or WEIGHT, and the slope is w when VARDEF= is DF or N.

Note: To produce probability plots using high-resolution graphics, use the
PROBPLOT statement in PROC UNIVARIATE; see the section PROBPLOT
Statement on page 241.

Side-by-Side Box Plots

When you use a BY statement with the PLOT option, PROC UNIVARIATE produces side-by-side box plots, one for each BY group. The box plots (also known
as schematic plots) use a common scale that enables you to compare the data distribution across BY groups. This plot appears after the univariate analyses of all BY
groups. Use the NOBYPLOT option to suppress this plot.
Note: To produce side-by-side box plots using high-resolution graphics, use the
BOXPLOT procedure in SAS/STAT software; refer to Chapter 18, The BOXPLOT
Procedure, in SAS/STAT Users Guide.

283

284

Chapter 3. The UNIVARIATE Procedure

Creating High-Resolution Graphics

If your site licenses SAS/GRAPH software, you can use the HISTOGRAM,
PROBPLOT, and QQPLOT statements to create high-resolution graphs. The
HISTOGRAM statement creates histograms that enable you to examine the data distribution. You can optionally t families of density curves and superimpose kernel
density estimates on the histograms. For additional information about the tted distributions and kernel density estimates, see the section Formulas for Fitted Continuous
Distributions on page 288 and the section Kernel Density Estimates on page 297.
The PROBPLOT statement creates a probability plot, which compares ordered values of a variable with percentiles of a specied theoretical distribution. The QQPLOT
statement creates a quantile-quantile plot, which compares ordered values of a variable with quantiles of a specied theoretical distribution. You can use these plots to
determine how well a theoretical distribution models a data distribution.
Note: You can use the CLASS statement with the HISTOGRAM, PROBPLOT, or
QQPLOT statements to create comparative histograms, probability plots, or Q-Q
plots, respectively.

Using the CLASS Statement to Create Comparative Plots

When you use the CLASS statement with the HISTOGRAM, PROBPLOT, or
QQPLOT statement, PROC UNIVARIATE creates comparative histograms, comparative probability plots, or comparative quantile-quantile plots. You can use these plot
statements with the CLASS statement to create one-way and two-way comparative
plots. When you use one class variable, PROC UNIVARIATE displays an array of
component plots (stacked or side-by-side), one for each level of the classication
variable. When you use two class variables, PROC UNIVARIATE displays a matrix
of component plots, one for each combination of levels of the classication variables.
The observations in a given level are referred to collectively as a cell.
When you create a one-way comparative plot, the observations in the input data set
are sorted by the method specied in the ORDER= option. PROC UNIVARIATE
creates a separate plot for the analysis variable values in each level, and arranges these
component plots in an array to form the comparative plot with uniform horizontal and
vertical axes. See Example 3.15.
When you create a two-way comparative plot, the observations in the input data
set are cross-classied according to the values (levels) of these variables. PROC
UNIVARIATE creates a separate plot for the analysis variable values in each cell of
the cross-classication and arranges these component plots in a matrix to form the
comparative plot with uniform horizontal and vertical axes. The levels of the rst
class variable are the labels for the rows of the matrix, and the levels of the second
class variable are the labels for the columns of the matrix. See Example 3.16.
PROC UNIVARIATE determines the layout of a two-way comparative plot by using the order for the rst class variable to obtain the order of the rows from top
to bottom. Then it applies the order for the second class variable to the observations that correspond to the rst row to obtain the order of the columns from left to

Positioning the Inset

right. If any columns remain unordered (that is, the categories are unbalanced), PROC
UNIVARIATE applies the order for the second class variable to the observations in
the second row, and so on, until all the columns have been ordered.
If you associate a label with a variable, PROC UNIVARIATE displays the variable
label in the comparative plot and this label is parallel to the column (or row) labels.
Use the MISSING option to treat missing values as valid levels.
To reduce the number of classication levels, use a FORMAT statement to combine
variable values.

Positioning the Inset

Positioning the Inset Using Compass Point Values
To position the inset by using a compass point position, specify the value N, NE, E,
SE, S, SW, W, or NW with the POSITION= option. The default position of the inset
is NW. The following statements produce a histogram to show the position of the
inset for the eight compass points:
data Score;
input Student $ PreTest PostTest @@;
label ScoreChange = Change in Test Scores;
ScoreChange = PostTest - PreTest;
datalines;
Capalleti 94 91 Dubose
51 65
Engles
95 97 Grant
63 75
Krupski
80 75 Lundsford 92 55
Mcbane
75 78 Mullen
89 82
Nguyen
79 76 Patel
71 77
Si
75 70 Tanaka
87 73
;
run;
title Test Scores for a College Course;
proc univariate data=Score noprint;
histogram PreTest / midpoints = 45 to 95 by 10;
inset n
/ cfill=blank
header=Position = NW pos=nw;
inset mean / cfill=blank
header=Position = N pos=n ;
inset sum
/ cfill=blank
header=Position = NE pos=ne;
inset max
/ cfill=blank
header=Position = E pos=e ;
inset min
/ cfill=blank
header=Position = SE pos=se;
inset nobs / cfill=blank
header=Position = S pos=s ;
inset range / cfill=blank
header=Position = SW pos=sw;
inset mode / cfill=blank
header=Position = W pos=w ;
label PreTest = Pretest Score;
run;

285

286

Chapter 3. The UNIVARIATE Procedure

Figure 3.7. Compass Positions for Inset

Positioning the Inset in the Margins

To position the inset in one of the four margins that surround the plot area, specify
the value LM, RM, TM, or BM with the POSITION= option.
Locating the Inset in the Margins

Margin positions are recommended if you list a large number of statistics in the
INSET statement. If you attempt to display a lengthy inset in the interior of the
plot, it is most likely that the inset will collide with the data display.
Positioning the Inset Using Coordinates

To position the inset with coordinates, use POSITION=(x,y). You specify the coordinates in axis data units or in axis percentage units (the default).
If you specify the DATA option immediately following the coordinates, PROC
UNIVARIATE positions the inset by using axis data units. For example, the following statements place the bottom left corner of the inset at 45 on the horizontal
axis and 10 on the vertical axis:
title Test Scores for a College Course;
proc univariate data=Score noprint;
histogram PreTest / midpoints = 45 to 95 by 10;
inset n / header
= Position=(45,10)
position = (45,10) data;
run;

Positioning the Inset

Figure 3.8. Coordinate Position for Inset

By default, the specied coordinates determine the position of the bottom left corner
of the inset. To change this reference point, use the REFPOINT= option (see the next
example).
If you omit the DATA option, PROC UNIVARIATE positions the inset by using axis
percentage units. The coordinates in axis percentage units must be between 0 and
100. The coordinates of the bottom left corner of the display are (0,0), while the
upper right corner is (100, 100). For example, the following statements create a
histogram and use coordinates in axis percentage units to position the two insets:
title Test Scores for a College Course;
proc univariate data=Score noprint;
histogram PreTest / midpoints = 45 to 95 by 10;
inset min / position = (5,25)
header
= Position=(5,25)
refpoint = tl;
inset max / position = (95,95)
header
= Position=(95,95)
refpoint = tr;
run;

The REFPOINT= option determines which corner of the inset to place at the coordinates that are specied with the POSITION= option. The rst inset uses
REFPOINT=TL, so that the top left corner of the inset is positioned 5% of the way
across the horizontal axis and 25% of the way up the vertical axis. The second inset
uses REFPOINT=TR, so that the top right corner of the inset is positioned 95% of
the way across the horizontal axis and 95% of the way up the vertical axis.

287

288

Chapter 3. The UNIVARIATE Procedure

Figure 3.9. Reference Point for Inset

A sample program, univar3.sas, for these examples is available in the SAS Sample
Library for Base SAS software.

Formulas for Fitted Continuous Distributions

The following sections provide information on the families of parametric distributions that you can t with the HISTOGRAM statement. Properties of these distributions are discussed by Johnson, Kotz, and Balakrishnan (1994, 1995).

Beta Distribution
The tted density function is
1

p(x) =

100h% (x) (+x)

B(,) (+1)
0

where B(, ) =

()()
(+)

for < x < +

for x or x +

and

= lower threshold parameter (lower endpoint parameter)

= scale parameter ( > 0)
= shape parameter ( > 0)
= shape parameter ( > 0)
h = width of histogram interval

Formulas for Fitted Continuous Distributions

Note: This notation is consistent with that of other distributions that you can t with
the HISTOGRAM statement. However, many texts, including Johnson, Kotz, and
Balakrishnan (1995), write the beta density function as

p(x) =

(xa)p1 (bx)q1
B(p,q)(ba)p+q1

for a < x < b

for x a or x b

The two parameterizations are related as follows:

=ba
=a
=p
=q
The range of the beta distribution is bounded below by a threshold parameter = a
and above by + = b. If you specify a tted beta curve using the BETA option,
must be less than the minimum data value, and + must be greater than the
maximum data value. You can specify and with the THETA= and SIGMA=
beta-options in parentheses after the keyword BETA. By default, = 1 and = 0.
If you specify THETA=EST and SIGMA=EST, maximum likelihood estimates are
computed for and . However, three- and four-parameter maximum likelihood
estimation may not always converge.
In addition, you can specify and with the ALPHA= and BETA= beta-options,
respectively. By default, the procedure calculates maximum likelihood estimates for
and . For example, to t a beta density curve to a set of data bounded below by 32
and above by 212 with maximum likelihood estimates for and , use the following
statement:
histogram Length / beta(theta=32 sigma=180);

The beta distributions are also referred to as Pearson Type I or II distributions. These
include the power-function distribution ( = 1), the arc-sine distribution ( = =
1
1
2 ), and the generalized arc-sine distributions ( + = 1, = 2 ).
You can use the DATA step function BETAINV to compute beta quantiles and the
DATA step function PROBBETA to compute beta probabilities.

Exponential Distribution
The tted density function is
p(x) =

100h%

exp(( x )) for x

for x <

where
= threshold parameter
= scale parameter ( > 0)
h = width of histogram interval

289

290

Chapter 3. The UNIVARIATE Procedure

The threshold parameter must be less than or equal to the minimum data value. You
can specify with the THRESHOLD= exponential-option. By default, = 0. If you
specify THETA=EST, a maximum likelihood estimate is computed for . In addition,
you can specify with the SCALE= exponential-option. By default, the procedure
calculates a maximum likelihood estimate for . Note that some authors dene the
1
scale parameter as .
The exponential distribution is a special case of both the gamma distribution (with
= 1) and the Weibull distribution (with c = 1). A related distribution is the
extreme value distribution. If Y = exp(X) has an exponential distribution, then X
has an extreme value distribution.

Gamma Distribution
The tted density function is

p(x) =

100h% x 1
exp(( x ))

() ( )

for x >
for x

where
= threshold parameter
= scale parameter ( > 0)
= shape parameter ( > 0)
h = width of histogram interval
The threshold parameter must be less than the minimum data value. You can specify with the THRESHOLD= gamma-option. By default, = 0. If you specify
THETA=EST, a maximum likelihood estimate is computed for . In addition, you
can specify and with the SCALE= and ALPHA= gamma-options. By default,
the procedure calculates maximum likelihood estimates for and .
The gamma distributions are also referred to as Pearson Type III distributions, and
they include the chi-square, exponential, and Erlang distributions. The probability
density function for the chi-square distribution is

p(x) =

1
2( )
2

x
2

1
2

exp( x ) for x > 0

2
for x 0

Notice that this is a gamma distribution with = , = 2, and = 0. The expo2

nential distribution is a gamma distribution with = 1, and the Erlang distribution
is a gamma distribution with being a positive integer. A related distribution is the
1 ,...,Xn )
Rayleigh distribution. If R = max(X1 ,...,Xn ) where the Xi s are independent 2 vari
min(X
ables, then log R is distributed with a distribution having a probability density
function of

p(x) =

2 2 1 ( )
2
0

x1 exp( x ) for x > 0

2
for x 0

Formulas for Fitted Continuous Distributions

If = 2, the preceding distribution is referred to as the Rayleigh distribution.
You can use the DATA step function GAMINV to compute gamma quantiles and the
DATA step function PROBGAM to compute gamma probabilities.

Lognormal Distribution
The tted density function is

p(x) =

100h%

2(x)

exp (log(x))
2 2

for x >
for x

0
where
= threshold parameter
= scale parameter ( < < )
= shape parameter ( > 0)
h = width of histogram interval

The threshold parameter must be less than the minimum data value. You can specify
with the THRESHOLD= lognormal-option. By default, = 0. If you specify
THETA=EST, a maximum likelihood estimate is computed for . You can specify
and with the SCALE= and SHAPE= lognormal-options, respectively. By default,
the procedure calculates maximum likelihood estimates for these parameters.
Note: The lognormal distribution is also referred to as the SL distribution in the
Johnson system of distributions.
Note: This book uses to denote the shape parameter of the lognormal distribution,
whereas is used to denote the scale parameter of the beta, exponential, gamma,
normal, and Weibull distributions. The use of to denote the lognormal shape pa1
rameter is based on the fact that (log(X ) ) has a standard normal distribution
if X is lognormally distributed. Based on this relationship, you can use the DATA
step function PROBIT to compute lognormal quantiles and the DATA step function
PROBNORM to compute probabilities.

Normal Distribution
The tted density function is
p(x) =

100h%

1
exp 2 ( x )2

where
= mean
= standard deviation ( > 0)
h = width of histogram interval

for < x <

291

292

Chapter 3. The UNIVARIATE Procedure

You can specify and with the MU= and SIGMA= normal-options, respectively.
By default, the procedure estimates with the sample mean and with the sample
standard deviation.
You can use the DATA step function PROBIT to compute normal quantiles and the
DATA step function PROBNORM to compute probabilities.
Note: The normal distribution is also referred to as the SN distribution in the Johnson
system of distributions.

Weibull Distribution
The tted density function is

p(x) =

c
100h% ( x )c1 exp(( x )c ) for x >

0
for x

where
= threshold parameter
= scale parameter ( > 0)
c = shape parameter (c > 0)
h = width of histogram interval
The threshold parameter must be less than the minimum data value. You can specify with the THRESHOLD= Weibull-option. By default, = 0. If you specify
THETA=EST, a maximum likelihood estimate is computed for . You can specify
and c with the SCALE= and SHAPE= Weibull-options, respectively. By default, the
procedure calculates maximum likelihood estimates for and c.
The exponential distribution is a special case of the Weibull distribution where c = 1.

Goodness-of-Fit Tests
When you specify the NORMAL option in the PROC UNIVARIATE statement or you
request a tted parametric distribution in the HISTOGRAM statement, the procedure
computes goodness-of-t tests for the null hypothesis that the values of the analysis
variable are a random sample from the specied theoretical distribution. See Example
3.22.
When you specify the NORMAL option, these tests, which are summarized in the
output table labeled Tests for Normality, include the following:
Shapiro-Wilk test
Kolmogorov-Smirnov test
Anderson-Darling test
Cramr-von Mises test

Goodness-of-Fit Tests
The Kolmogorov-Smirnov D statistic, the Anderson-Darling statistic, and the
Cramr-von Mises statistic are based on the empirical distribution function (EDF).
However, some EDF tests are not supported when certain combinations of the parameters of a specied distribution are estimated. See Table 3.62 on page 296 for a
list of the EDF tests available. You determine whether to reject the null hypothesis
by examining the p-value that is associated with a goodness-of-t statistic. When the
p-value is less than the predetermined critical value (), you reject the null hypothesis
and conclude that the data did not come from the specied distribution.
If you want to test the normality assumptions for analysis of variance methods, beware of using a statistical test for normality alone. A tests ability to reject the null
hypothesis (known as the power of the test) increases with the sample size. As the
sample size becomes larger, increasingly smaller departures from normality can be
detected. Since small deviations from normality do not severely affect the validity of analysis of variance tests, it is important to examine other statistics and plots
to make a nal assessment of normality. The skewness and kurtosis measures and
the plots that are provided by the PLOTS option, the HISTOGRAM statement, the
PROBPLOT statement, and the QQPLOT statement can be very helpful. For small
sample sizes, power is low for detecting larger departures from normality that may
be important. To increase the tests ability to detect such deviations, you may want to
declare signicance at higher levels, such as 0.15 or 0.20, rather than the often-used
0.05 level. Again, consulting plots and additional statistics will help you assess the
severity of the deviations from normality.

Shapiro-Wilk Statistic
If the sample size is less than or equal to 2000 and you specify the NORMAL option,
PROC UNIVARIATE computes the Shapiro-Wilk statistic, W (also denoted as Wn
to emphasize its dependence on the sample size n). The W statistic is the ratio of
the best estimator of the variance (based on the square of a linear combination of
the order statistics) to the usual corrected sum of squares estimator of the variance
(Shapiro and Wilk 1965). When n is greater than three, the coefcients to compute the linear combination of the order statistics are approximated by the method of
Royston (1992). The statistic W is always greater than zero and less than or equal to
one (0 < W 1).
Small values of W lead to the rejection of the null hypothesis of normality. The
distribution of W is highly skewed. Seemingly large values of W (such as 0.90)
may be considered small and lead you to reject the null hypothesis. The method for
computing the p-value (the probability of obtaining a W statistic less than or equal
to the observed value) depends on n. For n = 3, the probability distribution of W is
known and is used to determine the p-value. For n > 4, a normalizing transformation
is computed:

Zn =

( log( log(1 Wn )) )/ if 4 n 11
(log(1 Wn ) )/
if 12 n 2000

293

294

Chapter 3. The UNIVARIATE Procedure

The values of , , and are functions of n obtained from simulation results. Large
values of Zn indicate departure from normality, and since the statistic Zn has an
approximately standard normal distribution, this distribution is used to determine the
p-values for n > 4.

EDF Goodness-of-Fit Tests

When you t a parametric distribution, PROC UNIVARIATE provides a series of
goodness-of-t tests based on the empirical distribution function (EDF). The EDF
tests offer advantages over traditional chi-square goodness-of-t test, including improved power and invariance with respect to the histogram midpoints. For a thorough
discussion, refer to DAgostino and Stephens (1986).
The empirical distribution function is dened for a set of n independent observations
X1 , . . . , Xn with a common distribution function F (x). Denote the observations ordered from smallest to largest as X(1) , . . . , X(n) . The empirical distribution function,
Fn (x), is dened as
Fn (x) = 0, x < X(1)
i
Fn (x) = n , X(i) x < X(i+1) i = 1, . . . , n 1
Fn (x) = 1, X(n) x
1
Note that Fn (x) is a step function that takes a step of height n at each observation.
This function estimates the distribution function F (x). At any value x, Fn (x) is the
proportion of observations less than or equal to x, while F (x) is the probability of an
observation less than or equal to x. EDF statistics measure the discrepancy between
Fn (x) and F (x).

The computational formulas for the EDF statistics make use of the probability integral
transformation U = F (X). If F (X) is the distribution function of X, the random
variable U is uniformly distributed between 0 and 1.
Given n observations X(1) , . . . , X(n) , the values U(i) = F (X(i) ) are computed by
applying the transformation, as discussed in the next three sections.
PROC UNIVARIATE provides three EDF tests:
Kolmogorov-Smirnov
Anderson-Darling
Cramr-von Mises
The following sections provide formal denitions of these EDF statistics.
Kolmogorov D Statistic

The Kolmogorov-Smirnov statistic (D) is dened as

D = supx |Fn (x) F (x)|
The Kolmogorov-Smirnov statistic belongs to the supremum class of EDF statistics.
This class of statistics is based on the largest vertical difference between F (x) and
Fn (x).

Goodness-of-Fit Tests
The Kolmogorov-Smirnov statistic is computed as the maximum of D+ and D ,
where D+ is the largest vertical distance between the EDF and the distribution function when the EDF is greater than the distribution function, and D is the largest
vertical distance when the EDF is less than the distribution function.
i
D+ = maxi n U(i)
D = maxi U(i) i1
n
D
= max (D+ , D )

PROC UNIVARIATE uses a modied Kolmogorov D statistic to test the data against
a normal distribution with mean and variance equal to the sample mean and variance.
Anderson-Darling Statistic

The Anderson-Darling statistic and the Cramr-von Mises statistic belong to the
quadratic class of EDF statistics. This class of statistics is based on the squared
difference (Fn (x) F (x))2 . Quadratic statistics have the following general form:
+

(Fn (x) F (x))2 (x)dF (x)

Q=n

The function (x) weights the squared difference (Fn (x) F (x))2 .
The Anderson-Darling statistic (A2 ) is dened as
+

A2 = n

(Fn (x) F (x))2 [F (x) (1 F (x))]1 dF (x)

Here the weight function is (x) = [F (x) (1 F (x))]1 .

The Anderson-Darling statistic is computed as
1
A = n
n

(2i 1) log U(i) + (2n + 1 2i) log(1 U(i) )

i=1

Cramr-von Mises Statistic

The Cramr-von Mises statistic (W 2 ) is dened as

W2 = n

(Fn (x) F (x))2 dF (x)

Here the weight function is (x) = 1.

The Cramr-von Mises statistic is computed as
n
2

U(i)

W =
i=1

2i 1
2n

1
12n

295

296

Chapter 3. The UNIVARIATE Procedure

Probability Values of EDF Tests

Once the EDF test statistics are computed, PROC UNIVARIATE computes the associated probability values (p-values). The UNIVARIATE procedure uses internal tables of probability levels similar to those given by DAgostino and Stephens (1986).
If the value is between two probability levels, then linear interpolation is used to
estimate the probability value.
The probability value depends upon the parameters that are known and the parameters
that are estimated for the distribution. Table 3.62 summarizes different combinations
tted for which EDF tests are available.
Table 3.62. Availability of EDF Tests

Distribution
Beta
Exponential

Gamma

Lognormal

Normal

Weibull

Threshold
known
known
known,
known
unknown
unknown
known
known
known
known
unknown
unknown
unknown
unknown
known
known
known
known
unknown
unknown
unknown
unknown
known
known
unknown
unknown
known
known
known
known
unknown
unknown
unknown
unknown

Parameters
Scale
Shape
known
, known
known
, < 5 unknown
known
unknown
known
unknown
known
known
unknown
known
known
unknown
unknown
unknown
known
> 1 known
unknown
> 1 known
known
> 1 unknown
unknown
> 1 unknown
known
known
known
unknown
unknown
known
unknown
unknown
known
< 3 known
known
< 3 unknown
unknown
< 3 known
unknown
< 3 unknown
known
unknown
known
unknown
known
c known
unknown
c known
known
c unknown
unknown
c unknown
known
c > 2 known
unknown
c > 2 known
known
c > 2 unknown
unknown
c > 2 unknown

Tests Available
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
A2 and W 2
A2 and W 2
all
all
all
all
all
all
A2 and W 2
A2 and W 2
all
all
A2 and W 2
A2 and W 2
A2 and W 2
all
all
all
all

Kernel Density Estimates

You can use the KERNEL option to superimpose kernel density estimates on histograms. Smoothing the data distribution with a kernel density estimate can be more
effective than using a histogram to identify features that might be obscured by the
choice of histogram bins or sampling variation. A kernel density estimate can also
be more effective than a parametric curve t when the process distribution is multimodal. See Example 3.23.
The general form of the kernel density estimator is
100h%

f (x) =
n

K0
i=1

x xi

where K0 () is the kernel function, is the bandwidth, n is the sample size and xi is
the ith observation.
The KERNEL option provides three kernel functions (K0 ): normal, quadratic, and
triangular. You can specify the function with the K= kernel-option in parentheses
after the KERNEL option. Values for the K= option are NORMAL, QUADRATIC,
and TRIANGULAR (with aliases of N, Q, and T, respectively). By default, a normal
kernel is used. The formulas for the kernel functions are
Normal

K0 (t) =

Quadratic

K0 (t) =

1 exp( 1 t2 )
2
2
3
2)
4 (1 t

Triangular K0 (t) = 1 |t|

for < t <

for |t| 1
for |t| 1

The value of , referred to as the bandwidth parameter, determines the degree of

smoothness in the estimated density function. You specify indirectly by specifying
a standardized bandwidth c with the C= kernel-option. If Q is the interquartile range,
and n is the sample size, then c is related to by the formula
1

= cQn 5

For a specic kernel function, the discrepancy between the density estimator f (x)
and the true density f (x) is measured by the mean integrated square error (MISE):

{E(f (x)) f (x)}2 dx +

MISE() =
x

var(f (x))dx
x

The MISE is the sum of the integrated squared bias and the variance. An approximate
mean integrated square error (AMISE) is
1
AMISE() = 4
4

t2 K(t)dt
t

f (x)
x

dx +

1
n

K(t)2 dt
t

297

298

Chapter 3. The UNIVARIATE Procedure

A bandwidth that minimizes AMISE can be derived by treating f (x) as the normal
density having parameters and estimated by the sample mean and standard deviation. If you do not specify a bandwidth parameter or if you specify C=MISE, the
bandwidth that minimizes AMISE is used. The value of AMISE can be used to compare different density estimates. For each estimate, the bandwidth parameter c, the
kernel function type, and the value of AMISE are reported in the SAS log.
The general kernel density estimates assume that the domain of the density to estimate
can take on all values on a real line. However, sometimes the domain of a density is an
interval bounded on one or both sides. For example, if a variable Y is a measurement
of only positive values, then the kernel density curve should be bounded so that is
zero for negative Y values. You can use the LOWER= and UPPER= kernel-options
to specify the bounds.
The UNIVARIATE procedure uses a reection technique to create the bounded kernel
density curve, as described in Silverman (1986, pp. 30-31). It adds the reections
of the kernel density that are outside the boundary to the bounded kernel estimates.
The general form of the bounded kernel density estimator is computed by replacing
K0 xxi in the original equation with

x xi

+ K0

(x xl ) + (xi xl )

+ K0

(xu x) + (xu xi )

where xl is the lower bound and xu is the upper bound.

Without a lower bound, xl = and K0
without an upper bound, xu = and K0

(xxl )+(xi xl )

(xu x)+(xu xi )

equals zero. Similarly,

equals zero.

When C=MISE is used with a bounded kernel density, the UNIVARIATE procedure
uses a bandwidth that minimizes the AMISE for its corresponding unbounded kernel.

Construction of Quantile-Quantile and Probability Plots

Figure 3.10 illustrates how a Q-Q plot is constructed for a specied theoretical distribution. First, the n nonmissing values of the variable are ordered from smallest to
largest:
x(1) x(2) x(n)
Then the ith ordered value x(i) is plotted as a point whose y-coordinate is x(i) and
whose x-coordinate is F 1 i0.375 , where F () is the specied distribution with
n+0.25
zero location parameter and unit scale parameter.
You can modify the adjustment constants 0.375 and 0.25 with the RANKADJ= and
NADJ= options. This default combination is recommended by Blom (1958). For
additional information, refer to Chambers et al. (1983). Since x(i) is a quantile of the
empirical cumulative distribution function (ecdf), a Q-Q plot compares quantiles of
the ecdf with quantiles of a theoretical distribution. Probability plots (see the section
PROBPLOT Statement on page 241) are constructed the same way, except that the
x-axis is scaled nonlinearly in percentiles.

Interpretation of Quantile-Quantile and Probability Plots

x (i)

i th point

-1

F (

i - 0.375
n + 0.25

Figure 3.10. Construction of a Q-Q Plot

Interpretation of Quantile-Quantile and Probability Plots

The following properties of Q-Q plots and probability plots make them useful diagnostics of how well a specied theoretical distribution ts a set of measurements:
If the quantiles of the theoretical and data distributions agree, the plotted points
fall on or near the line y = x.
If the theoretical and data distributions differ only in their location or scale, the
points on the plot fall on or near the line y = ax + b. The slope a and intercept
b are visual estimates of the scale and location parameters of the theoretical
distribution.
Q-Q plots are more convenient than probability plots for graphical estimation of the
location and scale parameters since the x-axis of a Q-Q plot is scaled linearly. On
the other hand, probability plots are more convenient for estimating percentiles or
probabilities.
There are many reasons why the point pattern in a Q-Q plot may not be linear.
Chambers et al. (1983) and Fowlkes (1987) discuss the interpretations of commonly
encountered departures from linearity, and these are summarized in Table 3.63.
In some applications, a nonlinear pattern may be more revealing than a linear pattern.
However, Chambers et al. (1983) note that departures from linearity can also be due
to chance variation.

299

300

Chapter 3. The UNIVARIATE Procedure

Table 3.63. Quantile-Quantile Plot Diagnostics

Description of Point Pattern

Possible Interpretation

All but a few points fall on a line

Outliers in the data

Left end of pattern is below the line;

right end of pattern is above the line

Long tails at both ends of the data

distribution

Left end of pattern is above the line;

right end of pattern is below the line

Short tails at both ends of the data

distribution

Curved pattern with slope increasing

from left to right
Curved pattern with slope decreasing
from left to right
Staircase pattern (plateaus and gaps)

Data distribution is skewed to the right

Data distribution is skewed to the left
Data have been rounded or are discrete

When the pattern is linear, you can use Q-Q plots to estimate shape, location, and
scale parameters and to estimate percentiles. See Example 3.26 through Example
3.34.

Distributions for Probability and Q-Q Plots

You can use the PROBPLOT and QQPLOT statements to request probability and Q-Q
plots that are based on the theoretical distributions summarized in Table 3.64.
Table 3.64. Distributions and Parameters

Parameters
Location Scale Shape

Distribution

Density Function p(x)

Range

Beta

(x)1 (+x)1
B(,) (+1)

<x<+

all x

x > 0

Exponential

exp x

x 1
exp

Gamma

1
()

Lognormal

1
2(x)

exp (log(x))
2 2

(3-parameter)
1

Normal
c

Weibull

exp (x)
2 2

x c1
exp

x c

(3-parameter)
Weibull
(2-parameter)

exp

(known)

Distributions for Probability and Q-Q Plots

You can request these distributions with the BETA, EXPONENTIAL, GAMMA,
LOGNORMAL, NORMAL, WEIBULL, and WEIBULL2 options, respectively. If
you do not specify a distribution option, a normal probability plot or a normal Q-Q
plot is created.
The following sections provide details for constructing Q-Q plots that are based on
these distributions. Probability plots are constructed similarly except that the horizontal axis is scaled in percentile units.

Beta Distribution
To create the plot, the observations are ordered from smallest to largest, and the ith or1
1
dered observation is plotted against the quantile B i0.375 , where B () is the
n+0.25
inverse normalized incomplete beta function, n is the number of nonmissing observations, and and are the shape parameters of the beta distribution. In a probability
plot, the horizontal axis is scaled in percentile units.
The pattern on the plot for ALPHA= and BETA= tends to be linear with intercept
and slope if the data are beta distributed with the specic density function
p(x) =

(x)1 (+x)1
B(,) (+1)

where B(, ) =

()()
(+)

for < x < +

for x or x +

and

= lower threshold parameter

= scale parameter ( > 0)
= rst shape parameter ( > 0)
= second shape parameter ( > 0)

Exponential Distribution
To create the plot, the observations are ordered from smallest to largest, and the ith
ordered observation is plotted against the quantile log 1 i0.375 , where n is
n+0.25
the number of nonmissing observations. In a probability plot, the horizontal axis is
scaled in percentile units.
The pattern on the plot tends to be linear with intercept and slope if the data are
exponentially distributed with the specic density function
p(x) =

exp x

for x
for x <

where is a threshold parameter, and is a positive scale parameter.

Gamma Distribution
To create the plot, the observations are ordered from smallest to largest, and the ith
ordered observation is plotted against the quantile G1 i0.375 , where G1 () is

n+0.25
the inverse normalized incomplete gamma function, n is the number of nonmissing
observations, and is the shape parameter of the gamma distribution. In a probability
plot, the horizontal axis is scaled in percentile units.

301

302

Chapter 3. The UNIVARIATE Procedure

The pattern on the plot for ALPHA= tends to be linear with intercept and slope
if the data are gamma distributed with the specic density function
p(x) =

1
()

x 1
exp

for x >
for x

where
= threshold parameter
= scale parameter ( > 0)
= shape parameter ( > 0)

Lognormal Distribution
To create the plot, the observations are ordered from smallest to largest, and the
ith ordered observation is plotted against the quantile exp 1 i0.375 , where
n+0.25
1 () is the inverse cumulative standard normal distribution, n is the number of
nonmissing observations, and is the shape parameter of the lognormal distribution.
In a probability plot, the horizontal axis is scaled in percentile units.
The pattern on the plot for SIGMA= tends to be linear with intercept and slope
exp() if the data are lognormally distributed with the specic density function
p(x) =

1
2(x)

exp (log(x))
2 2

for x >
for x

0
where
= threshold parameter
= scale parameter
= shape parameter ( > 0)
See Example 3.26 and Example 3.33.

Normal Distribution
To create the plot, the observations are ordered from smallest to largest, and the ith
ordered observation is plotted against the quantile 1 i0.375 , where 1 () is the
n+0.25
inverse cumulative standard normal distribution, and n is the number of nonmissing
observations. In a probability plot, the horizontal axis is scaled in percentile units.
The point pattern on the plot tends to be linear with intercept and slope if the data
are normally distributed with the specic density function
p(x) =

exp (x)
2 2

for all x

where is the mean, and is the standard deviation ( > 0).

Three-Parameter Weibull Distribution

To create the plot, the observations are ordered from smallest to largest, and the ith
1
c

ordered observation is plotted against the quantile log 1 i0.375

, where
n+0.25
n is the number of nonmissing observations, and c is the Weibull distribution shape
parameter. In a probability plot, the horizontal axis is scaled in percentile units.

Estimating Shape Parameters Using Q-Q Plots

The pattern on the plot for C=c tends to be linear with intercept and slope if the
data are Weibull distributed with the specic density function
p(x) =

x c1
exp

x c

for x >
for x

0
where
= threshold parameter
= scale parameter ( > 0)
c = shape parameter (c > 0)
See Example 3.34.

Two-Parameter Weibull Distribution

To create the plot, the observations are ordered from smallest to largest, and the log of
the shifted ith ordered observation x(i) , denoted by log(x(i) 0 ), is plotted against
the quantile log log 1 i0.375 , where n is the number of nonmissing obsern+0.25
vations. In a probability plot, the horizontal axis is scaled in percentile units.
Unlike the three-parameter Weibull quantile, the preceding expression is free of distribution parameters. Consequently, the C= shape parameter is not mandatory with
the WEIBULL2 distribution option.
The pattern on the plot for THETA=0 tends to be linear with intercept log() and
slope 1 if the data are Weibull distributed with the specic density function
c
p(x) =

exp

for x > 0
for x 0

where
0 = known lower threshold
= scale parameter ( > 0)
c = shape parameter (c > 0)
See Example 3.34.

Estimating Shape Parameters Using Q-Q Plots

Some of the distribution options in the PROBPLOT or QQPLOT statements require
you to specify one or two shape parameters in parentheses after the distribution keyword. These are summarized in Table 3.65.
You can visually estimate the value of a shape parameter by specifying a list of values for the shape parameter option. A separate plot is produced for each value, and
you can then select the value of the shape parameter that produces the most nearly
linear point pattern. Alternatively, you can request that the plot be created using an
estimated shape parameter. See the entries for the distribution options in the section Dictionary of Options on page 245 (for the PROBPLOT statement) and in the
section Dictionary of Options on page 258 (for the QQPLOT statement).
Note: For Q-Q plots created with the WEIBULL2 option, you can estimate the shape
parameter c from a linear pattern using the fact that the slope of the pattern is 1 .
c

303

304

Chapter 3. The UNIVARIATE Procedure

Table 3.65. Shape Parameter Options

Distribution Keyword
BETA

Mandatory Shape Parameter Option

ALPHA=, BETA=

Range
> 0, > 0

EXPONENTIAL

None

GAMMA

ALPHA=

LOGNORMAL

SIGMA=

NORMAL

None

WEIBULL

C=c

WEIBULL2

None

c>0

Estimating Location and Scale Parameters Using Q-Q Plots

If you specify location and scale parameters for a distribution in a PROBPLOT or
QQPLOT statement (or if you request estimates for these parameters), a diagonal distribution reference line is displayed on the plot. (An exception is the two-parameter
Weibull distribution, for which a line is displayed when you specify or estimate the
scale and shape parameters.) Agreement between this line and the point pattern indicates that the distribution with these parameters is a good t.
When the point pattern on a Q-Q plot is linear, its intercept and slope provide estimates of the location and scale parameters. (An exception to this rule is the twoparameter Weibull distribution, for which the intercept and slope are related to the
scale and shape parameters.)
Table 3.66 shows how the specied parameters determine the intercept and slope of
the line. The intercept and slope are based on the quantile scale for the horizontal
axis, which is used in Q-Q plots.
Table 3.66. Intercept and Slope of Distribution Reference Line

Distribution
Beta
Exponential
Gamma
Lognormal
Normal
Weibull (3-parameter)
Weibull (2-parameter)

Parameters
Location
Scale

0 (known)

Shape
,

c
c

Linear Pattern
Intercept Slope

exp()

1
log()
c

For instance, specifying MU=3 and SIGMA=2 with the NORMAL option requests
a line with intercept 3 and slope 2. Specifying SIGMA=1 and C=2 with the
1
WEIBULL2 option requests a line with intercept log(1) = 0 and slope 2 . On a
probability plot with the LOGNORMAL and WEIBULL2 options, you can specify
the slope directly with the SLOPE= option. That is, for the LOGNORMAL option,
specifying THETA= 0 and SLOPE=exp(0 ) displays the same line as specifying
THETA= 0 and ZETA= 0 . For the WEIBULL2 option, specifying SIGMA= 0 and
SLOPE= c1 displays the same line as specifying SIGMA= 0 and C= c0 .
0

Input Data Sets

Estimating Percentiles Using Q-Q Plots

There are two ways to estimate percentiles from a Q-Q plot:
Specify the PCTLAXIS option, which adds a percentile axis opposite the theoretical quantile axis. The scale for the percentile axis ranges between 0 and
100 with tick marks at percentile values such as 1, 5, 10, 25, 50, 75, 90, 95,
and 99.
Specify the PCTLSCALE option, which relabels the horizontal axis tick marks
with their percentile equivalents but does not alter their spacing. For example,
on a normal Q-Q plot, the tick mark labeled 0 is relabeled as 50 since the
50th percentile corresponds to the zero quantile.
You can also estimate percentiles using probability plots created with the PROBPLOT
statement. See Example 3.32.

Input Data Sets

DATA= Data Set
The DATA= data set provides the set of variables that are analyzed. The
UNIVARIATE procedure must have a DATA= data set. If you do not specify one
with the DATA= option in the PROC UNIVARIATE statement, the procedure uses
the last data set created.

ANNOTATE= Data Sets

You can add features to plots by specifying ANNOTATE= data sets either in the
PROC UNIVARIATE statement or in individual plot statements.
Information contained in an ANNOTATE= data set specied in the PROC
UNIVARIATE statement is used for all plots produced in a given PROC step; this
is a global ANNOTATE= data set. By using this global data set, you can keep
information common to all high-resolution plots in one data set.
Information contained in the ANNOTATE= data set specied in a plot statement is
used only for plots produced by that statement; this is a local ANNOTATE= data
set. By using this data set, you can add statement-specic features to plots. For
example, you can add different features to plots produced by the HISTOGRAM and
QQPLOT statements by specifying an ANNOTATE= data set in each plot statement.
You can specify an ANNOTATE= data set in the PROC UNIVARIATE statement and
in plot statements. This enables you to add some features to all plots and also add
statement-specic features to plots. See Example 3.25.

305

306

Chapter 3. The UNIVARIATE Procedure

OUT= Output Data Set in the OUTPUT Statement

PROC UNIVARIATE creates an OUT= data set for each OUTPUT statement. This
data set contains an observation for each combination of levels of the variables in the
BY statement, or a single observation if you do not specify a BY statement. Thus
the number of observations in the new data set corresponds to the number of groups
for which statistics are calculated. Without a BY statement, the procedure computes
statistics and percentiles by using all the observations in the input data set. With a BY
statement, the procedure computes statistics and percentiles by using the observations
within each BY group.
The variables in the OUT= data set are as follows:
BY statement variables. The values of these variables match the values in the
corresponding BY group in the DATA= data set and indicate which BY group
each observation summarizes.
variables created by selecting statistics in the OUTPUT statement. The statistics are computed using all the nonmissing data, or they are computed for each
BY group if you use a BY statement.
variables created by requesting new percentiles with the PCTLPTS= option.
The names of these new variables depend on the values of the PCTLPRE= and
PCTLNAME= options.
If the output data set contains a percentile variable or a quartile variable, the percentile denition assigned with the PCTLDEF= option in the PROC UNIVARIATE
statement is recorded in the output data set label. See Example 3.8.
The following table lists variables available in the OUT= data set.
Table 3.67. Variables Available in the OUT= Data Set

Variable Name Description

Descriptive Statistics
CSS
Sum of squares corrected for the mean
CV
Percent coefcient of variation
KURTOSIS
Measurement of the heaviness of tails
MAX
Largest (maximum) value
MEAN
Arithmetic mean
MIN
Smallest (minimum) value
MODE
Most frequent value (if not unique, the smallest mode)
N
Number of observations on which calculations are based
NMISS
Number of missing observations
NOBS
Total number of observations
RANGE
Difference between the maximum and minimum values
SKEWNESS
Measurement of the tendency of the deviations to be larger in one direction than in the other
STD
Standard deviation
STDMEAN
Standard error of the mean
SUM
Sum
SUMWGT
Sum of the weights

OUT= Output Data Set in the OUTPUT Statement

Table 3.67. (continued)

Variable Name Description

USS
Uncorrected sum of squares
VAR
Variance
Quantile Statistics
MEDIAN|P50
Middle value (50th percentile)
P1
1st percentile
P5
5th percentile
P10
10th percentile
P90
90th percentile
P95
95th percentile
P99
99th percentile
Q1|P25
Lower quartile (25th percentile)
Q3|P75
Upper quartile (75th percentile)
QRANGE
Difference between the upper and lower quartiles (also known as the
inner quartile range)
Robust Statistics
GINI
Ginis mean difference
MAD
Median absolute difference
QN
2nd variation of median absolute difference
SN
1st variation of median absolute difference
STD GINI
Standard deviation for Ginis mean difference
STD MAD
Standard deviation for median absolute difference
STD QN
Standard deviation for the second variation of the median absolute difference
STD QRANGE Estimate of the standard deviation, based on interquartile range
STD SN
Standard deviation for the rst variation of the median absolute difference
Hypothesis Test Statistics
MSIGN
Sign statistic
NORMAL
Test statistic for normality. If the sample size is less than or equal
to 2000, this is the Shapiro-Wilk W statistic. Otherwise, it is the
Kolmogorov D statistic.
PROBM
Probability of a greater absolute value for the sign statistic
PROBN
Probability that the data came from a normal distribution
PROBS
Probability of a greater absolute value for the signed rank statistic
PROBT
Two-tailed p-value for Students t statistic with n 1 degrees of freedom
SIGNRANK
Signed rank statistic
T
Students t statistic to test the null hypothesis that the population mean
is equal to 0

307

308

Chapter 3. The UNIVARIATE Procedure

OUTHISTOGRAM= Output Data Set

You can create an OUTHISTOGRAM= data set with the HISTOGRAM statement.
This data set contains information about histogram intervals. Since you can specify
multiple HISTOGRAM statements with the UNIVARIATE procedure, you can create
multiple OUTHISTOGRAM= data sets.
An OUTHISTOGRAM= data set contains a group of observations for each variable
in the HISTOGRAM statement. The group contains an observation for each interval
of the histogram, beginning with the leftmost interval that contains a value of the variable and ending with the rightmost interval that contains a value of the variable. These
intervals will not necessarily coincide with the intervals displayed in the histogram
since the histogram may be padded with empty intervals at either end. If you superimpose one or more tted curves on the histogram, the OUTHISTOGRAM= data set
contains multiple groups of observations for each variable (one group for each curve).
If you use a BY statement, the OUTHISTOGRAM= data set contains groups of observations for each BY group. ID variables are not saved in an OUTHISTOGRAM=
data set.
By default, an OUTHISTOGRAM= data set contains the MIDPT variable, whose
values identify histogram intervals by their midpoints. When the ENDPOINTS= or
NENDPOINTS option is specied, intervals are identied by endpoint values instead.
If the RTINCLUDE option is specied, the MAXPT variable contains upper endpoint values. Otherwise, the MINPT variable contains lower endpoint values. See
Example 3.18.
Table 3.68. Variables in the OUTHISTOGRAM= Data Set

Variable
CURVE

Description
Name of tted distribution (if requested in HISTOGRAM
statement)

EXPPCT

Estimated percent of population in histogram interval determined

from optional tted distribution

MAXPT
MIDPT

Upper endpoint of histogram interval

Midpoint of histogram interval

MINPT
OBSPCT

Percent of variable values in histogram interval

VAR

Variable name

Lower endpoint of histogram interval

Tables for Summary Statistics

By default, PROC UNIVARIATE produces ODS tables of moments, basic statistical
measures, tests for location, quantiles, and extreme observations. You must specify
options in the PROC UNIVARIATE statement to request other statistics and tables.
The CIBASIC option produces a table that displays condence limits for the mean,
standard deviation, and variance. The CIPCTLDF and CIPCTLNORMAL options
request tables of condence limits for the quantiles. The LOCCOUNT option requests a table that shows the number of values greater than, not equal to, and less
than the value of MU0=. The FREQ option requests a table of frequencies counts.

ODS Table Names

The NEXTRVAL= option requests a table of extreme values. The NORMAL option
requests a table with tests for normality.
The TRIMMED=, WINSORIZED=, and ROBUSTCALE options request tables with
robust estimators. The table of trimmed or Winsorized means includes the percentage
and the number of observations that are trimmed or Winsorized at each end, the mean
and standard error, condence limits, and the Students t test. The table with robust
measures of scale includes interquartile range, Ginis mean difference G, MAD, Qn ,
and Sn , with their corresponding estimates of .
See the section ODS Table Names on page 309 for the names of ODS tables created
by PROC UNIVARIATE.

ODS Table Names

PROC UNIVARIATE assigns a name to each table that it creates. You can use these
names to reference the table when using the Output Delivery System (ODS) to select
tables and create output data sets.
Table 3.69. ODS Tables Produced with the PROC UNIVARIATE Statement

ODS Table Name

BasicIntervals
BasicMeasures
ExtremeObs
ExtremeValues
Frequencies
LocationCounts
MissingValues
Modes
Moments
Plots
Quantiles
RobustScale
SSPlots
TestsForLocation
TestsForNormality
TrimmedMeans
WinsorizedMeans

Description
Condence intervals for mean,
standard deviation, variance
Measures of location and variability
Extreme observations
Extreme values
Frequencies
Counts used for sign test and
signed rank test
Missing values
Modes
Sample moments
Line printer plots
Quantiles
Robust measures of scale
Line printer side-by-side box
plots
Tests for location
Tests for normality
Trimmed means
Winsorized means

Option
CIBASIC
Default
Default
NEXTRVAL=
FREQ
LOCCOUNT
Default, if missing values exist
MODES
Default
PLOTS
Default
ROBUSTSCALE
PLOTS (with BY statement)
Default
NORMALTEST
TRIMMED=
WINSORIZED=

309

310

Chapter 3. The UNIVARIATE Procedure

Table 3.70. ODS Tables Produced with the HISTOGRAM Statement

ODS Table Name

Bins

Description
Histogram bins

FitQuantiles
GoodnessOfFit

Quantiles of tted distribution

Goodness-of-t tests for tted
distribution
Histogram bins
Parameter estimates for tted
distribution

HistogramBins
ParameterEstimates

Option
MIDPERCENTS secondary option
Any distribution option
Any distribution option
MIDPERCENTS option
Any distribution option

ODS Tables for Fitted Distributions

If you request a tted parametric distribution with a HISTOGRAM statement, PROC
UNIVARIATE creates a summary that is organized into the ODS tables described in
this section.

Parameters
The ParameterEstimates table lists the estimated (or specied) parameters for the
tted curve as well as the estimated mean and estimated standard deviation. See
Formulas for Fitted Continuous Distributions on page 288.

EDF Goodness-of-Fit Tests

When you t a parametric distribution, the HISTOGRAM statement provides a series
of goodness-of-t tests based on the empirical distribution function (EDF). See EDF
Goodness-of-Fit Tests on page 294. These are displayed in the GoodnessOfFit table.

Histogram Intervals
The Bins table is included in the summary only if you specify the MIDPERCENTS
option in parentheses after the distribution option. This table lists the midpoints for
the histogram bins along with the observed and estimated percentages of the observations that lie in each bin. The estimated percentages are based on the tted
distribution.
If you specify the MIDPERCENTS option without requesting a tted distribution,
the HistogramBins table is included in the summary. This table lists the interval
midpoints with the observed percent of observations that lie in the interval. See the
entry for the MIDPERCENTS option on page 225.

Quantiles
The FitQuantiles table lists observed and estimated quantiles. You can use the
PERCENTS= option to specify the list of quantiles in this table. See the entry for the
PERCENTS= option on page 227. By default, the table lists observed and estimated
quantiles for the 1, 5, 10, 25, 50, 75, 90, 95, and 99 percent of a tted parametric
distribution.

Computational Resources

Computational Resources
Because the UNIVARIATE procedure computes quantile statistics, it requires additional memory to store a copy of the data in memory. By default, the MEANS,
SUMMARY, and TABULATE procedures require less memory because they do not
automatically compute quantiles. These procedures also provide an option to use a
new xed-memory quantiles estimation method that is usually less memory intensive.
In the UNIVARIATE procedure, the only factor that limits the number of variables
that you can analyze is the computer resources that are available. The amount of
temporary storage and CPU time required depends on the statements and the options
that you specify. To calculate the computer resources the procedure needs, let
N
V
Ui

be the number of observations in the data set

be the number of variables in the VAR statement
be the number of unique values for the ith variable

Then the minimum memory requirement in bytes to process all variables is M =

24 i Ui . If M bytes are not available, PROC UNIVARIATE must process the data
multiple times to compute all the statistics. This reduces the minimum memory requirement to M = 24 max(Ui ).
Using the ROUND= option reduces the number of unique values (Ui ), thereby reducing memory requirements. The ROBUSTSCALE option requires 40Ui bytes of
temporary storage.
Several factors affect the CPU time:
The time to create V tree structures to internally store the observations is proportional to N V log(N ).
The time to compute moments and quantiles for the ith variable is proportional
to Ui .
The time to compute the NORMAL option test statistics is proportional to N .
The time to compute the ROBUSTSCALE option test statistics is proportional
to Ui log(Ui ).
The time to compute the exact signicance level of the sign rank statistic may
increase when the number of nonzero values is less than or equal to 20.
Each of these factors has a different constant of proportionality. For additional information on optimizing CPU performance and memory usage, see the SAS documentation for your operating environment.

311

312

Chapter 3. The UNIVARIATE Procedure

Examples
Example 3.1. Computing Descriptive Statistics for Multiple
Variables
This example computes univariate statistics for two variables. The following statements create the data set BPressure, which contains the systolic (Systolic) and diastolic (Diastolic) blood pressure readings for 22 patients:
data BPressure;
length PatientID $2;
input PatientID $ Systolic Diastolic @@;
datalines;
CK 120 50 SS 96 60 FR 100 70
CP 120 75 BL 140 90 ES 120 70
CP 165 110 JI 110 40 MC 119 66
FC 125 76 RW 133 60 KD 108 54
DS 110 50 JW 130 80 BH 120 65
JW 134 80 SB 118 76 NS 122 78
GS 122 70 AB 122 78 EC 112 62
HH 122 82
;
run;

The following statements produce descriptive statistics and quantiles for the variables
Systolic and Diastolic:
title Systolic and Diastolic Blood Pressure;
ods select BasicMeasures Quantiles;
proc univariate data=BPressure;
var Systolic Diastolic;
run;

The ODS SELECT statement restricts the output, which is shown in Output 3.1.1, to
the BasicMeasures and Quantiles tables; see the section ODS Table Names
on page 309. You use the PROC UNIVARIATE statement to request univariate
statistics for the variables listed in the VAR statement, which species the analysis
variables and their order in the output. Formulas for computing the statistics in the
BasicMeasures table are provided in the section Descriptive Statistics on page
269. The quantiles are calculated using Denition 5, which is the default denition;
see the section Calculating Percentiles on page 273.
A sample program, uniex01.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.1. Computing Descriptive Statistics for Multiple Variables

Output 3.1.1. Display Basic Measures and Quantiles

Systolic and Diastolic Blood Pressure
The UNIVARIATE Procedure
Variable: Systolic
Basic Statistical Measures
Location
Mean
Median
Mode

Variability

121.2727
120.0000
120.0000

Std Deviation
Variance
Range
Interquartile Range

14.28346
204.01732
69.00000
13.00000

NOTE: The mode displayed is the smallest of 2 modes with a count of 4.

Quantiles (Definition 5)
Quantile
100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

Estimate
165
165
140
134
125
120
112
108
100
96
96

Systolic and Diastolic Blood Pressure

The UNIVARIATE Procedure
Variable: Diastolic
Basic Statistical Measures
Location
Mean
Median
Mode

Variability

70.09091
70.00000
70.00000

Std Deviation
Variance
Range
Interquartile Range

Quantiles (Definition 5)
Quantile
100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

Estimate
110
110
90
82
78
70
60
50
50
40
40

15.16547
229.99134
70.00000
18.00000

313

314

Chapter 3. The UNIVARIATE Procedure

Example 3.2. Calculating Modes

An instructor is interested in calculating all the modes of the scores on a recent exam.
The following statements create a data set named Exam, which contains the exam
scores in the variable Score:
data Exam;
label Score
input Score
datalines;
81 97 78 99 77
85 86 94 76 75
97 97 89 69 72
;
run;

= Exam Score;
@@;
81 84 86 86 97
42 91 90 88 86
82 83 81 80 81

The following statements use the MODES option to request a table of all possible
modes:
title Table of Modes for Exam Scores;
ods select Modes;
proc univariate data=Exam modes;
var Score;
run;

The ODS SELECT statement restricts the output to the Modes table; see the section
ODS Table Names on page 309.
Output 3.2.1. Table of Modes Display
Table of Modes for Exam Scores
The UNIVARIATE Procedure
Variable: Score (Exam Score)
Modes
Mode

Count

81
86
97

4
4
4

By default, when the MODES option is used, and there is more than one mode, the
lowest mode is displayed in the BasicMeasures table. The following statements
illustrate the default behavior:
title Default Output;
ods select BasicMeasures;
proc univariate data=Exam;
var Score;
run;

Example 3.3. Identifying Extreme Observations and Extreme Values

Output 3.2.2. Default Output (Without MODES Option)

Default Output
The UNIVARIATE Procedure
Variable: Score (Exam Score)
Basic Statistical Measures
Location
Mean
Median
Mode

83.66667
84.50000
81.00000

Variability
Std Deviation
Variance
Range
Interquartile Range

11.08069
122.78161
57.00000
10.00000

NOTE: The mode displayed is the smallest of 3 modes with a count of 4.

The default output displays a mode of 81 and includes a note regarding the number of
modes; the modes 86 and 97 are not displayed. The ODS SELECT statement restricts
the output to the BasicMeasures table; see the section ODS Table Names on page
309.
A sample program, uniex02.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.3. Identifying Extreme Observations and Extreme

Values
This example, which uses the data set BPressure introduced in Example 3.1, illustrates how to produce a table of the extreme observations and a table of the extreme
values in a data set. The following statements generate the Extreme Observations
tables for Systolic and Diastolic, which enable you to identify the extreme observations for each variable:
title Extreme Blood Pressure Observations;
ods select ExtremeObs;
proc univariate data=BPressure;
var Systolic Diastolic;
id PatientID;
run;

The ODS SELECT statement restricts the output to the ExtremeObs table; see
the section ODS Table Names on page 309. The ID statement requests that the
extreme observations are to be identied using the value of PatientID as well as the
observation number. By default, the ve lowest and ve highest observations are
displayed. You can use the NEXTROBS= option to request a different number of
extreme observations.
Output 3.3.1 shows that the patient identied as CP (Observation 7) has the highest
values for both Systolic and Diastolic. To visualize extreme observations, you can
create histograms; see Example 3.14.

315

316

Chapter 3. The UNIVARIATE Procedure

Output 3.3.1. Blood Pressure Extreme Observations

Extreme Blood Pressure Observations
The UNIVARIATE Procedure
Variable: Systolic
Extreme Observations
---------Lowest---------

Value
96
100
108
110
110

Patient
ID
SS
FR
KD
DS
JI

---------Highest--------

Obs

Value

2
3
12
13
8

130
133
134
140
165

Patient
ID
JW
RW
JW
BL
CP

Obs
14
11
16
5
7

Extreme Blood Pressure Observations

The UNIVARIATE Procedure
Variable: Diastolic
Extreme Observations
---------Lowest---------

Value
40
50
50
54
60

Patient
ID
JI
DS
CK
KD
RW

---------Highest--------

Obs

Value

8
13
1
12
11

80
80
82
90
110

Patient
ID
JW
JW
HH
BL
CP

Obs
14
16
22
5
7

The following statements generate the Extreme Values tables for Systolic and
Diastolic, which tabulate the tails of the distributions:
title Extreme Blood Pressure Values;
ods select ExtremeValues;
proc univariate data=BPressure nextrval=5;
var Systolic Diastolic;
run;

The ODS SELECT statement restricts the output to the ExtremeValues table; see
the section ODS Table Names on page 309. The NEXTRVAL= option species the
number of extreme values at each end of the distribution to be shown in the tables in
Output 3.3.2.
Output 3.3.2 shows that the values 78 and 80 occurred twice for Diastolic and the
maximum of Diastolic is 110. Note that Output 3.3.1 displays the value of 80 twice
for Diastolic because there are two observations with that value. In Output 3.3.2, the
value 80 is only displayed once.

Example 3.4. Creating a Frequency Table

Output 3.3.2. Blood Pressure Extreme Values

Extreme Blood Pressure Values
The UNIVARIATE Procedure
Variable: Systolic
Extreme Values
---------Lowest--------

--------Highest--------

Order

Value

Freq

Order

Value

Freq

1
2
3
4
5

96
100
108
110
112

1
1
1
2
1

11
12
13
14
15

130
133
134
140
165

1
1
1
1
1

Extreme Blood Pressure Values

The UNIVARIATE Procedure
Variable: Diastolic
Extreme Values
---------Lowest--------

--------Highest--------

Order

Value

Freq

Order

Value

Freq

1
2
3
4
5

40
50
54
60
62

1
2
1
2
1

11
12
13
14
15

78
80
82
90
110

2
2
1
1
1

A sample program, uniex01.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.4. Creating a Frequency Table

An instructor is interested in creating a frequency table of score changes between a
pair of tests given in one of his college courses. The data set Score contains test
scores for his students who took a pretest and a posttest on the same material. The
variable ScoreChange contains the difference between the two test scores. The
following statements create the data set:

317

318

Chapter 3. The UNIVARIATE Procedure

data Score;
input Student $ PreTest PostTest @@;
label ScoreChange = Change in Test Scores;
ScoreChange = PostTest - PreTest;
datalines;
Capalleti 94 91 Dubose
51 65
Engles
95 97 Grant
63 75
Krupski
80 75 Lundsford 92 55
Mcbane
75 78 Mullen
89 82
Nguyen
79 76 Patel
71 77
Si
75 70 Tanaka
87 73
;
run;

The following statements produce a frequency table for the variable ScoreChange:
title Analysis of Score Changes;
ods select Frequencies;
proc univariate data=Score freq;
var ScoreChange;
run;

The ODS SELECT statement restricts the output to the Frequencies table; see
the section ODS Table Names on page 309. The FREQ option on the PROC
UNIVARIATE statement requests the table of frequencies shown in Output 3.4.1.
Output 3.4.1. Table of Frequencies
Analysis of Score Changes

Variable:

The UNIVARIATE Procedure

ScoreChange (Change in Test Scores)
Frequency Counts

Value Count
-37
-14
-7
-5

1
1
1
2

Percents
Cell
Cum
8.3
8.3
8.3
16.7

8.3
16.7
25.0
41.7

Value Count
-3
2
3

2
1
1

Percents
Cell
Cum
16.7
8.3
8.3

58.3
66.7
75.0

Value Count
6
12
14

1
1
1

Percents
Cell
Cum
8.3 83.3
8.3 91.7
8.3 100.0

From Output 3.4.1, the instructor sees that only score changes of 3 and 5 occurred
more than once.
A sample program, uniex03.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.5. Creating Plots for Line Printer Output

The PLOT option in the PROC UNIVARIATE statement requests several basic plots
for display in line printer output. For more information on plots created by the PLOT
option, see the section Creating Line Printer Plots on page 281. This example illustrates the use of the PLOT option as well as BY processing in PROC UNIVARIATE.
A researcher is analyzing a data set consisting of air pollution data from three different measurement sites. The data set AirPoll, created by the following statements,
contains the variables Site and Ozone, which are the site number and ozone level,
respectively.
data AirPoll (keep = Site Ozone);
label Site = Site Number
Ozone = Ozone level (in ppb);
do i = 1 to 3;
input Site @@;
do j = 1 to 15;
input Ozone @@;
output;
end;
end;
datalines;
102 4 6 3 4 7 8 2 3 4 1 3 8 9 5 6
134 5 3 6 2 1 2 4 3 2 4 6 4 6 3 1
137 8 9 7 8 6 7 6 7 9 8 9 8 7 8 5
;
run;

The following statements produce stem-and-leaf plots, box plots, and normal probability plots for each site in the AirPoll data set:
ods select Plots SSPlots;
proc univariate data=AirPoll plot;
by Site;
var Ozone;
run;

The PLOT option produces a stem-and-leaf plot, a box plot, and a normal probability
plot for the Ozone variable at each site. Since the BY statement is used, a side-byside box plot is also created to compare the ozone levels across sites. Note that AirPoll
is sorted by Site; in general, the data set should be sorted by the BY variable using
the SORT procedure. The ODS SELECT statement restricts the output to the Plots
and SSPlots tables; see the section ODS Table Names on page 309. Optionally,
you can specify the PLOTSIZE=n option to control the approximate number of rows
(between 8 and the page size) that the plots occupy.
Output 3.5.1 through Output 3.5.3 show the plots produced for each BY group.
Output 3.5.4 shows the side-by-side box plot for comparing Ozone values across
sites.

319

320

Chapter 3. The UNIVARIATE Procedure

Leaf
0
00
0
00
0
000
000
0
0
----+----+----+----+

#
1
2
1
2
1
3
3
1
1

Note that you can use the PROBPLOT statement with the NORMAL option to produce high-resolution normal probability plots; see the section Modeling a Data
Distribution on page 200.
Note that you can use the BOXPLOT procedure to produce box plots using highresolution graphics; refer to Chapter 18, The BOXPLOT Procedure, in SAS/STAT
Users Guide.
A sample program, uniex04.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.6. Analyzing a Data Set With a FREQ Variable

This example illustrates how to use PROC UNIVARIATE to analyze a data set with
a variable that contains the frequency of each observation. The data set Speeding
contains data on the number of cars pulled over for speeding on a stretch of highway
with a 65 mile per hour speed limit. Speed is the speed at which the cars were
traveling, and Number is the number of cars at each speed. The following statements
create the data set:
data Speeding;
label Speed = Speed (in miles per hour);
do Speed = 66 to 85;
input Number @@;
output;
end;
datalines;
2 3 2 1 3 6 8 9 10 13
12 14 6 2 0 0 1 1 0 1
;
run;

The following statements create a table of moments for the variable Speed:
title Analysis of Speeding Data;
ods select Moments;
proc univariate data=Speeding;
freq Number;
var Speed;
run;

The ODS SELECT statement restricts the output, which is shown in Output 3.6.1, to
the Moments table; see the section ODS Table Names on page 309. The FREQ
statement species that the value of the variable Number represents the frequency of
each observation.
For the formulas used to compute these moments, see the section Descriptive
Statistics on page 269. A sample program, uniex05.sas, for this example is available in the SAS Sample Library for Base SAS software.

Example 3.7. Saving Summary Statistics in an OUT= Output Data Set

Output 3.6.1. Table of Moments

Analysis of Speeding Data

Variable:

The UNIVARIATE Procedure

Speed (Speed (in miles per hour))
Freq:

Number

Moments
N
Mean
Std Deviation
Skewness
Uncorrected SS
Coeff Variation

94
74.3404255
3.44403237
-0.1275543
520594
4.63278538

Sum Weights
Sum Observations
Variance
Kurtosis
Corrected SS
Std Error Mean

94
6988
11.861359
0.92002287
1103.10638
0.35522482

Example 3.7. Saving Summary Statistics in an OUT= Output

Data Set
This example illustrates how to save summary statistics in an output data set. The
following statements create a data set named Belts, which contains the breaking
strengths (Strength) and widths (Width) of a sample of 50 automotive seat belts:
data Belts;
label Strength = Breaking Strength (lb/in)
Width
= Width in Inches;
input Strength Width @@;
datalines;
1243.51 3.036 1221.95 2.995 1131.67 2.983
1198.08 3.106 1273.31 2.947 1250.24 3.018
1126.78 2.965 1174.62 3.033 1250.79 2.941
1285.30 2.893 1214.14 3.035 1270.24 2.957
1166.02 3.067 1278.85 3.037 1280.74 2.984
1101.73 2.961 1165.79 3.075 1186.19 3.058
1213.62 2.984 1213.93 3.029 1289.59 2.956
1247.48 3.027 1284.34 3.073 1209.09 3.004
1224.03 2.915 1200.43 2.974 1183.42 3.033
1258.31 2.958 1136.05 3.022 1177.44 3.090
1183.67 3.045 1206.50 3.024 1195.69 3.005
1147.47 2.944 1171.76 3.005 1207.28 3.065
1215.92 3.003 1202.17 3.058
;
run;

1129.70
1225.47
1216.75
1249.55
1201.96
1124.46
1208.27
1146.78
1195.66
1246.13
1223.49
1131.33

3.019
2.980
3.037
2.958
3.002
2.929
3.029
3.061
2.995
3.022
2.971
2.984

The following statements produce two output data sets containing summary statistics:
proc univariate data=Belts noprint;
var Strength Width;
output out=Means
mean=StrengthMean WidthMean;
output out=StrengthStats mean=StrengthMean std=StrengthSD
min=StrengthMin
max=StrengthMax;
run;

323

324

Chapter 3. The UNIVARIATE Procedure

When you specify an OUTPUT statement, you must also specify a VAR statement.
You can use multiple OUTPUT statements with a single procedure statement. Each
OUTPUT statement creates a new data set with the name specied by the OUT=
option. In this example, two data sets, Means and StrengthStats, are created. See
Output 3.7.1 for a listing of Means and Output 3.7.2 for a listing of StrengthStats.
Output 3.7.1. Listing of Output Data Set Means
Obs
1

Strength
Mean
1205.75

Width
Mean
3.00584

Output 3.7.2. Listing of Output Data Set StrengthStats

Obs
1

Strength
Mean

Strength
SD

Strength
Max

Strength
Min

1205.75

48.3290

1289.59

1101.73

Summary statistics are saved in an output data set by specifying keyword=names

after the OUT= option. In the preceding statements, the rst OUTPUT statement species the keyword MEAN followed by the names StrengthMean and
WidthMean. The second OUTPUT statement species the keywords MEAN, STD,
MAX, and MIN, for which the names StrengthMean, StrengthSD, StrengthMax,
and StrengthMin are given.
The keyword species the statistic to be saved in the output data set, and the names
determine the names for the new variables. The rst name listed after a keyword
contains that statistic for the rst variable listed in the VAR statement; the second
name contains that statistic for the second variable in the VAR statement, and so on.
The data set Means contains the mean of Strength in a variable named
StrengthMean and the mean of Width in a variable named WidthMean. The
data set StrengthStats contains the mean, standard deviation, maximum value,
and minimum value of Strength in the variables StrengthMean, StrengthSD,
StrengthMax, and StrengthMin, respectively.
See the section OUT= Output Data Set in the OUTPUT Statement on page 306 for
more information on OUT= output data sets.
A sample program, uniex06.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.8. Saving Percentiles in an Output Data Set

This example, which uses the Belts data set from the previous example, illustrates
how to save percentiles in an output data set. The UNIVARIATE procedure automatically computes the 1st, 5th, 10th, 25th, 75th, 90th, 95th, and 99th percentiles for
each variable. You can save these percentiles in an output data set by specifying the
appropriate keywords. For example, the following statements create an output data
set named PctlStrength, which contains the 5th and 95th percentiles of the variable
Strength:
proc univariate data=Belts noprint;
var Strength Width;
output out=PctlStrength p5=p5str p95=p95str;
run;

The output data set PctlStrength is listed in Output 3.8.1.

Output 3.8.1. Listing of Output Data Set PctlStrength
Obs

p95str

p5str

1284.34

1126.78

You can use the PCTLPTS=, PCTLPRE=, and PCTLNAME= options to save percentiles not automatically computed by the UNIVARIATE procedure. For example,
the following statements create an output data set named Pctls, which contains the
20th and 40th percentiles of the variables Strength and Width:
proc univariate data=Belts noprint;
var Strength Width;
output out=Pctls pctlpts = 20 40
pctlpre = Strength Width
pctlname = pct20 pct40;
run;

The PCTLPTS= option species the percentiles to compute (in this case, the 20th
and 40th percentiles). The PCTLPRE= and PCTLNAME= options build the names
for the variables containing the percentiles. The PCTLPRE= option gives prexes
for the new variables, and the PCTLNAME= option gives a sufx to add to the prex. When you use the PCTLPTS= specication, you must also use the PCTLPRE=
specication.
The OUTPUT statement saves the 20th and 40th percentiles of Strength and Width
in the variables Strengthpct20, Widthpct20, Strengthpct40, and Weightpct40.
The output data set Pctls is listed in Output 3.8.2.

325

326

Chapter 3. The UNIVARIATE Procedure

Output 3.8.2. Listing of Output Data Set Pctls

Obs

Strengthpct20

Strengthpct40

Widthpct20

Widthpct40

1165.91

1199.26

2.9595

2.995

A sample program, uniex06.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.9. Computing Condence Limits for the Mean,

Standard Deviation, and Variance
This example illustrates how to compute condence limits for the mean, standard
deviation, and variance of a population. A researcher is studying the heights of a
certain population of adult females. She has collected a random sample of heights of
75 females, which are saved in the data set Heights:
data Heights;
label Height = Height (in);
input Height @@;
datalines;
64.1 60.9 64.1 64.7 66.7 65.0 63.7
64.0 67.5 62.8 63.9 65.9 62.3 64.1
63.7 63.0 64.7 68.2 66.7 62.8 64.0
62.7 60.9 61.6 64.6 65.7 66.6 66.7
60.5 63.0 60.0 61.6 64.3 60.2 63.5
63.6 62.0 63.6 65.8 66.0 65.4 63.5
65.8 63.1 65.8 64.4 64.0 64.9 65.7
68.6 66.6 65.7 65.1 70.0
;
run;

67.4
60.6
64.1
66.0
64.7
66.3
61.0

64.9
68.6
62.1
68.5
66.0
66.2
64.1

63.7
68.6
62.9
64.4
65.1
67.5
65.5

The following statements produce condence limits for the mean, standard deviation,
and variance of the population of heights:
title Analysis of Female Heights;
ods select BasicIntervals;
proc univariate data=Heights cibasic;
var Height;
run;

The CIBASIC option requests condence limits for the mean, standard deviation, and
variance. For example, Output 3.9.1 shows that the 95% condence interval for the
population mean is (64.06, 65.07). The ODS SELECT statement restricts the output
to the BasicIntervals table; see the section ODS Table Names on page 309.
The condence limits in Output 3.9.1 assume that the heights are normally distributed, so you should check this assumption before using these condence limits.

Example 3.9. Computing Condence Limits for the Mean, Standard Deviation, and
Variance
See the section Shapiro-Wilk Statistic on page 293 for information on the ShapiroWilk test for normality in PROC UNIVARIATE. See Example 3.19 for an example
using the test for normality.
Output 3.9.1. Default 95% Condence Limits
Analysis of Female Heights
The UNIVARIATE Procedure
Variable: Height (Height (in))
Basic Confidence Limits Assuming Normality
Parameter

Estimate

Mean
Std Deviation
Variance

64.56667
2.18900
4.79171

95% Confidence Limits

64.06302
1.88608
3.55731

65.07031
2.60874
6.80552

By default, the condence limits produced by the CIBASIC option produce 95%
condence intervals. You can request different level condence limits by using the
ALPHA= option in parentheses after the CIBASIC option. The following statements
produce 90% condence limits:
title Analysis of Female Heights;
ods select BasicIntervals;
proc univariate data=Heights cibasic(alpha=.1);
var Height;
run;

The 90% condence limits are displayed in Output 3.9.2.

Output 3.9.2. 90% Condence Limits
Analysis of Female Heights
The UNIVARIATE Procedure
Variable: Height (Height (in))
Basic Confidence Limits Assuming Normality
Parameter

Estimate

Mean
Std Deviation
Variance

64.56667
2.18900
4.79171

90% Confidence Limits

64.14564
1.93114
3.72929

64.98770
2.53474
6.42492

For the formulas used to compute these limits, see the section Condence Limits for
Parameters of the Normal Distribution on page 277.
A sample program, uniex07.sas, for this example is available in the SAS Sample
Library for Base SAS software.

327

328

Chapter 3. The UNIVARIATE Procedure

Example 3.10. Computing Condence Limits for Quantiles

and Percentiles
This example, which is a continuation of Example 3.9, illustrates how to compute
condence limits for quantiles and percentiles. A second researcher is more interested in summarizing the heights with quantiles than the mean and standard deviation. He is also interested in computing 90% condence intervals for the quantiles.
The following statements produce estimated quantiles and condence limits for the
population quantiles:
title Analysis of Female Heights;
ods select Quantiles;
proc univariate data=Heights ciquantnormal(alpha=.1);
var Height;
run;

The ODS SELECT statement restricts the output to the Quantiles table; see the section ODS Table Names on page 309. The CIQUANTNORMAL option produces
condence limits for the quantiles. As noted in Output 3.10.1, these limits assume
that the data are normally distributed. You should check this assumption before using these condence limits. See the section Shapiro-Wilk Statistic on page 293
for information on the Shapiro-Wilk test for normality in PROC UNIVARIATE; see
Example 3.19 for an example using the test for normality.
Output 3.10.1. Normal-Based Quantile Condence Limits
Analysis of Female Heights
The UNIVARIATE Procedure
Variable: Height (Height (in))
Quantiles (Definition 5)

Quantile
100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

Estimate
70.0
70.0
68.6
67.5
66.0
64.4
63.1
61.6
60.6
60.0
60.0

90% Confidence Limits

Assuming Normality

68.94553
67.59184
66.85981
65.60757
64.14564
62.59071
61.13060
60.24022
58.55106

70.58228
68.89311
68.00273
66.54262
64.98770
63.52576
62.27352
61.54149
60.18781

It is also possible to use PROC UNIVARIATE to compute condence limits for quantiles without assuming normality. The following statements use the CIQUANTDF
option to request distribution-free condence limits for the quantiles of the population of heights:

Example 3.11. Computing Robust Estimates

329

title Analysis of Female Heights;

ods select Quantiles;
proc univariate data=Heights ciquantdf(alpha=.1);
var Height;
run;

The distribution-free condence limits are shown in Output 3.10.2.

Output 3.10.2. Distribution-Free Quantile Condence Limits
Analysis of Female Heights
The UNIVARIATE Procedure
Variable: Height (Height (in))
Quantiles (Definition 5)

Quantile
100% Max
99%
95%
90%
75% Q3
50% Median
25% Q1
10%
5%
1%
0% Min

Estimate
70.0
70.0
68.6
67.5
66.0
64.4
63.1
61.6
60.6
60.0
60.0

90% Confidence Limits

Distribution Free

68.6
67.5
66.6
65.7
64.1
62.7
60.6
60.0
60.0

70.0
70.0
68.6
66.6
65.1
63.7
62.7
61.6
60.5

-------Order Statistics------LCL Rank UCL Rank

Coverage

73
68
63
50
31
13
4
1
1

75
75
72
63
46
26
13
8
3

The table in Output 3.10.2 includes the ranks from which the condence limits are
computed. For more information on how these condence limits are calculated, see
the section Condence Limits for Percentiles on page 274. Note that condence
limits for quantiles are not produced when the WEIGHT statement is used.
A sample program, uniex07.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.11. Computing Robust Estimates

This example illustrates how you can use the UNIVARIATE procedure to compute
robust estimates of location and scale. The following statements compute these estimates for the variable Systolic in the data set BPressure, which was introduced in
Example 3.1:
title Robust Estimates for Blood Pressure Data;
ods select TrimmedMeans WinsorizedMeans RobustScale;
proc univariate data=BPressure trimmed=1 .1
winsorized=.1 robustscale;
var Systolic;
run;

The ODS SELECT statement restricts the output to the TrimmedMeans,

WinsorizedMeans, and RobustScale tables; see the section ODS Table Names

48.97
94.50
91.53
91.77
91.54
91.77
91.53
94.50
48.97

330

Chapter 3. The UNIVARIATE Procedure

on page 309. The TRIMMED= option computes two trimmed means, the rst after
removing one observation and the second after removing 10% of the observations.
If the value of TRIMMED= is greater than or equal to one, it is interpreted as the
number of observations to be trimmed. The WINSORIZED= option computes a
Winsorized mean that replaces three observations from the tails with the next closest
observations. (Three observations are replaced because np = (22)(.1) = 2.2, and
three is the smallest integer greater than 2.2.) The trimmed and Winsorized means
for Systolic are displayed in Output 3.11.1.
Output 3.11.1. Computation of Trimmed and Winsorized Means
Robust Estimates for Blood Pressure Data
The UNIVARIATE Procedure
Variable: Systolic
Trimmed Means

Percent
Trimmed
in Tail

Number
Trimmed
in Tail

Trimmed
Mean

Std Error
Trimmed
Mean

4.55
13.64

1
3

120.3500
120.3125

2.573536
2.395387

95% Confidence
Limits
114.9635
115.2069

125.7365
125.4181

t for H0:
Mu0=0.00

Pr > |t|

19
15

46.76446
50.22675

<.0001
<.0001

Winsorized Means
Percent
Winsorized
in Tail

Number
Winsorized
in Tail

Winsorized
Mean

Std Error
Winsorized
Mean

13.64

120.6364

2.417065

95% Confidence
Limits
115.4845

125.7882

t for H0:
Mu0=0.00

Pr > |t|

49.91027

<.0001

Output 3.11.1 shows the trimmed mean for Systolic is 120.35 after one observation
has been trimmed, and 120.31 after 3 observations are trimmed. The Winsorized
mean for Systolic is 120.64. For details on trimmed and Winsorized means, see the
section Robust Estimators on page 278. The trimmed means can be compared with
the means shown in Output 3.1.1 (from Example 3.1), which displays the mean for
Systolic as 121.273.
The ROBUSTSCALE option requests a table, displayed in Output 3.11.2, which includes the interquartile range, Ginis mean difference, the median absolute deviation
about the median, Qn , and Sn .
Output 3.11.2 shows the robust estimates of scale for Systolic. For instance, the
interquartile range is 13. The estimates of range from 9.54 to 13.32. See the
section Robust Estimators on page 278.
A sample program, uniex01.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.13. Testing for Location

Output 3.11.2. Computation of Robust Estimates of Scale

Robust Estimates for Blood Pressure Data
Variable:

Systolic

Robust Measures of Scale

Measure
Interquartile Range
Ginis Mean Difference
MAD
Sn
Qn

Value

Estimate
of Sigma

13.00000
15.03030
6.50000
9.54080
13.33140

9.63691
13.32026
9.63690
9.54080
11.36786

Example 3.12. Testing for Location

This example, which is a continuation of Example 3.9, illustrates how to carry out
three tests for location: the Students t test, the sign test, and the Wilcoxon signed
rank test. These tests are discussed in the section Tests for Location on page 275.
The following statements demonstrate the tests for location using the Heights data
set introduced in Example 3.9. Since the data consists of adult female heights, the
researchers are not interested in testing if the mean of the population is equal to zero
inches, which is the default 0 value. Instead, they are interested in testing if the mean
is equal to 66 inches. The following statements test the null hypothesis H0 : 0 = 66:
title Analysis of Female Height Data;
ods select TestsForLocation LocationCounts;
proc univariate data=Heights mu0=66 loccount;
var Height;
run;

The ODS SELECT statement restricts the output to the TestsForLocation and
LocationCounts tables; see the section ODS Table Names on page 309. The
MU0= option species the null hypothesis value of 0 for the tests for location; by
default, 0 = 0. The LOCCOUNT option produces the table of the number of observations greater than, not equal to, and less than 66 inches.
Output 3.12.1 contains the results of the tests for location. All three tests are highly
signicant, causing the researchers to reject the hypothesis that the mean is 66 inches.
A sample program, uniex07.sas, for this example is available in the SAS Sample
Library for Base SAS software.

331

332

Chapter 3. The UNIVARIATE Procedure

Output 3.12.1. Tests for Location with MU0=66 and LOCCOUNT

Analysis of Female Height Data
The UNIVARIATE Procedure
Variable: Height (Height (in))
Tests for Location: Mu0=66
Test

-Statistic-

-----p Value------

Students t
Sign
Signed Rank

t
M
S

Pr > |t|
Pr >= |M|
Pr >= |S|

-5.67065
-20
-849

<.0001
<.0001
<.0001

Location Counts: Mu0=66.00

Count
Num Obs > Mu0
Num Obs ^= Mu0
Num Obs < Mu0

Value
16
72
56

Example 3.13. Performing a Sign Test Using Paired Data

This example demonstrates a sign test for paired data, which is a specic application
of the Tests for Location discussed in Example 3.12.
The instructor from Example 3.4 is now interested in performing a sign test for the
pairs of test scores in his college course. The following statements request basic
statistical measures and tests for location:
title Test Scores for a College Course;
ods select BasicMeasures TestsForLocation;
proc univariate data=Score;
var ScoreChange;
run;

The ODS SELECT statement restricts the output to the BasicMeasures and
TestsForLocation tables; see the section ODS Table Names on page 309. The
instructor is not willing to assume the ScoreChange variable is normal or even
symmetric, so he decides to examine the sign test. The large p-value (0.7744) of the
sign test provides insufcient evidence of a difference in test score medians.

Example 3.14. Creating a Histogram

Output 3.13.1. Sign Test for ScoreChange

Test Scores for a College Course

Variable:

The UNIVARIATE Procedure

ScoreChange (Change in Test Scores)
Basic Statistical Measures

Location
Mean
Median
Mode

Variability

-3.08333
-3.00000
-5.00000

Std Deviation
Variance
Range
Interquartile Range

13.33797
177.90152
51.00000
10.50000

NOTE: The mode displayed is the smallest of 2 modes with a count of 2.

Tests for Location: Mu0=0

Test

-Statistic-

-----p Value------

Students t
Sign
Signed Rank

t
M
S

Pr > |t|
Pr >= |M|
Pr >= |S|

-0.80079
-1
-8.5

0.4402
0.7744
0.5278

A sample program, uniex03.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.14. Creating a Histogram

This example illustrates how to create a histogram. A semiconductor manufacturer
produces printed circuit boards that are sampled to determine the thickness of their
copper plating. The following statements create a data set named Trans, which contains the plating thicknesses (Thick) of 100 boards:
data Trans;
input Thick @@;
label Thick = Plating Thickness
datalines;
3.468 3.428 3.509 3.516 3.461 3.492
3.490 3.467 3.498 3.519 3.504 3.469
3.458 3.478 3.443 3.500 3.449 3.525
3.561 3.506 3.444 3.479 3.524 3.531
3.481 3.497 3.461 3.513 3.528 3.496
3.512 3.550 3.441 3.541 3.569 3.531
3.505 3.523 3.475 3.470 3.457 3.536
3.510 3.461 3.431 3.502 3.491 3.506
3.469 3.481 3.515 3.535 3.460 3.575
3.517 3.483 3.467 3.467 3.502 3.471
;
run;

(mils);
3.478
3.497
3.461
3.501
3.533
3.468
3.528
3.439
3.488
3.516

3.556
3.495
3.489
3.495
3.450
3.564
3.477
3.513
3.515
3.474

3.482
3.518
3.514
3.443
3.516
3.522
3.536
3.496
3.484
3.500

3.512
3.523
3.470
3.458
3.476
3.520
3.491
3.539
3.482
3.466

333

334

Chapter 3. The UNIVARIATE Procedure

The following statements create the histogram shown in Output 3.14.1.
title Analysis of Plating Thickness;
proc univariate data=Trans noprint;
histogram Thick;
run;

The NOPRINT option in the PROC UNIVARIATE statement suppresses tables of

summary statistics for the variable Thick that would be displayed by default. A histogram is created for each variable listed in the HISTOGRAM statement.
Output 3.14.1. Histogram for Plating Thickness

A sample program, uniex08.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.15. Creating a One-Way Comparative Histogram

This example illustrates how to create a comparative histogram. The effective channel
length (in microns) is measured for 1225 eld effect transistors. The channel lengths
(Length) are stored in a data set named Channel, which is partially listed in Output
3.15.1:

Example 3.15. Creating a One-Way Comparative Histogram

Length
1

1
2
2
3
3

0.91
.
1.17
1.47
.
1.39
2.04
.
1.91

The following statements request a histogram, which is shown in Output 3.15.2 of

Length ignoring the lot source:
title Histogram of Length Ignoring Lot Source;
proc univariate data=Channel noprint;
histogram Length;
run;

Output 3.15.2. Histogram for Length Ignoring Lot Source

335

336

Chapter 3. The UNIVARIATE Procedure

To investigate whether the peaks (modes) in Output 3.15.2 are related to the lot
source, you can create a comparative histogram using Lot as a classication variable.
The following statements create the histogram shown in Output 3.15.3:
title Comparative Analysis of Lot Source;
proc univariate data=Channel noprint;
class Lot;
histogram Length / nrows = 3;
run;

The CLASS statement requests comparisons for each level (distinct value) of the
classication variable Lot. The HISTOGRAM statement requests a comparative histogram for the variable Length. The NROWS= option species the number of rows
in the comparative histogram. By default, comparative histograms are displayed in
two rows per panel.
Output 3.15.3. Comparison by Lot Source

Output 3.15.3 reveals that the distributions of Length are similarly distributed except
for shifts in mean.
A sample program, uniex09.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.16. Creating a Two-Way Comparative Histogram

This example illustrates how to create a two-way comparative histogram. Two suppliers (A and B) provide disk drives for a computer manufacturer. The manufacturer
measures the disk drive opening width to determine whether there has been a change
in variability from 2002 to 2003 for each supplier.
The following statements save the measurements in a data set named Disk. There are
two classication variables, Supplier and Year, and a user-dened format is associated with Year.
proc format ;
value mytime

1 = 2002 2 = 2003;

data Disk;
input @1 Supplier $10. Year Width;
label Width = Opening Width (inches);
format Year mytime.;
datalines;
Supplier A
1
1.8932
.
.
.
Supplier B
1
1.8986
Supplier A
2
1.8978
.
.
.
Supplier B
2
1.8997
;

The following statements create the comparative histogram in Output 3.16.1:

title Results of Supplier Training Program;
proc univariate data=Disk noprint;
class Supplier Year / keylevel = (Supplier A 2003);
histogram Width / intertile = 1.0
vaxis
= 0 10 20 30
ncols
= 2
nrows
= 2
cfill
= ligr
cframetop = yellow
cframeside = yellow;
run;

The KEYLEVEL= option species the key cell as the cell for which Supplier is equal
to SUPPLIER A and Year is equal to 2003. This cell determines the binning for
the other cells, and the columns are arranged so that this cell is displayed in the upper
left corner. Without the KEYLEVEL= option, the default key cell would be the
cell for which Supplier is equal to SUPPLIER A and Year is equal to 2002; the
column labeled 2002 would be displayed to the left of the column labeled 2003.
The VAXIS= option species the tick mark labels for the vertical axis. The
NROWS=2 and NCOLS=2 options specify a 2 2 arrangement for the tiles. The
CFRAMESIDE= and CFRAMETOP= options specify ll colors for the row and column labels, and the CFILL= option species a ll color for the bars. Output 3.16.1
provides evidence that both suppliers have reduced variability from 2002 to 2003.

337

338

Chapter 3. The UNIVARIATE Procedure

Output 3.16.1. Two-Way Comparative Histogram

A sample program, uniex10.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.17. Adding Insets with Descriptive Statistics

This example illustrates how to add insets with descriptive statistics to a comparative
histogram; see Output 3.17.1. Three similar machines are used to attach a part to an
assembly. One hundred assemblies are sampled from the output of each machine,
and a part position is measured in millimeters. The following statements create the
data set Machines, which contains the measurements in a variable named Position:
data Machines;
input Position @@;
label Position = Position in Millimeters;
if
(_n_ <= 100) then Machine = Machine 1;
else if (_n_ <= 200) then Machine = Machine 2;
else
Machine = Machine 3;
datalines;
-0.17 -0.19 -0.24 -0.24 -0.12
0.07 -0.61
0.22
-0.59
0.05 -0.38
0.82 -0.14
0.32
0.12 -0.02

1.91
0.26

-0.08
0.19

0.79
1.17

0.92
1.03

...
0.48
0.79
;
run;

0.41
0.66

0.78
0.22

0.58
0.71

0.43
0.53

0.07
0.57

0.27
0.90

0.49
0.48

Example 3.17. Adding Insets with Descriptive Statistics

The following statements create the comparative histogram in Output 3.17.1:
title Machine Comparision Study;
proc univariate data=Machines noprint;
class Machine;
histogram Position / nrows
= 3
intertile = 1
midpoints = -1.2 to 2.2 by 0.1
vaxis
=
0 to 16 by 4
cfill
= ligr;
inset mean std="Std Dev" / pos = ne format = 6.3;
run;

The INSET statement requests insets containing the sample mean and standard deviation for each machine in the corresponding tile. The MIDPOINTS= option species
the midpoints of the histogram bins.
Output 3.17.1. Comparative Histograms

Output 3.17.1 shows that the average position for Machines 2 and 3 are similar and
that the spread for Machine 1 is much larger than for Machines 2 and 3.
A sample program, uniex11.sas, for this example is available in the SAS Sample
Library for Base SAS software.

339

340

Chapter 3. The UNIVARIATE Procedure

Example 3.18. Binning a Histogram

This example, which is a continuation of Example 3.14, demonstrates various methods for binning a histogram. This example also illustrates how to save bin percentages
in an OUTHISTOGRAM= data set.
The manufacturer from Example 3.14 now wants to enhance the histogram by changing the endpoints of the bins using the ENDPOINTS= option. The following statements create a histogram with bins that have end points 3.425 and 3.6 and width
0.025:
title Enhancing a Histogram;
ods select HistogramBins MyHist;
proc univariate data=Trans;
histogram Thick / midpercents name=MyHist
endpoints = 3.425 to 3.6 by .025;
run;

The ODS SELECT statement restricts the output to the HistogramBins table and
the MyHist histogram; see the section ODS Table Names on page 309. The
ENDPOINTS= option species the endpoints for the histogram bins. By default, if
the ENDPOINTS= option is not specied, the automatic binning algorithm computes
values for the midpoints of the bins. The MIDPERCENTS option requests a table
of the midpoints of each histogram bin and the percent of the observations that fall
in each bin. This table is displayed in Output 3.18.1; the histogram is displayed in
Output 3.18.2. The NAME= option species a name for the histogram that can be
used in the ODS SELECT statement.
Output 3.18.1. Table of Bin Percentages Requested with MIDPERCENTS Option
Enhancing a Histogram
The UNIVARIATE Procedure
Variable: Thick
Histogram Bins for Thick
Bin
Minimum
Point

Observed
Percent

3.425
3.450
3.475
3.500
3.525
3.550
3.575

8.000
21.000
25.000
29.000
11.000
5.000
1.000

Example 3.18. Binning a Histogram

Output 3.18.2. Histogram with ENDPOINTS= Option

The MIDPOINTS= option is an alternative to the ENDPOINTS= option for specifying histogram bins. The following statements create a similar histogram, which is
shown in Output 3.18.3, to the one in Output 3.18.2:
title Enhancing a Histogram;
proc univariate data=Trans noprint;
histogram Thick / midpoints
= 3.4375 to 3.5875 by .025
rtinclude
outhistogram = OutMdpts;
run;

Output 3.18.3 differs from Output 3.18.2 in two ways:

The MIDPOINTS= option species the bins for the histogram by specifying
the midpoints of the bins instead of specifying the endpoints. Note that the
histogram displays midpoints instead of endpoints.
The RTINCLUDE option request that the right endpoint of each bin be included in the histogram interval instead of the default, which is to include the
left endpoint in the interval. This changes the histogram slightly from Output
3.18.2. Six observations have a thickness equal to an endpoint of an interval.
For instance, there is one observation with a thickness of 3.45 mils. In Output
3.18.3, this observation is included in the bin from 3.425 to 3.45.

341

342

Chapter 3. The UNIVARIATE Procedure

Output 3.18.3. Histogram with MIDPOINTS= and RTINCLUDE Options

The OUTHISTOGRAM= option produces an output data set named OutMdpts, displayed in Output 3.18.4. This data set provides information on the bins of the histogram. For more information, see the section OUTHISTOGRAM= Output Data
Set on page 308.
Output 3.18.4. The OUTHISTOGRAM= Data Set OutMdpts
OUTHISTOGRAM= Data Set
Obs

_VAR_

_MIDPT_

1
2
3
4
5
6
7

Thick
Thick
Thick
Thick
Thick
Thick
Thick

3.4375
3.4625
3.4875
3.5125
3.5375
3.5625
3.5875

_OBSPCT_
9
21
26
28
11
5
0

A sample program, uniex08.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.19. Adding a Normal Curve to a Histogram

343

Example 3.19. Adding a Normal Curve to a Histogram

This example is a continuation of Example 3.14. The following statements t a normal distribution to the thickness measurements in the Trans data set and superimpose
the tted density curve on the histogram:
title Analysis of Plating Thickness;
ods select ParameterEstimates GoodnessOfFit FitQuantiles Bins MyPlot;
proc univariate data=Trans;
histogram Thick / normal(percents=20 40 60 80 midpercents)
name=MyPlot;
inset n normal(ksdpval) / pos = ne format = 6.3;
run;

The ODS SELECT statement restricts the output to the ParameterEstimates,

GoodnessOfFit, FitQuantiles, and Bins tables; see the section ODS Table
Names on page 309. The NORMAL option requests species that the normal
curve is to be displayed on the histogram shown in Output 3.19.3. It also requests
a summary of the tted distribution, which is shown in Output 3.19.1 and Output
3.19.2. This summary includes goodness-of-t tests, parameter estimates, and quantiles of the tted distribution. (If you specify the NORMALTEST option in the PROC
UNIVARIATE statement, the Shapiro-Wilk test for normality will be included in the
tables of statistical output.)
Two secondary options are specied in parentheses after the NORMAL primary option. The PERCENTS= option species quantiles, which are to be displayed in the
FitQuantiles table. The MIDPERCENTS option requests a table that lists the midpoints, the observed percentage of observations, and the estimated percentage of the
population in each interval (estimated from the tted normal distribution). See Table
3.3 on page 214 and Table 3.8 on page 215 for the secondary options that can be
specied with after the NORMAL primary option.
Output 3.19.1. Summary of Fitted Normal Distribution
Analysis of Plating Thickness
The UNIVARIATE Procedure
Fitted Distribution for Thick
Parameters for Normal Distribution
Parameter

Symbol

Estimate

Mean
Std Dev

Mu
Sigma

3.49533
0.032117

Goodness-of-Fit Tests for Normal Distribution

Test

---Statistic----

-----p Value-----

Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

D
W-Sq
A-Sq

Pr > D
Pr > W-Sq
Pr > A-Sq

0.05563823
0.04307548
0.27840748

>0.150
>0.250
>0.250

344

Chapter 3. The UNIVARIATE Procedure

Output 3.19.2. Summary of Fitted Normal Distribution (cont.)

Analysis of Plating Thickness
Fitted Distribution for Thick
Histogram Bin Percents for Normal Distribution
Bin
Midpoint
3.43
3.45
3.47
3.49
3.51
3.53
3.55
3.57

-------Percent-----Observed
Estimated
3.000
9.000
23.000
19.000
24.000
15.000
3.000
4.000

3.296
9.319
18.091
24.124
22.099
13.907
6.011
1.784

Quantiles for Normal Distribution

Percent
20.0
40.0
60.0
80.0

------Quantile-----Observed
Estimated
3.46700
3.48350
3.50450
3.52250

3.46830
3.48719
3.50347
3.52236

Output 3.19.3. Histogram Superimposed with Normal Curve

Example 3.20. Adding Fitted Normal Curves to a Comparative Histogram

The histogram of the variable Thick with a superimposed normal curve is shown in
Output 3.19.3.
The estimated parameters for the normal curve ( = 3.50 and = 0.03) are shown

in Output 3.19.1. By default, the parameters are estimated unless you specify values
with the MU= and SIGMA= secondary options after the NORMAL primary option.
The results of three goodness-of-t tests based on the empirical distribution function
(EDF) are displayed in Output 3.19.1. Since the p-values are all greater than 0.15,
the hypothesis of normality is not rejected.
A sample program, uniex08.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.20. Adding Fitted Normal Curves to a Comparative

Histogram
This example is a continuation of Example 3.15, which introduced the data set
Channel on page 334. In Output 3.15.3, it appears that the channel lengths in each
lot are normally distributed. The following statements use the NORMAL option to
t a normal distribution for each lot:
title Comparative Analysis of Lot Source;
proc univariate data=Channel noprint;
class Lot;
histogram Length / nrows
= 3
intertile
= 1
cprop
= orange
normal(color = black noprint);
inset n = "N" / pos = nw;
run;

The NOPRINT option in the PROC UNIVARIATE statement suppresses the tables
of statistical output produced by default; the NOPRINT option in parentheses after
the NORMAL option suppresses the tables of statistical output related to the t of the
normal distribution. The normal parameters are estimated from the data for each lot,
and the curves are superimposed on each component histogram. The INTERTILE=
option species the space between the framed areas, which are referred to as tiles. The
CPROP= option requests the shaded bars above each tile, which represent the relative
frequencies of observations in each lot. The comparative histogram is displayed in
Output 3.20.1.
A sample program, uniex09.sas, for this example is available in the SAS Sample
Library for Base SAS software.

345

346

Chapter 3. The UNIVARIATE Procedure

Output 3.20.1. Fitting Normal Curves to a Comparative Histogram

Example 3.21. Fitting a Beta Curve

You can use a beta distribution to model the distribution of a variable that is known
to vary between lower and upper bounds. In this example, a manufacturing company
uses a robotic arm to attach hinges on metal sheets. The attachment point should be
offset 10.1 mm from the left edge of the sheet. The actual offset varies between 10.0
and 10.5 mm due to variation in the arm. The following statements save the offsets
for 50 attachment points as the values of the variable Length in the data set Robots:
data Robots;
input Length @@;
label Length = Attachment Point Offset (in mm);
datalines;
10.147 10.070 10.032 10.042 10.102
10.034 10.143 10.278 10.114 10.127
10.122 10.018 10.271 10.293 10.136
10.240 10.205 10.186 10.186 10.080
10.158 10.114 10.018 10.201 10.065
10.061 10.133 10.153 10.201 10.109
10.122 10.139 10.090 10.136 10.066
10.074 10.175 10.052 10.059 10.077
10.211 10.122 10.031 10.322 10.187
10.094 10.067 10.094 10.051 10.174
;
run;

Example 3.21. Fitting a Beta Curve

The following statements create a histogram with a tted beta density curve, shown
in Output 3.21.1:
title Fitted Beta Distribution of Offsets;
ods select ParameterEstimates FitQuantiles MyHist;
proc univariate data=Robots;
histogram Length /
beta(theta=10 scale=0.5 color=red fill)
cfill
= yellow
cframe
= ligr
href
= 10
hreflabel = Lower Bound
lhref
= 2
vaxis
= axis1
name
= MyHist;
axis1 label=(a=90 r=0);
inset n = Sample Size
beta / pos=ne cfill=blank;
run;

The ODS SELECT statement restricts the output to the ParameterEstimates and
FitQuantiles tables; see the section ODS Table Names on page 309. The BETA
primary option requests a tted beta distribution. The THETA= secondary option
species the lower threshold. The SCALE= secondary option species the range
between the lower threshold and the upper threshold. Note that the default THETA=
and SCALE= values are zero and one, respectively.
Output 3.21.1. Superimposing a Histogram with a Fitted Beta Curve

347

348

Chapter 3. The UNIVARIATE Procedure

The FILL secondary option species that the area under the curve is to be lled with
the CFILL= color. (If FILL were omitted, the CFILL= color would be used to ll the
histogram bars instead.)
The HREF= option draws a reference line at the lower bound, and the HREFLABEL=
option adds the label Lower Bound. The LHREF= option species a dashed line
type for the reference line. The INSET statement adds an inset with the sample size
positioned in the northeast corner of the plot.
In addition to displaying the beta curve, the BETA option requests a summary of the
curve t. This summary, which includes parameters for the curve and the observed
and estimated quantiles, is shown in Output 3.21.2. A sample program, uniex12.sas,
for this example is available in the SAS Sample Library for Base SAS software.
Output 3.21.2. Summary of Fitted Beta Distribution
Fitted Beta Distribution of Offsets
Fitted Distribution for Length
Parameters for Beta Distribution
Parameter

Symbol

Estimate

Threshold
Scale
Shape
Shape
Mean
Std Dev

Theta
Sigma
Alpha
Beta

10
0.5
2.06832
6.022479
10.12782
0.072339

Quantiles for Beta Distribution

Percent
1.0
5.0
10.0
25.0
50.0
75.0
90.0
95.0
99.0

------Quantile-----Observed
Estimated
10.0180
10.0310
10.0380
10.0670
10.1220
10.1750
10.2255
10.2780
10.3220

Example 3.22. Fitting Lognormal,

Curves

10.0124
10.0285
10.0416
10.0718
10.1174
10.1735
10.2292
10.2630
10.3237

Weibull,

and

Gamma

To determine an appropriate model for a data distribution, you should consider

curves from several distribution families. As shown in this example, you can use
the HISTOGRAM statement to t more than one distribution and display the density
curves on a histogram.

Example 3.22. Fitting Lognormal, Weibull, and Gamma Curves

The gap between two plates is measured (in cm) for each of 50 welded assemblies
selected at random from the output of a welding process. The following statements
save the measurements (Gap) in a data set named Plates:
data Plates;
label Gap = Plate Gap in cm;
input Gap @@;
datalines;
0.746 0.357 0.376 0.327 0.485
0.252 0.512 0.534 1.656 0.742
0.541 0.805 0.682 0.418 0.506
0.519 1.302 0.275 0.601 0.388
1.547 0.690 0.676 0.314 0.736
;
run;

1.741
0.378
0.501
0.450
0.643

0.241
0.714
0.247
0.845
0.483

0.777
1.121
0.922
0.319
0.352

0.768
0.597
0.880
0.486
0.636

0.409
0.231
0.344
0.529
1.080

The following statements t three distributions (lognormal, Weibull, and gamma) and
display their density curves on a single histogram:
title Distribution of Plate Gaps;
ods select ParameterEstimates GoodnessOfFit FitQuantiles MyHist;
proc univariate data=Plates;
var Gap;
histogram / midpoints=0.2 to 1.8 by 0.2
lognormal (l=1)
weibull
(l=2)
gamma
(l=8)
vaxis
= axis1
name
= MyHist;
inset n mean(5.3) std=Std Dev(5.3) skewness(5.3)
/ pos = ne header = Summary Statistics;
axis1 label=(a=90 r=0);
run;

The ODS SELECT statement restricts the output to the ParameterEstimates,

GoodnessOfFit, and FitQuantiles tables; see the section ODS Table Names on
page 309. The LOGNORMAL, WEIBULL, and GAMMA primary options request
superimposed tted curves on the histogram in Output 3.22.1. The L= secondary options specify distinct line types for the curves. Note that a threshold parameter = 0
is assumed for each curve. In applications where the threshold is not zero, you can
specify with the THETA= secondary option.
The LOGNORMAL, WEIBULL, and GAMMA options also produce the summaries
for the tted distributions shown in Output 3.22.2 through Output 3.22.5.
Output 3.22.2 provides three EDF goodness-of-t tests for the lognormal distribution: the Anderson-Darling, the Cramr-von Mises, and the Kolmogorov-Smirnov
tests. At the = 0.10 signicance level, all tests support the conclusion that the

two-parameter lognormal distribution with scale parameter = 0.58 and shape

parameter = 0.50 provides a good model for the distribution of plate gaps.

349

350

Chapter 3. The UNIVARIATE Procedure

Output 3.22.1. Superimposing a Histogram with Fitted Curves

Output 3.22.2. Summary of Fitted Lognormal Distribution

Distribution of Plate Gaps
Fitted Distributions for Gap
Parameters for Lognormal Distribution
Parameter

Symbol

Estimate

Threshold
Scale
Shape
Mean
Std Dev

Theta
Zeta
Sigma

0
-0.58375
0.499546
0.631932
0.336436

Goodness-of-Fit Tests for Lognormal Distribution

Test

---Statistic----

-----p Value-----

Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

D
W-Sq
A-Sq

Pr > D
Pr > W-Sq
Pr > A-Sq

0.06441431
0.02823022
0.24308402

>0.150
>0.500
>0.500

Example 3.22. Fitting Lognormal, Weibull, and Gamma Curves

Output 3.22.3. Summary of Fitted Lognormal Distribution (cont.)

Distribution of Plate Gaps
Fitted Distributions for Gap
Quantiles for Lognormal Distribution
------Quantile-----Observed
Estimated

Percent
1.0
5.0
10.0
25.0
50.0
75.0
90.0
95.0
99.0

0.23100
0.24700
0.29450
0.37800
0.53150
0.74600
1.10050
1.54700
1.74100

0.17449
0.24526
0.29407
0.39825
0.55780
0.78129
1.05807
1.26862
1.78313

Output 3.22.4. Summary of Fitted Weibull Distribution

Distribution of Plate Gaps
Fitted Distributions for Gap
Parameters for Weibull Distribution
Parameter

Symbol

Estimate

Threshold
Scale
Shape
Mean
Std Dev

Theta
Sigma
C

0
0.719208
1.961159
0.637641
0.339248

Goodness-of-Fit Tests for Weibull Distribution

Test

---Statistic----

-----p Value-----

Cramer-von Mises
Anderson-Darling

W-Sq
A-Sq

Pr > W-Sq
Pr > A-Sq

0.15937281
1.15693542

Quantiles for Weibull Distribution

Percent
1.0
5.0
10.0
25.0
50.0
75.0
90.0
95.0
99.0

------Quantile-----Observed
Estimated
0.23100
0.24700
0.29450
0.37800
0.53150
0.74600
1.10050
1.54700
1.74100

0.06889
0.15817
0.22831
0.38102
0.59661
0.84955
1.10040
1.25842
1.56691

0.016
<0.010

351

352

Chapter 3. The UNIVARIATE Procedure

Output 3.22.4 provides two EDF goodness-of-t tests for the Weibull distribution: the
Anderson-Darling and the Cramr-von Mises tests. The p-values for the EDF tests
are all less than 0.10, indicating that the data do not support a Weibull model.
Output 3.22.5. Summary of Fitted Gamma Distribution
Distribution of Plate Gaps
Fitted Distributions for Gap
Parameters for Gamma Distribution
Parameter

Symbol

Estimate

Threshold
Scale
Shape
Mean
Std Dev

Theta
Sigma
Alpha

0
0.155198
4.082646
0.63362
0.313587

Goodness-of-Fit Tests for Gamma Distribution

Test

---Statistic----

-----p Value-----

Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

D
W-Sq
A-Sq

Pr > D
Pr > W-Sq
Pr > A-Sq

0.09695325
0.07398467
0.58106613

>0.250
>0.250
0.137

Quantiles for Gamma Distribution

Percent
1.0
5.0
10.0
25.0
50.0
75.0
90.0
95.0
99.0

------Quantile-----Observed
Estimated
0.23100
0.24700
0.29450
0.37800
0.53150
0.74600
1.10050
1.54700
1.74100

0.13326
0.21951
0.27938
0.40404
0.58271
0.80804
1.05392
1.22160
1.57939

Output 3.22.5 provides three EDF goodness-of-t tests for the gamma distribution:
the Anderson-Darling, the Cramr-von Mises, and the Kolmogorov-Smirnov tests.
At the = 0.10 signicance level, all tests support the conclusion that the gamma
distribution with scale parameter = 0.16 and shape parameter = 4.08 provides a
good model for the distribution of plate gaps.
Based on this analysis, the tted lognormal distribution and the tted gamma distribution are both good models for the distribution of plate gaps. A sample program,
uniex13.sas, for this example is available in the SAS Sample Library for Base SAS
software.

Example 3.23. Computing Kernel Density Estimates

This example illustrates the use of kernel density estimates to visualize a nonnormal
data distribution. This example uses the data set Channel, which is introduced in
Example 3.15.
When you compute kernel density estimates, you should try several choices for the
bandwidth parameter c since this determines the smoothness and closeness of the t.

Example 3.23. Computing Kernel Density Estimates

You can specify a list of up to ve C= values with the KERNEL option to request
multiple density estimates, as shown in the following statements:
title FET Channel Length Analysis;
proc univariate data=Channel noprint;
histogram Length / kernel(c = 0.25 0.50 0.75 1.00
l = 1 20 2 34
color=red
noprint);
run;

The L= secondary option species distinct line types for the curves (the L= values are
paired with the C= values in the order listed). Output 3.23.1 demonstrates the effect
of c. In general, larger values of c yield smoother density estimates, and smaller
values yield estimates that more closely t the data distribution.
Output 3.23.1. Multiple Kernel Density Estimates

Output 3.23.1 reveals strong trimodality in the data, which is displayed with comparative histograms in Example 3.15.
A sample program, uniex09.sas, for this example is available in the SAS Sample
Library for Base SAS software.

353

354

Chapter 3. The UNIVARIATE Procedure

Example 3.24. Fitting a Three-Parameter Lognormal Curve

If you request a lognormal t with the LOGNORMAL primary option, a twoparameter lognormal distribution is assumed. This means that the shape parameter
and the scale parameter are unknown (unless specied) and that the threshold is
known (it is either specied with the THETA= option or assumed to be zero).
If it is necessary to estimate in addition to and , the distribution is referred to as
a three-parameter lognormal distribution. This example shows how you can request
a three-parameter lognormal distribution.
A manufacturing process produces a plastic laminate whose strength must exceed a
minimum of 25 psi. Samples are tested, and a lognormal distribution is observed for
the strengths. It is important to estimate to determine whether the process meets the
strength requirement. The following statements save the strengths for 49 samples in
the data set Plastic:
data Plastic;
label Strength
input Strength
datalines;
30.26 31.23 71.96
81.37 78.48 72.65
43.27 41.76 57.24
31.29 32.48 51.54
25.80 29.95 60.89
43.41 54.67 99.43
35.57 60.41 54.92
;
run;

= Strength in psi;
@@;
47.39
61.63
23.80
44.06
55.33
50.76
35.66

33.93
34.90
34.03
42.66
39.44
48.81
59.30

76.15
24.83
33.38
47.98
34.50
31.86
41.96

42.21
68.93
21.87
33.73
73.51
33.88
45.32

The following statements use the LOGNORMAL primary option in the

HISTOGRAM statement to display the tted three-parameter lognormal curve
shown in Output 3.24.1:
title Three-Parameter Lognormal Fit;
proc univariate data=Plastic noprint;
histogram Strength / lognormal(fill theta = est noprint)
cfill = white;
inset lognormal
/ format=6.2 pos=ne;
run;

The NOPRINT option suppresses the tables of statistical output produced by default.
Specifying THETA=EST requests a local maximum likelihood estimate (LMLE) for
, as described by Cohen (1951). This estimate is then used to compute maximum
likelihood estimates for and .
Note: You can also specify THETA=EST with the WEIBULL primary option to t a
three-parameter Weibull distribution.
A sample program, uniex14.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.25. Annotating a Folded Normal Curve

Output 3.24.1. Three-Parameter Lognormal Fit

Example 3.25. Annotating a Folded Normal Curve

This example shows how to display a tted curve that is not supported by the
HISTOGRAM statement. The offset of an attachment point is measured (in mm) for
a number of manufactured assemblies, and the measurements (Offset) are saved in a
data set named Assembly. The following statements create the data set Assembly:
data Assembly;
label Offset = Offset (in
input Offset @@;
datalines;
11.11 13.07 11.42 3.92 11.08
9.18 5.07 3.51 16.65 14.10
9.97 3.28 13.03 13.78 3.13
11.98 3.90 7.67 4.32 12.69
3.18 6.02 6.63 1.72 2.42
1.29 1.70 0.65 2.62 2.04
0.31 8.91 13.62 14.94 4.83
5.16 4.14 1.92 12.70 1.97
15.84 10.85 2.35 1.93 9.19
0.05 2.15 1.95 4.39 0.48
0.23 0.38 12.71 0.06 10.11
12.93 10.39 2.05 15.49 8.12
8.60 22.22 1.74 5.84 12.90
15.60 2.36 3.97 6.17 0.62
12.91 0.95 0.89 3.82 7.86
;
run;

mm);

5.40
9.69
9.53
6.17
11.32
11.08
16.84
2.10
1.39
10.16
18.38
9.52
13.06
8.56
5.33

11.22
16.61
4.58
11.48
16.49
18.85
7.09
9.38
11.40
4.81
5.53
7.77
5.08
9.36
12.92

14.69 6.27 9.76

5.67 2.89 8.13
7.94 13.51 11.43
2.82 20.42 1.01
1.22 9.13 3.34
11.94 8.34 2.07
3.37 0.49 15.19
3.18 4.18 7.22
12.20 16.07 9.23
8.28 5.68 22.81
9.36 9.32 3.63
10.70 6.37 1.91
2.09 6.41 1.40
10.19 7.16 2.37
2.64 7.92 14.06

355

356

Chapter 3. The UNIVARIATE Procedure

It is decided to t a folded normal distribution to the offset measurements. A variable
X has a folded normal distribution if X = |Y |, where Y is distributed as N (, ).
The tted density is

h(x) =

1
(x )2
exp
2 2
2

+ exp

(x + )2
2 2

, x0

You can use SAS/IML to compute preliminary estimates of and based on a

method of moments given by Elandt (1961). These estimates are computed by solving equation (19) of Elandt (1961), which is given by

f () =

2
2 e /2
2

[1 2()]
1 + 2

where () is the standard normal distribution function, and

1
n

n
2
i=1 xi

Then the estimates of and are given by

0 =

n
i=1

1
n

x2
i

1+2

0 = 0

Begin by using PROC MEANS to compute the rst and second moments and using
the following DATA step to compute the constant A:
proc means data = Assembly noprint;
var Offset;
output out=stat mean=m1 var=var n=n min = min;
run;
* Compute constant A from equation (19) of Elandt (1961);
data stat;
keep m2 a min;
set stat;
a = (m1*m1);
m2 = ((n-1)/n)*var + a;
a = a/m2;
run;

Next, use the SAS/IML subroutine NLPDD to solve equation (19) by minimizing
(f () A)2 , and compute 0 and 0 :

Example 3.25. Annotating a Folded Normal Curve

proc iml;
use stat;
read all var {m2} into m2;
read all var {a}
into a;
read all var {min} into min;
* f(t) is the function in equation (19) of Elandt (1961);
start f(t) global(a);
y = .39894*exp(-0.5*t*t);
y = (2*y-(t*(1-2*probnorm(t))))**2/(1+t*t);
y = (y-a)**2;
return(y);
finish;
* Minimize (f(t)-A)**2 and estimate mu and sigma;
if ( min < 0 ) then do;
print "Warning: Observations are not all nonnegative.";
print "
The folded normal is inappropriate.";
stop;
end;
if ( a < 0.637 ) then do;
print "Warning: the folded normal may be inappropriate";
end;
opt = { 0 0 };
con = { 1e-6 };
x0 = { 2.0 };
tc = { . . . . . 1e-8 . . . . . . .};
call nlpdd(rc,etheta0,"f",x0,opt,con,tc);
esig0 = sqrt(m2/(1+etheta0*etheta0));
emu0 = etheta0*esig0;
create prelim var {emu0 esig0 etheta0};
append;
close prelim;

The preliminary estimates are saved in the data set Prelim, as shown in Output 3.25.1:
Output 3.25.1. Preliminary Estimates of , , and
The Data Set Prelim
EMU0

ESIG0

6.51735

ETHETA0

6.54953

0.99509

Now, using 0 and 0 as initial estimates, call the NLPDD subroutine to maximize

the log likelihood, l(, ), of the folded normal distribution, where, up to a constant,
n

log exp

l(, ) = n log +
i=1

(xi )2
2 2

+ exp

(xi + )2
2 2

357

358

Chapter 3. The UNIVARIATE Procedure

* Define the log likelihood of the folded normal;
start g(p) global(x);
y = 0.0;
do i = 1 to nrow(x);
z = exp( (-0.5/p[2])*(x[i]-p[1])*(x[i]-p[1]) );
z = z + exp( (-0.5/p[2])*(x[i]+p[1])*(x[i]+p[1]) );
y = y + log(z);
end;
y = y - nrow(x)*log( sqrt( p[2] ) );
return(y);
finish;
* Maximize the log likelihood with subroutine NLPDD;
use assembly;
read all var {offset} into x;
esig0sq = esig0*esig0;
x0
= emu0||esig0sq;
opt
= { 1 0 };
con
= { . 0.0, . . };
call nlpdd(rc,xr,"g",x0,opt,con);
emu
= xr[1];
esig
= sqrt(xr[2]);
etheta = emu/esig;
create parmest var{emu esig etheta};
append;
close parmest;
quit;

The data set ParmEst contains the maximum likelihood estimates and (as well

as / ), as shown in Output 3.25.2:

Output 3.25.2. Final Estimates of , , and
The Data Set ParmEst
EMU
6.66761

ESIG
6.39650

ETHETA
1.04239

To annotate the curve on a histogram, begin by computing the width and endpoints
of the histogram intervals. The following statements save these values in a data set
called OutCalc. Note that a plot is not produced at this point.

Example 3.25. Annotating a Folded Normal Curve

proc univariate data = Assembly noprint;
histogram Offset / outhistogram = out normal(noprint) noplot;
run;
data OutCalc (drop = _MIDPT_);
set out (keep = _MIDPT_) end = eof;
retain _MIDPT1_ _WIDTH_;
if _N_ = 1 then _MIDPT1_ = _MIDPT_;
if eof then do;
_MIDPTN_ = _MIDPT_;
_WIDTH_ = (_MIDPTN_ - _MIDPT1_) / (_N_ - 1);
output;
end;
run;

Output 3.25.3 provides a listing of the data set OutCalc. The width of the histogram
bars is saved as the value of the variable WIDTH ; the midpoints of the rst and last
histogram bars are saved as the values of the variables MIDPT1 and MIDPTN .
Output 3.25.3. The Data Set OutCalc
Data Set OutCalc
_MIDPT1_
1.5

_WIDTH_

_MIDPTN_

22.5

The following statements create an annotate data set named Anno, which contains
the coordinates of the tted curve:
data Anno;
merge ParmEst OutCalc;
length function color $ 8;
function = point;
color
= black;
size
= 2;
xsys
= 2;
ysys
= 2;
when
= a;
constant = 39.894*_width_;;
left
= _midpt1_ - .5*_width_;
right
= _midptn_ + .5*_width_;
inc
= (right-left)/100;
do x = left to right by inc;
z1 = (x-emu)/esig;
z2 = (x+emu)/esig;
y = (constant/esig)*(exp(-0.5*z1*z1)+exp(-0.5*z2*z2));
output;
function = draw;
end;
run;

The following statements read the ANNOTATE= data set and display the histogram
and tted curve:

359

360

Chapter 3. The UNIVARIATE Procedure

title Folded Normal Distribution;
proc univariate data=assembly noprint;
histogram Offset / annotate = anno
cbarline = black
cfill
= ligr;
run;

Output 3.25.4 displays the histogram and tted curve:

Output 3.25.4. Histogram with Annotated Folded Normal Curve

A sample program, uniex15.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.26. Creating Lognormal Probability Plots

This example is a continuation of the example explored in the section Modeling a
Data Distribution on page 200.
In the normal probability plot shown in Figure 3.6, the nonlinearity of the point pattern indicates a departure from normality in the distribution of Deviation. Since the
point pattern is curved with slope increasing from left to right, a theoretical distribution that is skewed to the right, such as a lognormal distribution, should provide
a better t than the normal distribution. See the section Interpretation of QuantileQuantile and Probability Plots on page 299.

Example 3.26. Creating Lognormal Probability Plots

You can explore the possibility of a lognormal t with a lognormal probability plot.
When you request such a plot, you must specify the shape parameter for the lognormal distribution. This value must be positive, and typical values of range from
0.1 to 1.0. You can specify values for with the SIGMA= secondary option in the
LOGNORMAL primary option, or you can specify that is to be estimated from the
data.
The following statements illustrate the rst approach by creating a series of three
lognormal probability plots for the variable Deviation introduced in the section
Modeling a Data Distribution on page 200:
symbol v=plus height=3.5pct;
title Lognormal Probability Plot for Position Deviations;
proc univariate data=Aircraft noprint;
probplot Deviation /
lognormal(theta=est zeta=est sigma=0.7 0.9 1.1)
href = 95
lhref = 1
square;
run;

The LOGNORMAL primary option requests plots based on the lognormal family of
distributions, and the SIGMA= secondary option requests plots for equal to 0.7,
0.9, and 1.1. These plots are displayed in Output 3.26.1, Output 3.26.2, and Output
3.26.3, respectively. Alternatively, you can specify to be estimated using the sample
standard deviation by using the option SIGMA=EST.
The SQUARE option displays the probability plot in a square format, the HREF=
option requests a reference line at the 95th percentile, and the LHREF= option
species the line type for the reference line.
Output 3.26.1. Probability Plot Based on Lognormal Distribution with =0.7

361

362

Chapter 3. The UNIVARIATE Procedure

Output 3.26.2. Probability Plot Based on Lognormal Distribution with =0.9

Output 3.26.3. Probability Plot Based on Lognormal Distribution with =1.1

The value = 0.9 in Output 3.26.2 most nearly linearizes the point pattern.
The 95th percentile of the position deviation distribution seen in Output 3.26.2 is
approximately 0.001, since this is the value corresponding to the intersection of the
point pattern with the reference line.

Example 3.27. Creating a Histogram to Display Lognormal Fit

Note: Once the that produces the most linear t is found, you can then estimate the
threshold parameter and the scale parameter . See Example 3.31.
The following statements illustrate how you can create a lognormal probability plot
for Deviation using a local maximum likelihood estimate for .
symbol v=plus height=3.5pct;
title Lognormal Probability Plot for Position Deviations;
proc univariate data=Aircraft noprint;
probplot Deviation / lognormal(theta=est zeta=est sigma=est)
href
= 95
lhref = 1
square;
run;

The plot is displayed in Output 3.26.4. Note that the maximum likelihood estimate
of (in this case 0.882) does not necessarily produce the most linear point pattern.
Output 3.26.4. Probability Plot Based on Lognormal Distribution with Estimated

A sample program, uniex16.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.27. Creating a Histogram to Display Lognormal Fit

This example uses the data set Aircraft from the previous example to illustrate how
to display a lognormal t with a histogram. To determine whether the lognormal distribution is an appropriate model for a distribution, you should consider the graphical
t as well as conduct goodness-of-t tests.

363

364

Chapter 3. The UNIVARIATE Procedure

The following statements t a lognormal distribution and display the density curve
on a histogram:
title Distribution of Position Deviations;
ods select Lognormal.ParameterEstimates Lognormal.GoodnessOfFit MyPlot;
proc univariate data=Aircraft;
var Deviation;
histogram / lognormal(w=3 theta=est)
vaxis = axis1
name = MyPlot;
inset n mean (5.3) std=Std Dev (5.3) skewness (5.3) /
pos = ne header = Summary Statistics;
axis1 label=(a=90 r=0);
run;

The ODS SELECT statement restricts the output to the ParameterEstimates and
GoodnessOfFit tables; see the section ODS Table Names on page 309. The
LOGNORMAL primary option superimposes a tted curve on the histogram in
Output 3.27.1. The W= option species the line width for the curve. The INSET
statement species that the mean, standard deviation, and skewness be displayed in
an inset in the northeast corner of the plot. Note that the default value of the threshold
parameter is zero. In applications where the threshold is not zero, you can specify
with the THETA= option. The variable Deviation includes values that are less than
the default threshold; therefore, the option THETA= EST is used.
Output 3.27.1. Normal Probability Plot Created with Graphics Device

Output 3.27.2 provides three EDF goodness-of-t tests for the lognormal distribution: the Anderson-Darling, the Cramr-von Mises, and the Kolmogorov-Smirnov
tests. The null hypothesis for the three tests is that a lognormal distribution holds for
the sample data.

Example 3.28. Creating a Normal Quantile Plot

Output 3.27.2. Summary of Fitted Lognormal Distribution

Distribution of Position Deviations
Fitted Distribution for Deviation
Parameters for Lognormal Distribution
Parameter

Symbol

Estimate

Threshold
Scale
Shape
Mean
Std Dev

Theta
Zeta
Sigma

-0.00834
-6.14382
0.882225
-0.00517
0.003438

Goodness-of-Fit Tests for Lognormal Distribution

Test

---Statistic----

-----p Value-----

Kolmogorov-Smirnov
Cramer-von Mises
Anderson-Darling

D
W-Sq
A-Sq

Pr > D
Pr > W-Sq
Pr > A-Sq

0.09419634
0.02919815
0.21606642

>0.500
>0.500
>0.500

The p-values for all three tests are greater than 0.5, so the null hypothesis is not
rejected. The tests support the conclusion that the two-parameter lognormal distri
bution with scale parameter = 6.14, and shape parameter = 0.88 provides

a good model for the distribution of position deviations. For further discussion of
goodness-of-t interpretation, see the section Goodness-of-Fit Tests on page 292.
A sample program, uniex16.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.28. Creating a Normal Quantile Plot

This example illustrates how to create a normal quantile plot. An engineer is analyzing the distribution of distances between holes cut in steel sheets. The following
statements save measurements of the distance between two holes cut into 50 steel
sheets as values of the variable Distance in the data set Sheets:
data Sheets;
input Distance
label Distance
datalines;
9.80 10.20 10.27
10.11 10.24 10.20
9.99 9.78 10.10
9.96 9.79 10.08
10.10 9.95 9.84
10.56 10.47 9.42
10.11 10.36 9.94
9.89 9.62 10.05
9.99 10.16 10.58
10.31 10.07 10.33
;
run;

@@;
= Hole Distance (cm);
9.70
10.24
10.21
9.79
10.11
10.44
9.77
9.72
10.70
9.98

9.76
9.63
10.00
10.06
9.93
10.16
9.36
9.82
9.54
10.15

365

366

Chapter 3. The UNIVARIATE Procedure

The engineer decides to check whether the distribution of distances is normal. The
following statements create a Q-Q plot for Distance, shown in Output 3.28.1:
symbol v=plus;
title Normal Quantile-Quantile Plot for Hole Distance;
proc univariate data=Sheets noprint;
qqplot Distance;
run;

The plot compares the ordered values of Distance with quantiles of the normal distribution. The linearity of the point pattern indicates that the measurements are normally
distributed. Note that a normal Q-Q plot is created by default.
Output 3.28.1. Normal Quantile-Quantile Plot for Distance

A sample program, uniex17.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.29. Adding a Distribution Reference Line

This example, which is a continuation of Example 3.28, illustrates how to add a
reference line to a normal Q-Q plot, which represents the normal distribution with
mean 0 and standard deviation 0 . The following statements reproduce the Q-Q
plot in Output 3.28.1 and add the reference line:

Example 3.29. Adding a Distribution Reference Line

symbol v=plus;
title Normal Quantile-Quantile Plot for Hole Distance;
proc univariate data=Sheets noprint;
qqplot Distance / normal(mu=est sigma=est color=red
l=2 noprint)
square;
run;

The plot is displayed in Output 3.29.1.

Specifying MU=EST and SIGMA=EST with the NORMAL primary option requests
the reference line for which 0 and 0 are estimated by the sample mean and standard
deviation. Alternatively, you can specify numeric values for 0 and 0 with the MU=
and SIGMA= secondary options. The COLOR= and L= options specify the color and
type of the line, and the SQUARE option displays the plot in a square format. The
NOPRINT options in the PROC UNIVARIATE statement and after the NORMAL
option suppress all the tables of statistical output produced by default.
Output 3.29.1. Adding a Distribution Reference Line to a Q-Q Plot

The data clearly follow the line, which indicates that the distribution of the distances
is normal.
A sample program, uniex17.sas, for this example is available in the SAS Sample
Library for Base SAS software.

367

368

Chapter 3. The UNIVARIATE Procedure

Example 3.30. Interpreting a Normal Quantile Plot

This example illustrates how to interpret a normal quantile plot when the data
are from a non-normal distribution. The following statements create the data set
Measures, which contains the measurements of the diameters of 50 steel rods in the
variable Diameter:
data Measures;
input Diameter @@;
label Diameter = Diameter (mm);
datalines;
5.501 5.251 5.404 5.366 5.445 5.576
5.200 5.977 5.177 5.332 5.399 5.661
5.252 5.404 5.739 5.525 5.160 5.410
5.376 5.202 5.470 5.410 5.394 5.146
5.309 5.480 5.388 5.399 5.360 5.368
5.248 5.409 5.304 6.239 5.781 5.247
5.208 5.143 5.304 5.603 5.164 5.209
5.223
;
run;

5.607
5.512
5.823
5.244
5.394
5.907
5.475

The following statements request the normal Q-Q plot in Output 3.30.1:
symbol v=plus;
title Normal Q-Q Plot for Diameters;
proc univariate data=Measures noprint;
qqplot Diameter / normal(noprint)
square
vaxis=axis1;
axis1 label=(a=90 r=0);
run;

The nonlinearity of the points in Output 3.30.1 indicates a departure from normality.
Since the point pattern is curved with slope increasing from left to right, a theoretical distribution that is skewed to the right, such as a lognormal distribution, should
provide a better t than the normal distribution. The mild curvature suggests that you
should examine the data with a series of lognormal Q-Q plots for small values of the
shape parameter , as illustrated in Example 3.31. For details on interpreting a Q-Q
plot, see the section Interpretation of Quantile-Quantile and Probability Plots on
page 299.

Example 3.31. Estimating Three Parameters from Lognormal Quantile Plots

Output 3.30.1. Normal Quantile-Quantile Plot of Nonnormal Data

A sample program, uniex18.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.31. Estimating Three Parameters from Lognormal

Quantile Plots
This example, which is a continuation of Example 3.30, demonstrates techniques for
estimating the shape, location, and scale parameters, and the theoretical percentiles
for a three-parameter lognormal distribution.
The three-parameter lognormal distribution depends on a threshold parameter , a
scale parameter , and a shape parameter . You can estimate from a series of
lognormal Q-Q plots which use the SIGMA= secondary option to specify different
values of ; the estimate of is the value that linearizes the point pattern. You can
then estimate the threshold and scale parameters from the intercept and slope of the
point pattern. The following statements create the series of plots in Output 3.31.1,
Output 3.31.2, and Output 3.31.3 for values of 0.2, 0.5, and 0.8, respectively:
symbol v=plus;
title Lognormal Q-Q Plot for Diameters;
proc univariate data=Measures noprint;
qqplot Diameter / lognormal(sigma=0.2 0.5 0.8 noprint)
square;
run;

369

370

Chapter 3. The UNIVARIATE Procedure

Note: You must specify a value for the shape parameter for a lognormal Q-Q plot
with the SIGMA= option or its alias, the SHAPE= option.
Output 3.31.1. Lognormal Quantile-Quantile Plot ( =0.2)

Output 3.31.2. Lognormal Quantile-Quantile Plot ( =0.5)

Example 3.31. Estimating Three Parameters from Lognormal Quantile Plots

Output 3.31.3. Lognormal Quantile-Quantile Plot ( =0.8)

The plot in Output 3.31.2 displays the most linear point pattern, indicating that the
lognormal distribution with = 0.5 provides a reasonable t for the data distribution.
Data with this particular lognormal distribution have the following density function:

p(x) =

2
(x)

exp 2(log(x ) )2

for x >
for x

The points in the plot fall on or near the line with intercept and slope exp(). Based
on Output 3.31.2, 5 and exp() 1.2 = 0.4, giving log(0.4) 0.92.
3
You can also request a reference line using the SIGMA=, THETA=, and ZETA=
options together. The following statements produce the lognormal Q-Q plot in Output
3.31.4:
symbol v=plus;
title Lognormal Q-Q Plot for Diameters;
proc univariate data=Measures noprint;
qqplot Diameter / lognormal(theta=5 zeta=est sigma=est
color=black l=2 noprint)
square;
run;

Output 3.31.1 through Output 3.31.3 show that the threshold parameter is not equal
to zero. Specifying THETA=5 overrides the default value of zero. The SIGMA=EST
and ZETA=EST secondary options request estimates for and exp using the sample
mean and standard deviation.

371

372

Chapter 3. The UNIVARIATE Procedure

Output 3.31.4. Lognormal Quantile-Quantile Plot ( =est, =est, =5)

From the plot in Output 3.31.2, can be estimated as 0.51, which is consistent with
the estimate of 0.5 derived from the plot in Output 3.31.2. The next example illustrates how to estimate percentiles using lognormal Q-Q plots.
A sample program, uniex18.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.32. Estimating Percentiles

Quantile Plots

from

Lognormal

This example, which is a continuation of the previous example, shows how to use a QQ plot to estimate percentiles such as the 95th percentile of the lognormal distribution.
A probability plot can also be used for this purpose, as illustrated in Example 3.26.
The point pattern in Output 3.31.4 has a slope of approximately 0.39 and an intercept
of 5. The following statements reproduce this plot, adding a lognormal reference line
with this slope and intercept:
symbol v=plus;
title Lognormal Q-Q Plot for Diameters;
proc univariate data=Measures noprint;
qqplot Diameter / lognormal(sigma=0.5 theta=5 slope=0.39 noprint)
pctlaxis(grid)
vref = 5.8 5.9 6.0
square;
run;

The result is shown in Output 3.32.1:

Example 3.33. Estimating Parameters from Lognormal Quantile Plots

Output 3.32.1. Lognormal Q-Q Plot Identifying Percentiles

The PCTLAXIS option labels the major percentiles, and the GRID option draws
percentile axis reference lines. The 95th percentile is 5.9, since the intersection of
the distribution reference line and the 95th reference line occurs at this value on the
vertical axis.
Alternatively, you can compute this percentile from the estimated lognormal parameters. The th percentile of the lognormal distribution is
P = exp(1 () + ) +
where 1 () is the inverse cumulative standard normal distribution. Consequently,

P0.95 = exp

1 1
2 (0.95)

+ log(0.39) + 5 = 5.89

A sample program, uniex18.sas, for this example is available in the SAS Sample
Library for Base SAS software.

Example 3.33. Estimating Parameters

Quantile Plots

from

Lognormal

This example, which is a continuation of Example 3.31, demonstrates techniques for

estimating the shape, location, and scale parameters, and the theoretical percentiles
for a two-parameter lognormal distribution.
If the threshold parameter is known, you can construct a two-parameter lognormal
Q-Q plot by subtracting the threshold from the data values and making a normal Q-Q
plot of the log-transformed differences, as illustrated in the following statements:

373

374

Chapter 3. The UNIVARIATE Procedure

data ModifiedMeasures;
set Measures;
LogDiameter = log(Diameter-5);
label LogDiameter = log(Diameter-5);
run;
symbol v=plus;
title Two-Parameter Lognormal Q-Q Plot for Diameters;
proc univariate data=ModifiedMeasures noprint;
qqplot LogDiameter / normal(mu=est sigma=est noprint)
square
vaxis=axis1;
inset n mean (5.3) std (5.3)
/ pos = nw header = Summary Statistics;
axis1 label=(a=90 r=0);
run;

Output 3.33.1. Two-Parameter Lognormal Q-Q Plot for Diameters

Because the point pattern in Output 3.33.1 is linear, you can estimate the lognormal
parameters and as the normal plot estimates of and , which are 0.99 and
0.51. These values correspond to the previous estimates of 0.92 for and 0.5 for
from Example 3.31. A sample program, uniex18.sas, for this example is available
in the SAS Sample Library for Base SAS software.

Example 3.34. Comparing Weibull Quantile Plots

This example compares the use of three-parameter and two-parameter Weibull Q-Q
plots for the failure times in months for 48 integrated circuits. The times are assumed
to follow a Weibull distribution. The following statements save the failure times as
the values of the variable Time in the data set Failures:
data Failures;
input Time @@;
label Time = Time in Months;
datalines;
29.42 32.14 30.58 27.50 26.08 29.06
29.14 33.96 30.64 27.32 29.86 26.28
29.32 30.82 27.26 27.92 30.92 24.64
30.28 28.36 25.86 31.36 25.26 36.32
26.72 27.42 29.02 27.54 31.60 33.46
29.18 27.94 27.66 26.42 31.00 26.64
;
run;

25.10
29.68
32.90
28.58
26.78
31.44

31.34
33.76
35.46
28.88
27.82
32.52

If no assumption is made about the parameters of this distribution, you can use the
WEIBULL option to request a three-parameter Weibull plot. As in the previous example, you can visually estimate the shape parameter c by requesting plots for different
values of c and choosing the value of c that linearizes the point pattern. Alternatively,
you can request a maximum likelihood estimate for c, as illustrated in the following
statements:
symbol v=plus;
title Three-Parameter Weibull Q-Q Plot for Failure Times;
proc univariate data=Failures noprint;
qqplot Time / weibull(c=est theta=est sigma=est noprint)
square
href=0.5 1 1.5 2
vref=25 27.5 30 32.5 35
lhref=4 lvref=4
chref=tan cvref=tan;
run;

Note: When using the WEIBULL option, you must either specify a list of values for
the Weibull shape parameter c with the C= option, or you must specify C=EST.
Output 3.34.1 displays the plot for the estimated value c = 1.99. The reference

line corresponds to the estimated values for the threshold and scale parameters of

0 = 24.19 and 0 = 5.83, respectively.

375

376

Chapter 3. The UNIVARIATE Procedure

Output 3.34.1. Three-Parameter Weibull Q-Q Plot

Now, suppose it is known that the circuit lifetime is at least 24 months. The following
statements use the known threshold value 0 = 24 to produce the two-parameter
Weibull Q-Q plot shown in Output 3.31.4:
symbol v=plus;
title Two-Parameter Weibull Q-Q Plot for Failure Times;
proc univariate data=Failures noprint;
qqplot Time / weibull(theta=24 c=est sigma=est noprint)
square
vref= 25 to 35 by 2.5
href= 0.5 to 2.0 by 0.5
lhref=4 lvref=4
chref=tan cvref=tan;
run;

The reference line is based on maximum likelihood estimates c = 2.08 and = 6.05.

References

Output 3.34.2. Two-Parameter Weibull Q-Q Plot for 0 = 24

A sample program, uniex19.sas, for this example is available in the SAS Sample
Library for Base SAS software.

References
Blom, G. (1958), Statistical Estimates and Transformed Beta-Variables, New York:
John Wiley & Sons, Inc.
Chambers, J. M., Cleveland, W. S., Kleiner, B., and Tukey, P. A. (1983), Graphical
Methods for Data Analysis, Belmont, Calif.: Wadsworth International Group.
Cohen, A. C., Jr.
(1951), Estimating Parameters of Logarithmic-Normal
Distributions by Maximum Likelihood, Journal of the American Statistical
Association, 46, 206212.
Conover, W. J. (1999), Practical Nonparametric Statistics, Third Edition, New York:
John Wiley & Sons, Inc.
Croux, C. and Rousseeuw, P. J. (1992), Time-Efcient Algorithms for Two Highly
Robust Estimators of Scale, Computational Statistics, Vol. 1, 411428.
DAgostino, R. B. and Stephens, M. (1986), Goodness-of-Fit Techniques, New York:
Marcel Dekker, Inc.
Dixon, W. J. and Tukey, J. W. (1968), Approximate Behavior of the Distribution of
Winsorized t (Trimming/Winsorization 2), Technometrics, 10, 8398.

377

378

Chapter 3. The UNIVARIATE Procedure

Elandt, R. C. (1961), The Folded Normal Distribution: Two Methods of Estimating
Parameters from Moments, Technometrics, 3, 551562.
Fisher, R. A. (1973), Statistical Methods and Scientic Inference, New York: Hafner
Press.
Fowlkes, E. B. (1987), A Folio of Distributions: A Collection of Theoretical QuantileQuantile Plots, New York: Marcel Dekker, Inc.
Hahn, G. J. and Meeker, W. Q. (1991), Statistical Intervals:
Practitioners, New York: John Wiley & Sons, Inc.

A Guide for

Hampel, F. R. (1974), The Inuence Curve and Its Role in Robust Estimation,
Journal of the American Statistical Association, 69, 383393.
Iman, R. L. (1974), Use of a t-statistic as an Approximation to the Exact Distribution
of the Wilcoxon Signed Ranks Test Statistic, Communications in Statistics, 3,
795806.
Johnson, N. L., Kotz, S., and Balakrishnan, N. (1994), Continuous Univariate
Distributions, Vol. 1, New York: John Wiley & Sons, Inc.
Johnson, N. L., Kotz, S., and Balakrishnan, N. (1995), Continuous Univariate
Distributions, Vol. 2, New York: John Wiley & Sons, Inc.
Lehmann, E. L. (1998), Nonparametrics: Statistical Methods Based on Ranks, New
Jersey: Prentice Hall.
Odeh, R. E. and Owen, D. B. (1980), Tables for Normal Tolerance Limits, Sampling
Plans, and Screening, New York: Marcel Dekker, Inc.
Owen, D. B. and Hua, T. A. (1977), Tables of Condence Limits on the Tail Area of
the Normal Distribution, Communication and Statistics, Part BSimulation and
Computation, 6, 285311.
Rousseeuw, P. J. and Croux, C. (1993), Alternatives to the Median Absolute
Deviation, Journal of the American Statistical Association. 88, 12731283.
Royston, J. P. (1992), Approximating the Shapiro-Wilk W-Test for Non-normality,
Statistics and Computing, 2, 117119.
Shapiro, S. S. and Wilk, M. B. (1965), An Analysis of Variance Test for Normality
(complete samples), Biometrika, 52, 591611.
Silverman, B. W. (1986), Density Estimation for Statistics and Data Analysis, New
York: Chapman and Hall.
Terrell, G. R. and Scott, D. W. (1985), Oversmoothed Nonparametric Density
Estimates, Journal of the American Statistical Association, 80, 209214.
Tukey, J. W. (1977), Exploratory Data Analysis, Reading, Massachusetts: AddisonWesley.
Tukey, J. W. and McLaughlin, D. H. (1963), Less Vulnerable Condence
and Signicance Procedures for Location Based on a Single Sample:
Trimming/Winsorization 1, Sankhya A, 25, 331352.

Subject Index
A
adjusted odds ratio, 137
agreement, measures of, 127
alpha level
FREQ procedure, 79, 87
Anderson-Darling statistic, 295
Anderson-Darling test, 206
annotating
histograms, 218
probability plots, 245
quantile plots, 258
ANOVA (row mean scores) statistic, 136
association, measures of
FREQ procedure, 108
asymmetric lambda, 108, 116

B
beta distribution, 288, 301
deviation from theoretical distribution, 294
EDF goodness-of-t test, 294
estimation of parameters, 218
tting, 218, 288
formulas for, 288
probability plots, 245, 301
quantile plots, 258, 301
binomial proportion test, 118
examples, 166
Bowkers test of symmetry, 127, 128
box plots, line printer, 207, 282
side-by-side, 206, 283
Breslow-Day test, 142

C
case-control studies
odds ratio, 122, 137, 138
cell count data, 98
example (FREQ), 161
chi-square tests
examples (FREQ), 164, 169, 172
FREQ procedure, 103, 104
Cicchetti-Allison weights, 131
Cochrans Q test, 127, 133, 182
Cochran-Armitage test for trend, 124, 177
Cochran-Mantel-Haenszel statistics (FREQ), 81, 134,
See also chi-square tests
ANOVA (row mean scores) statistic, 136
correlation statistic, 135
examples, 174

general association statistic, 136

cohort studies, 174
relative risk, 123, 141
comparative plots, 210212, 284
histograms, 224, 226, 249, 334, 337, 345
probability plots, 250
quantile plots, 262, 263
concordant observations, 108
condence limits
asymptotic (FREQ), 109
exact (FREQ), 77
for percentiles, 274
means, for, 277
parameters of normal distribution, for, 277
standard deviations, for, 277
variances, for, 278
condence limits for the correlation
Fishers z transformation, 22
contingency coefcient, 103, 108
contingency tables, 65, 84
continuity-adjusted chi-square, 103, 105
CORR procedure
computer resources, 28
concepts, 14
details, 14
examples, 34
missing values, 26
ODS graph names, 34
ODS table names, 30
output, 26
output data sets, 27
overview, 3
results, 26
syntax, 6
task tables, 7
corrected sums of squares and crossproducts, 8
correlation coefcients, 3
limited combinations of, 14
printing, for each variable, 8
suppressing probabilities, 8
correlation statistic, 135
covariances, 8
Cramers V statistic, 103, 108
Cramr-von Mises statistic, 295
Cramr-von Mises test, 206
Cronbachs coefcient alpha, 24
calculating and printing, 8
example, 48

380

Subject Index

for estimating reliability, 3

crosstabulation tables, 65, 84

D
data summarization tools, 196
density estimation,
See kernel density estimation
descriptive statistics
computing, 269
discordant observations, 108
distribution of variables, 196

E
EDF,
See empirical distribution function
EDF goodness-of-t tests, 294
probability values of, 296
empirical distribution function
denition of, 294
EDF test statistics, 294
exact tests
computational algorithms (FREQ), 143
computational resources (FREQ), 145
condence limits, 77
FREQ procedure, 142, 177
network algorithm (FREQ), 143
p-value, denitions, 144
exponential distribution, 289, 301
deviation from theoretical distribution, 294
EDF goodness-of-t test, 294
estimation of parameters, 221
tting, 289
formulas for, 289
probability plots, 247, 301
quantile plots, 260, 301
extreme observations, 230, 315
extreme values, 315

F
Fishers exact test
FREQ procedure, 103, 106, 107
Fishers z transformation, 8, 21
applications, 23
condence limits for the correlation, 22
tted parametric distributions, 288
beta distribution, 288
exponential distribution, 289
folded normal distribution, 355
gamma distribution, 290
lognormal distribution, 291
normal distribution, 291
Weibull distribution, 292
Fleiss-Cohen weights, 131
folded normal distribution, 355
Freeman-Halton test, 107
FREQ procedure
alpha level, 79, 87
binomial proportion, 118, 166
Bowkers test of symmetry, 127

Breslow-Day test, 142

cell count data, 98, 161
chi-square tests, 103105, 164, 169
Cochrans Q test, 127
Cochran-Mantel-Haenszel statistics, 174
computation time, limiting, 79
computational methods, 143
computational resources, 145, 147
contingency coefcient, 103
contingency table, 169
continuity-adjusted chi-square, 103, 105
correlation statistic, 135
Cramers V statistic, 103
default tables, 84
displayed output, 151
exact p-values, 144
EXACT statement, used with TABLES, 80
exact tests, 77, 142, 177
Fishers exact test, 103
Friedmans chi-square statistic, 180
gamma statistic, 108
general association statistic, 136
grouping variables, 99
input data sets, 75, 98
kappa coefcient, 131, 132
Kendalls tau-b statistic, 108
lambda asymmetric, 108
lambda symmetric, 108
likelihood-ratio chi-square test, 103
Mantel-Haenszel chi-square test, 103
McNemars test, 127
measures of association, 108
missing values, 100
Monte Carlo estimation, 77, 79, 146
multiway tables, 152, 154, 155
network algorithm, 143
odds ratio, 122, 137, 138
ODS table names, 158
one-way frequency tables, 103, 104, 151, 152,
164
order of variables, 76
output data sets, 80, 148150, 161, 172
output variable prexes, 150
OUTPUT, used with TABLES or EXACT, 83
overall kappa coefcient, 127
Pearson chi-square test, 103, 105
Pearson correlation coefcient, 108
phi coefcient, 103
polychoric correlation coefcient, 108
relative risk, 123, 137, 141
row mean scores statistic, 136
scores, 102
simple kappa coefcient, 127
Somers D statistics, 108
Spearman rank correlation coefcient, 108
statistical computations, 102
stratied table, 174
Stuarts tau-c statistic, 108
two-way frequency tables, 104, 105, 169

Subject Index
uncertainty coefcients, 108
weighted kappa coefcient, 127
frequency tables, 65, 84
creating (UNIVARIATE), 317
one-way (FREQ), 103, 104, 151, 152, 164
two-way (FREQ), 104, 105, 169
Friedmans chi-square statistic, 180

381

gamma distribution, 290, 301

deviation from theoretical distribution, 294
EDF goodness-of-t test, 294
estimation of parameters, 222
tting, 222, 290
formulas for, 290
probability plots, 247, 301
quantile plots, 260, 301
gamma statistic, 108, 110
general association statistic, 136
Ginis mean difference, 280
goodness-of-t tests, 206, 292, 348,
See empirical distribution function
Anderson-Darling, 295
Cramr-von Mises, 295
EDF, 294, 296
Kolmogorov D, 294
Shapiro-Wilk, 293
graphics, 196
annotating, 204
descriptions, 220, 247, 260
high-resolution, 196, 284
insets, 230, 285, 286
line printer, 281
naming, 226
probability plots, 241
quantile plots, 253
saving, 205

intervals, 227, 310

kernel density estimation, options, 218, 223,
224, 228, 229
kernel density estimation, superimposed, 297,
352
line type, 224
lognormal curve, superimposed, 224
lognormal distribution, 363
midpoints, 225
multiple distributions, example, 348
normal curve, superimposed, 226
normal distribution, 226
options summarized by function, 214
output data sets, 308
parameters for tted density curves, 217, 218,
226229
plots, suppressing, 226
quantiles, 310
reference lines, options, 219, 220, 222224, 228,
229
saving histogram intervals, 227
tables of statistical output, 310
tables of statistical output, suppressing, 226
three-parameter lognormal distribution, superimposed, 354
three-parameter Weibull distribution, superimposed, 354
tick marks on horizontal axis, 222
tiles for comparative plots, 223
Weibull curve, superimposed, 229
Hoeffdings measure of dependence, 3, 18
calculating and printing, 8
example, 34
output data set with, 8
probability values, 18
hypothesis tests
exact (FREQ), 77

high-resolution graphics, 196, 284

histograms, 212, 310
adding a grid, 222
annotating, 218
appearance, 218224, 226229
axis color, 219
axis scaling, 229
bar width, 218, 225
bars, suppressing, 226
beta curve, superimposed, 218
binning, 340
color, options, 219, 220
comparative, 224, 226, 334, 337, 345
creating, 333
endpoints of intervals, 227
exponential curve, superimposed, 221
extreme observations, 315
lling area under density curve, 221
gamma curve, superimposed, 222
insets, 338

insets, 230, 338

appearance, 236
appearance, color, 235
positioning, 235, 236, 285
positioning in margins, 286
positioning with compass point, 285
positioning with coordinates, 286
statistics associated with distributions, 232, 233
insets for descriptive statistics,
See insets
interquartile range, 280

J
Jonckheere-Terpstra test, 125

K
kappa coefcient, 129, 130
tests, 132
weights, 131
Kendall correlation statistics, 8

382

Subject Index

Kendalls partial tau-b, 3, 13

Kendalls tau-b, 3, 16
probability values, 17
Kendalls tau-b statistic, 108, 111
kernel density estimation, 297, 352
adding density curve to histogram, 223
bandwidth parameter, specifying, 218
color, 219
density curve, width of, 229
kernel function, specifying type of, 223
line type for density curve, 224
lower bound, specifying, 224
upper bound, specifying, 228
kernel function,
See kernel density estimation
key cell for comparative plots, 211
Kolmogorov D statistic, 294
Kolmogorov-Smirnov test, 206

L
lambda asymmetric, 108, 116
lambda symmetric, 108, 117
likelihood-ratio chi-square test, 103
likelihood-ratio test
chi-square (FREQ), 105
line printer plots, 281
box plots, 282, 283
normal probability plots, 282
stem-and-leaf plots, 282
listwise deletion, 26
location estimates
robust, 207, 209
location parameters, 304
probability plots, estimation with, 304
quantile plots, estimation with, 304
location, tests for
UNIVARIATE procedure, 331
lognormal distribution, 291, 302
deviation from theoretical distribution, 294
EDF goodness-of-t test, 294
estimation of parameters, 224
tting, 224, 291
formulas for, 291
histograms, 354, 363
probability plots, 249, 302, 360
quantile plots, 261, 302, 373

M
Mantel-Haenszel chi-square test, 103, 106
McNemars test, 127, 128
measures of agreement, 127
measures of association, 34
nonparametric, 3
measures of location
means, 270
modes, 272, 314
trimmed means, 279
Winsorized means, 278
median absolute deviation (MAD), 280

Mehta and Patel, network algorithm, 143

missing values
UNIVARIATE procedure, 268
mode calculation, 272
modied ridit scores, 103
Monte Carlo estimation
FREQ procedure, 77, 79, 146

N
network algorithm, 143
nonparametric density estimation,
See kernel density estimation
nonparametric measures of association, 3
normal distribution, 291, 302
deviation from theoretical distribution, 294
EDF goodness-of-t test, 294
estimation of parameters, 226
tting, 226, 291
formulas for, 291
histograms, 226
probability plots, 241, 250, 302
quantile plots, 262, 302, 365
normal probability plots,
See probability plots
line printer, 207, 282

O
odds ratio
adjusted, 137
Breslow-Day test, 142
case-control studies, 122, 137, 138
logit estimate, 138
Mantel-Haenszel estimate, 137
ODS (Output Delivery System)
CORR procedure and, 30
UNIVARIATE procedure table names, 309
ODS graph names
CORR procedure, 34
output data sets
saving correlations in, 52
overall kappa coefcient, 127, 132

P
paired data, 275, 332
pairwise deletion, 26
parameters for tted density curves, 217, 218, 226
229
partial correlations, 18
probability values, 21
Pearson chi-square test, 103, 105
Pearson correlation coefcient, 108, 113
Pearson correlation statistics, 3
example, 34
in output data set, 8
Pearson partial correlation, 3, 13
Pearson product-moment correlation, 3, 8, 14, 34
Pearson weighted product-moment correlation,
3, 13
probability values, 16

Subject Index
suppressing, 8
percentiles
axes, quantile plots, 263, 305
calculating, 273
condence limits for, 274, 328
dening, 207, 273
empirical distribution function, 273
options, 239, 240
probability plots and, 241
quantile plots and, 253
saving to an output data set, 325
visual estimates, probability plots, 305
visual estimates, quantile plots, 305
weighted, 273
weighted average, 273
phi coefcient, 103, 108
plot statements, UNIVARIATE procedure, 195
plots
box plots, 206, 207, 282, 283
comparative, 210212, 284
comparative histograms, 224, 226, 249, 334,
337, 345
comparative probability plots, 250
comparative quantile plots, 262, 263
line printer plots, 281
normal probability plots, 207, 282
probability plots, 241, 300
quantile plots, 253, 300
size of, 207
stem-and-leaf, 207, 282
suppressing, 226
polychoric correlation coefcient, 94, 108, 116
probability plots, 241
annotating, 245
appearance, 246250, 252
axis color, 246
beta distribution, 245, 301
comparative, 249, 250
distribution reference lines, 251
distributions for, 300
exponential distribution, 247, 301
frame, color, 246
gamma distribution, 247, 301
location parameters, estimation of, 304
lognormal distribution, 249, 302, 360, 363
normal distribution, 241, 250, 302
options summarized by function, 243
overview, 241
parameters for distributions, 243, 245, 246, 249
253
percentile axis, 250
percentiles, estimates of, 305
reference lines, 247, 248
reference lines, options, 246, 248250, 252
scale parameters, estimation of, 304
shape parameters, 251, 303
three-parameter Weibull distribution, 302
threshold parameter, 252
threshold parameters, estimation of, 304

383

tick marks on horizontal axis, 248

tiles for comparative plots, 248
two-parameter Weibull distribution, 303
Weibull distribution, 253
prospective study, 174

Q
Q-Q plots,
See quantile plots
Qn , 280
quantile plots, 253
appearance, 259263, 265
axes, percentile scale, 263, 305
axis color, 259
beta distribution, 258, 301
comparative, 262, 263
creating, 298
diagnostics, 299
distribution reference lines, 264, 366
distributions for, 300
exponential distribution, 260, 301
frame, color, 259
gamma distribution, 260, 301
interpreting, 299
legends, suppressing (UNIVARIATE), 366
location parameters, estimation of, 304
lognormal distribution, 261, 302, 369, 373
lognormal distribution, percentiles, 372
nonnormal data, 368
normal distribution, 262, 302, 365
options summarized by function, 255
overview, 253
parameters for distributions, 255, 258, 259, 262
266
percentiles, estimates of, 305
reference lines, 259, 260, 263
reference lines, options, 259, 261263, 265
scale parameters, estimation of, 304
shape parameters, 264, 303
three-parameter Weibull distribution, 302
threshold parameter, 265
threshold parameters, estimation of, 304
tick marks on horizontal axis, 260
tiles for comparative plots, 261
two-parameter Weibull distribution, 303
Weibull distribution, 266, 375
quantile-quantile plots,
See quantile plots
quantiles
dening, 273
empirical distribution function, 273
histograms and, 310
weighted average, 273

R
rank scores, 103
relative risk, 137
cohort studies, 123
logit estimate, 141

384

Subject Index

Mantel-Haenszel estimate, 141

reliability estimation, 3
ridit scores, 103
risks and risk differences, 120
robust estimates, 207, 209
robust estimators, 278, 329
Ginis mean difference, 280
interquartile range, 280
median absolute deviation (MAD), 280
Qn , 280
Sn , 280
trimmed means, 279
Winsorized means, 278
robust measures of scale, 280
rounding, 207, 269
UNIVARIATE procedure, 269
row mean scores statistic, 136

S
saving correlations
example, 52
scale estimates
robust, 207
scale parameters, 304
probability plots, 304
quantile plots, 304
shape parameters, 303
probability plots, 251
quantile plots, 264
Shapiro-Wilk statistic, 293
Shapiro-Wilk test, 206
sign test, 275, 276
paired data and, 332
signed rank statistic, computing, 277
simple kappa coefcient, 127, 129
singularity of variables, 8
smoothing data distribution,
See kernel density estimation
Sn , 280
Somers D statistics, 108, 112
Spearman correlation statistics, 8
probability values, 16
Spearman partial correlation, 3, 13
Spearman rank-order correlation, 3, 16, 34
Spearman rank correlation coefcient, 108, 114
standard deviation, 8
stem-and-leaf plots, 207, 282
stratied analysis
FREQ procedure, 65, 84
stratied table
example, 174
Stuarts tau-c statistic, 108, 112
Students t test, 275, 276
summary statistics
insets of, 230
sums of squares and crossproducts, 8

T
t test

Students, 275, 276

table scores, 103
tables
frequency and crosstabulation (FREQ), 65, 84
multiway, 152, 154, 155
one-way frequency, 151, 152
one-way, tests, 103, 104
two-way, tests, 104, 105
Tarones adjustment
Breslow-Day test, 142
tests for location, 275, 331
paired data, 275, 332
sign test, 276
Students t test, 276
Wilcoxon signed rank test, 277
tetrachoric correlation coefcient, 94, 116
theoretical distributions, 300
three-parameter Weibull distribution, 302
probability plots, 302
quantile plots, 302
threshold parameter
probability plots, 252
quantile plots, 265
threshold parameters
probability plots, 304
quantile plots, 304
tiles for comparative plots, 248, 261
histograms, 223
probability plots, 248
quantile plots, 261
trend test, 124, 177
trimmed means, 207, 279
two-parameter Weibull distribution, 303
probability plots, 303
quantile plots, 303

U
uncertainty coefcients, 108, 117, 118
univariate analysis
for multiple variables, 312
UNIVARIATE procedure
calculating modes, 314
classication levels, 210
comparative plots, 210212, 284
computational resources, 311
concepts, 268
condence limits, 204, 277, 326
descriptive statistics, 269, 312
examples, 312
extreme observations, 230, 315
extreme values, 315
tted continuous distributions, 288
frequency variables, 322
goodness-of-t tests, 292
high-resolution graphics, 284
histograms, 310, 315
insets for descriptive statistics, 230
keywords for insets, 230
keywords for output data sets, 237

Subject Index
line printer plots, 281, 319
missing values, 210, 268
mode calculation, 272
normal probability plots, 282
ODS table names, 309
output data sets, 237, 306, 323
overview, 196
percentiles, 241, 253, 273
percentiles, condence limits, 204, 205, 328
plot statements, 195
probability plots, 241, 300
quantile plots, 253, 300
quantiles, condence limits, 204, 205
results, 308
robust estimates, 329
robust estimators, 278
robust location estimates, 207, 209
robust scale estimates, 207
rounding, 269
sign test, 276, 332
specifying analysis variables, 266
task tables, 253
testing for location, 331
tests for location, 275
weight variables, 267
UNIVARIATE procedure, OUTPUT statement
output data set, 306

V
variances, 8

W
Weibull distribution, 292
deviation from theoretical distribution, 294
EDF goodness-of-t test, 294
estimation of parameters, 229
tting, 229, 292
formulas for, 292
histograms, 354
probability plots, 253
quantile plots, 266, 375
three-parameter, 302
two-parameter, 303
weight values, 205
weighted kappa coefcient, 127, 130
weighted percentiles, 273
Wilcoxon signed rank test, 275, 277
Winsorized means, 209, 278

Y
Yules Q statistic, 110

Z
zeros, structural and random
FREQ procedure, 133

385

386

Subject Index

Syntax Index
A
AGREE option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 81
TABLES statement (FREQ), 87
TEST statement (FREQ), 97
AJCHI option
OUTPUT statement (FREQ), 81
ALL option
OUTPUT statement (FREQ), 81
PROC UNIVARIATE statement, 204
TABLES statement (FREQ), 87
ALPHA option
PROC CORR statement, 8
ALPHA= option
EXACT statement (FREQ), 79
HISTOGRAM statement (UNIVARIATE), 217,
289
PLOTS=SCATTER option (CORR), 32
PROBPLOT statement (UNIVARIATE), 245
PROC UNIVARIATE statement, 204
QQPLOT statement (UNIVARIATE), 258
TABLES statement (FREQ), 87
ANNOKEY option
HISTOGRAM statement (UNIVARIATE), 217
PROBPLOT statement (UNIVARIATE), 245
QQPLOT statement (UNIVARIATE), 258
ANNOTATE= option
HISTOGRAM statement (UNIVARIATE), 218,
355
PROBPLOT statement (UNIVARIATE), 245
PROC UNIVARIATE statement, 204, 305
QQPLOT statement (UNIVARIATE), 258

B
BARWIDTH= option
HISTOGRAM statement (UNIVARIATE), 218
BDCHI option
OUTPUT statement (FREQ), 81
BDT option
TABLES statement (FREQ), 87
BEST= option
PROC CORR statement, 8
BETA option
HISTOGRAM statement (UNIVARIATE), 218,
288, 346
PROBPLOT statement (UNIVARIATE), 245,
301
QQPLOT statement (UNIVARIATE), 258, 301

BETA= option
HISTOGRAM statement (UNIVARIATE), 218,
289
PROBPLOT statement (UNIVARIATE), 245
QQPLOT statement (UNIVARIATE), 258
BINOMIAL option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 81
TABLES statement (FREQ), 87, 166
BINOMIALC option
TABLES statement (FREQ), 88
BY statement
CORR procedure, 12
FREQ procedure, 77
UNIVARIATE procedure, 209

C
C= option
HISTOGRAM statement (UNIVARIATE), 218,
297, 352
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259, 375
CAXIS= option
HISTOGRAM statement (UNIVARIATE), 219
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259
CBARLINE= option
HISTOGRAM statement (UNIVARIATE), 219
CELLCHI2 option
TABLES statement (FREQ), 88
CFILL= option
HISTOGRAM statement (UNIVARIATE), 219
INSET statement (UNIVARIATE), 235
CFILLH= option
INSET statement (UNIVARIATE), 235
CFRAME= option
HISTOGRAM statement (UNIVARIATE), 219
INSET statement (UNIVARIATE), 235
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259
CFRAMESIDE= option
HISTOGRAM statement (UNIVARIATE), 219
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259
CFRAMETOP= option
HISTOGRAM statement (UNIVARIATE), 219
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259

388

Syntax Index

CGRID= option
HISTOGRAM statement (UNIVARIATE), 219
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259
CHEADER= option
INSET statement (UNIVARIATE), 235
CHISQ option
EXACT statement (FREQ), 78, 169
OUTPUT statement (FREQ), 81
TABLES statement (FREQ), 88, 104, 169
CHREF= option
HISTOGRAM statement (UNIVARIATE), 219
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259
CIBASIC option
PROC UNIVARIATE statement, 204, 326
CIPCTLDF option
PROC UNIVARIATE statement, 204
CIPCTLNORMAL option
PROC UNIVARIATE statement, 205
CIQUANTDF option
PROC UNIVARIATE statement, 328
CIQUANTNORMAL option
PROC UNIVARIATE statement, 205, 328
CL option
TABLES statement (FREQ), 89
CLASS statement
UNIVARIATE procedure, 210
CMH option
OUTPUT statement (FREQ), 81
TABLES statement (FREQ), 89
CMH1 option
OUTPUT statement (FREQ), 81
TABLES statement (FREQ), 89
CMH2 option
OUTPUT statement (FREQ), 81
TABLES statement (FREQ), 89
CMHCOR option
OUTPUT statement (FREQ), 81
CMHGA option
OUTPUT statement (FREQ), 81
CMHRMS option
OUTPUT statement (FREQ), 81
COCHQ option
OUTPUT statement (FREQ), 81
COLOR= option
HISTOGRAM statement (UNIVARIATE), 219
PROBPLOT statement (UNIVARIATE), 246
QQPLOT statement (UNIVARIATE), 259
COMOR option
EXACT statement (FREQ), 78
COMPRESS option
PROC FREQ statement, 75
CONTENTS= option
TABLES statement (FREQ), 89
CONTGY option
OUTPUT statement (FREQ), 82
CONVERGE= option
TABLES statement (FREQ), 90

CORR procedure
syntax, 6
CORR procedure, BY statement, 12
CORR procedure, FREQ statement, 13
CORR procedure, PARTIAL statement, 13
CORR procedure, PLOTS option
NMAXVAR= option, 31, 32
NMAXWITH= option, 31, 32
CORR procedure, PLOTS=SCATTER option
ALPHA= option, 32
ELLIPSE= option, 32
NOINSET, 32
NOLEGEND, 32
CORR procedure, PROC CORR statement, 7
ALPHA option, 8
BEST= option, 8
COV option, 8
CSSCP option, 8
DATA= option, 9
EXCLNPWGT option, 9
FISHER option, 9
HOEFFDING option, 9
KENDALL option, 9
NOCORR option, 10
NOMISS option, 10
NOPRINT option, 10
NOPROB option, 10
NOSIMPLE option, 10
OUT= option, 10
OUTH= option, 10
OUTK= option, 10
OUTP= option, 10
OUTS= option, 10
PEARSON option, 11
RANK option, 11
SINGULAR= option, 11
SPEARMAN option, 11
SSCP option, 11
VARDEF= option, 11
CORR procedure, VAR statement, 13
CORR procedure, WEIGHT statement, 13
CORR procedure, WITH statement, 14
COV option
PROC CORR statement, 8
CPROP= option
HISTOGRAM statement (UNIVARIATE), 220,
345
CRAMV option
OUTPUT statement (FREQ), 82
CROSSLIST option
TABLES statement (FREQ), 90
CSHADOW= option
INSET statement (UNIVARIATE), 235
CSSCP option
PROC CORR statement, 8
CTEXT= option
HISTOGRAM statement (UNIVARIATE), 220
INSET statement (UNIVARIATE), 235
PROBPLOT statement (UNIVARIATE), 247

Syntax Index
QQPLOT statement (UNIVARIATE), 259
CTEXTSIDE= option
HISTOGRAM statement (UNIVARIATE), 220
CTEXTTOP= option
HISTOGRAM statement (UNIVARIATE), 220
CUMCOL option
TABLES statement (FREQ), 91
CVREF= option
HISTOGRAM statement (UNIVARIATE), 220
PROBPLOT statement (UNIVARIATE), 247
QQPLOT statement (UNIVARIATE), 259

D
DATA option
INSET statement (UNIVARIATE), 235
DATA= option
INSET statement (UNIVARIATE), 235
PROC CORR statement, 9
PROC FREQ statement, 75
PROC UNIVARIATE statement, 205, 305
DESCENDING option
BY statement (UNIVARIATE), 209
DESCRIPTION= option
HISTOGRAM statement (UNIVARIATE), 220
PROBPLOT statement (UNIVARIATE), 247
QQPLOT statement (UNIVARIATE), 260
DEVIATION option
TABLES statement (FREQ), 91

E
ELLIPSE= option
PLOTS=SCATTER option (CORR), 32
ENDPOINTS= option
HISTOGRAM statement (UNIVARIATE), 220,
340
EQKAP option
OUTPUT statement (FREQ), 82
EQWKP option
OUTPUT statement (FREQ), 82
EXACT option
OUTPUT statement (FREQ), 82
EXACT statement
FREQ procedure, 77
EXCLNPWGT option
PROC CORR statement, 9
PROC UNIVARIATE statement, 205
EXPECTED option
TABLES statement (FREQ), 91
EXPONENTIAL option
HISTOGRAM statement (UNIVARIATE), 221,
289
PROBPLOT statement (UNIVARIATE), 247,
301
QQPLOT statement (UNIVARIATE), 260, 301

F
FILL option
HISTOGRAM statement (UNIVARIATE), 221
FISHER option

389

EXACT statement (FREQ), 78

OUTPUT statement (FREQ), 82
PROC CORR statement, 9
TABLES statement (FREQ), 91
FONT= option
HISTOGRAM statement (UNIVARIATE), 221
INSET statement (UNIVARIATE), 236
PROBPLOT statement (UNIVARIATE), 247
QQPLOT statement (UNIVARIATE), 260
FORCEHIST option
HISTOGRAM statement (UNIVARIATE), 222
FORMAT= option
INSET statement (UNIVARIATE), 236
TABLES statement (FREQ), 91
FORMCHAR= option
PROC FREQ statement, 75
FREQ option
PROC UNIVARIATE statement, 205, 317
FREQ procedure
syntax, 74
FREQ procedure, BY statement, 77
FREQ procedure, EXACT statement, 77
AGREE option, 78
ALPHA= option, 79
BINOMIAL option, 78
CHISQ option, 78, 169
COMOR option, 78
FISHER option, 78
JT option, 78
KAPPA option, 78
LRCHI option, 78
MAXTIME= option, 79
MC option, 79
MCNEM option, 78
MEASURES option, 78
MHCHI option, 78
N= option, 79
OR option, 78, 169
PCHI option, 78
PCORR option, 78
POINT option, 79
SCORR option, 78
SEED= option, 79
TREND option, 78, 177
WTKAP option, 78
FREQ procedure, OUTPUT statement, 80
AGREE option, 81
AJCHI option, 81
ALL option, 81
BDCHI option, 81
BINOMIAL option, 81
CHISQ option, 81
CMH option, 81
CMH1 option, 81
CMH2 option, 81
CMHCOR option, 81
CMHGA option, 81
CMHRMS option, 81
COCHQ option, 81

390

Syntax Index

CONTGY option, 82
CRAMV option, 82
EQKAP option, 82
EQWKP option, 82
EXACT option, 82
FISHER option, 82
GAMMA option, 82
JT option, 82
KAPPA option, 82
KENTB option, 82
LAMCR option, 82
LAMDAS option, 82
LAMRC option, 82
LGOR option, 82
LGRRC1 option, 82
LGRRC2 option, 82
LRCHI option, 82, 173
MCNEM option, 82
MEASURES option, 82
MHCHI option, 82
MHOR option, 82
MHRRC1 option, 82
MHRRC2 option, 82
N option, 82
NMISS option, 82
OR option, 83
OUT= option, 80
PCHI option, 83, 173
PCORR option, 83
PHI option, 83
PLCORR option, 83
RDIF1 option, 83
RDIF2 option, 83
RELRISK option, 83
RISKDIFF option, 83
RISKDIFF1 option, 83
RISKDIFF2 option, 83
RRC1 option, 83
RRC2 option, 83
RSK1 option, 83
RSK11 option, 83
RSK12 option, 83
RSK2 option, 83
RSK21 option, 83
RSK22 option, 83
SCORR option, 83
SMDCR option, 83
SMDRC option, 83
STUTC option, 83
TREND option, 83
TSYMM option, 83
U option, 83
UCR option, 83
URC option, 83
WTKAP option, 83
FREQ procedure, PROC FREQ statement, 75
COMPRESS option, 75
DATA= option, 75
FORMCHAR= option, 75

NLEVELS option, 76
NOPRINT option, 76
ORDER= option, 76
PAGE option, 77
FREQ procedure, TABLES statement, 84
ALL option, 87
ALPHA= option, 87
BDT option, 87
BINOMIAL option, 87, 166
BINOMIALC option, 88
CELLCHI2 option, 88
CHISQ option, 88, 104, 169
CL option, 89
CMH option, 89
CMH1 option, 89
CMH2 option, 89
CONTENTS= option, 89
CONVERGE= option, 90
CROSSLIST option, 90
CUMCOL option, 91
DEVIATION option, 91
EXPECTED option, 91
FISHER option, 91
FORMAT= option, 91
JT option, 91
LIST option, 91
MAXITER= option, 92
MEASURES option, 92
MISSING option, 92
MISSPRINT option, 92
NOCOL option, 92
NOCUM option, 92
NOFREQ option, 92
NOPERCENT option, 92
NOPRINT option, 93
NOROW option, 93
NOSPARSE option, 93
NOWARN option, 93
option, 87
OUT= option, 93
OUTCUM option, 93
OUTEXPECT option, 93, 161
OUTPCT option, 94
PLCORR option, 94
PRINTKWT option, 94
RELRISK option, 94, 169
RISKDIFF option, 94
RISKDIFFC option, 94
SCORES= option, 95, 181
SCOROUT option, 95
SPARSE option, 95, 161
TESTF= option, 96, 104
TESTP= option, 96, 104, 164
TOTPCT option, 96
TREND option, 96, 177
FREQ procedure, TEST statement, 96
AGREE option, 97
GAMMA option, 97
KAPPA option, 97

Syntax Index
KENTB option, 97
MEASURES option, 97
PCORR option, 97
SCORR option, 97
SMDCR option, 97, 177
SMDRC option, 97
STUTC option, 97
WTKAP option, 97
FREQ procedure, WEIGHT statement, 97
ZEROS option, 98
FREQ statement
CORR procedure, 13
UNIVARIATE procedure, 212
FRONTREF option
HISTOGRAM statement (UNIVARIATE), 222

G
GAMMA option
HISTOGRAM statement (UNIVARIATE), 222,
290, 348
OUTPUT statement (FREQ), 82
PROBPLOT statement (UNIVARIATE), 247,
301
QQPLOT statement (UNIVARIATE), 260, 301
TEST statement (FREQ), 97
GOUT= option
PROC UNIVARIATE statement, 205
GRID option
HISTOGRAM statement (UNIVARIATE), 222
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 260, 263,
372
GRIDCHAR= option
QQPLOT statement (UNIVARIATE), 263

H
HEADER= option
INSET statement (UNIVARIATE), 236
HEIGHT= option
HISTOGRAM statement (UNIVARIATE), 222
INSET statement (UNIVARIATE), 236
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 260
HISTOGRAM statement
UNIVARIATE procedure, 212
HMINOR= option
HISTOGRAM statement (UNIVARIATE), 222
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 260
HOEFFDING option
PROC CORR statement, 9
HOFFSET= option
HISTOGRAM statement (UNIVARIATE), 222
HREF= option
HISTOGRAM statement (UNIVARIATE), 222
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 261
HREFLABELS= option
HISTOGRAM statement (UNIVARIATE), 222

391

PROBPLOT statement (UNIVARIATE), 248

QQPLOT statement (UNIVARIATE), 261
HREFLABPOS= option
HISTOGRAM statement (UNIVARIATE), 223
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 261

I
ID statement
UNIVARIATE procedure, 230
INFONT= option
HISTOGRAM statement (UNIVARIATE), 223
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 261
INHEIGHT= option
HISTOGRAM statement (UNIVARIATE), 223
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 261
INSET statement
UNIVARIATE procedure, 230
INTERTILE= option
HISTOGRAM statement (UNIVARIATE), 223,
345
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 261

J
JT option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 82
TABLES statement (FREQ), 91

K
K= option
HISTOGRAM statement (UNIVARIATE), 223,
297
KAPPA option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 82
TEST statement (FREQ), 97
KENDALL option
PROC CORR statement, 9
KENTB option
OUTPUT statement (FREQ), 82
TEST statement (FREQ), 97
KERNEL option
HISTOGRAM statement (UNIVARIATE), 223,
297, 352
KEYLEVEL= option
CLASS statement (UNIVARIATE), 211
PROC UNIVARIATE statement, 337

L
L= option
HISTOGRAM statement (UNIVARIATE), 224
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 261
LABEL= option
QQPLOT statement (UNIVARIATE), 263

392

Syntax Index

LAMCR option
OUTPUT statement (FREQ), 82
LAMDAS option
OUTPUT statement (FREQ), 82
LAMRC option
OUTPUT statement (FREQ), 82
LGOR option
OUTPUT statement (FREQ), 82
LGRID= option
HISTOGRAM statement (UNIVARIATE), 224
PROBPLOT statement (UNIVARIATE), 248
QQPLOT statement (UNIVARIATE), 261, 263
LGRRC1 option
OUTPUT statement (FREQ), 82
LGRRC2 option
OUTPUT statement (FREQ), 82
LHREF= option
HISTOGRAM statement (UNIVARIATE), 224
PROBPLOT statement (UNIVARIATE), 249
QQPLOT statement (UNIVARIATE), 261
LIST option
TABLES statement (FREQ), 91
LOCCOUNT option
PROC UNIVARIATE statement, 205, 331
LOGNORMAL option
HISTOGRAM statement (UNIVARIATE), 224,
291, 348, 354, 363
PROBPLOT statement (UNIVARIATE), 249,
302, 360
QQPLOT statement (UNIVARIATE), 261, 302
LOWER= option
HISTOGRAM statement (UNIVARIATE), 224
LRCHI option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 82, 173
LVREF= option
HISTOGRAM statement (UNIVARIATE), 224
PROBPLOT statement (UNIVARIATE), 249
QQPLOT statement (UNIVARIATE), 262

M
MAXITER= option
TABLES statement (FREQ), 92
MAXNBIN= option
HISTOGRAM statement (UNIVARIATE), 224
MAXSIGMAS= option
HISTOGRAM statement (UNIVARIATE), 224
MAXTIME= option
EXACT statement (FREQ), 79
MC option
EXACT statement (FREQ), 79
MCNEM option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 82
MEASURES option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 82
TABLES statement (FREQ), 92
TEST statement (FREQ), 97

MHCHI option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 82
MHOR option
OUTPUT statement (FREQ), 82
MHRRC1 option
OUTPUT statement (FREQ), 82
MHRRC2 option
OUTPUT statement (FREQ), 82
MIDPERCENTS option
HISTOGRAM statement (UNIVARIATE), 225,
343
MIDPOINTS= option
HISTOGRAM statement (UNIVARIATE), 225,
338, 340
MISSING option
CLASS statement (UNIVARIATE), 210
TABLES statement (FREQ), 92
MISSPRINT option
TABLES statement (FREQ), 92
MODES option
PROC UNIVARIATE statement, 206, 314
MU0= option
PROC UNIVARIATE statement, 206
MU= option
HISTOGRAM statement (UNIVARIATE), 226,
343
PROBPLOT statement (UNIVARIATE), 249
QQPLOT statement (UNIVARIATE), 262, 366

N
N option
OUTPUT statement (FREQ), 82
N= option
EXACT statement (FREQ), 79
NADJ= option
PROBPLOT statement (UNIVARIATE), 249
QQPLOT statement (UNIVARIATE), 262, 298
NAME= option
HISTOGRAM statement (UNIVARIATE), 226
PROBPLOT statement (UNIVARIATE), 249
QQPLOT statement (UNIVARIATE), 262
NCOLS= option
HISTOGRAM statement (UNIVARIATE), 226
PROBPLOT statement (UNIVARIATE), 249
QQPLOT statement (UNIVARIATE), 262
NEXTROBS= option
PROC UNIVARIATE statement, 206, 315
NEXTRVAL= option
PROC UNIVARIATE statement, 206, 315
NLEVELS option
PROC FREQ statement, 76
NMAXVAR= option
PLOTS option (CORR), 31
PLOTS=SCATTER option (CORR), 32
NMAXWIDTH= option
PLOTS option (CORR), 31
PLOTS=SCATTER option (CORR), 32
NMISS option

Syntax Index
OUTPUT statement (FREQ), 82
NOBARS option
HISTOGRAM statement (UNIVARIATE), 226
NOBYPLOT option
PROC UNIVARIATE statement, 206
NOCOL option
TABLES statement (FREQ), 92
NOCORR option
PROC CORR statement, 10
NOCUM option
TABLES statement (FREQ), 92
NOFRAME option
HISTOGRAM statement (UNIVARIATE), 226
INSET statement (UNIVARIATE), 236
PROBPLOT statement (UNIVARIATE), 249
QQPLOT statement (UNIVARIATE), 262
NOFREQ option
TABLES statement (FREQ), 92
NOHLABEL option
HISTOGRAM statement (UNIVARIATE), 226
PROBPLOT statement (UNIVARIATE), 250
QQPLOT statement (UNIVARIATE), 262
NOINSET option
PLOTS=SCATTER option (CORR), 32
NOLEGEND option
PLOTS=SCATTER option (CORR), 32
NOMISS option
PROC CORR statement, 10
NOPERCENT option
TABLES statement (FREQ), 92
NOPLOT option
HISTOGRAM statement (UNIVARIATE), 226
NOPRINT option
HISTOGRAM statement (UNIVARIATE), 226
PROC CORR statement, 10
PROC FREQ statement, 76
PROC UNIVARIATE statement, 206
TABLES statement (FREQ), 93
NOPROB option
PROC CORR statement, 10
NORMAL option
HISTOGRAM statement (UNIVARIATE), 226,
291, 343
PROBPLOT statement (UNIVARIATE), 250,
302
PROC UNIVARIATE statement, 206
QQPLOT statement (UNIVARIATE), 262, 302
NORMALTEST option
PROC UNIVARIATE statement, 206
NOROW option
TABLES statement (FREQ), 93
NOSIMPLE option
PROC CORR statement, 10
NOSPARSE option
TABLES statement (FREQ), 93
NOTSORTED option
BY statement (UNIVARIATE), 209
NOVLABEL option
HISTOGRAM statement (UNIVARIATE), 226

393

PROBPLOT statement (UNIVARIATE), 250

QQPLOT statement (UNIVARIATE), 263
NOVTICK option
HISTOGRAM statement (UNIVARIATE), 226
PROBPLOT statement (UNIVARIATE), 250
QQPLOT statement (UNIVARIATE), 263
NOWARN option
TABLES statement (FREQ), 93
NROWS= option
HISTOGRAM statement (UNIVARIATE), 226,
334
PROBPLOT statement (UNIVARIATE), 250
QQPLOT statement (UNIVARIATE), 263

O
OR option
EXACT statement (FREQ), 78, 169
OUTPUT statement (FREQ), 83
ORDER= option
CLASS statement (UNIVARIATE), 210
PROC FREQ statement, 76
OUT= option
OUTPUT statement (FREQ), 80
OUTPUT statement (UNIVARIATE), 237
PROC CORR statement, 10
TABLES statement (FREQ), 93
OUTCUM option
TABLES statement (FREQ), 93
OUTEXPECT option
TABLES statement (FREQ), 93, 161
OUTH= option
PROC CORR statement, 10
OUTHISTOGRAM= option
HISTOGRAM statement (UNIVARIATE), 227,
308, 340
OUTK= option
PROC CORR statement, 10
OUTP= option
PROC CORR statement, 10
OUTPCT option
TABLES statement (FREQ), 94
OUTPUT statement
FREQ procedure, 80
UNIVARIATE procedure, 237, 267
OUTS= option
PROC CORR statement, 10

P
PAGE option
PROC FREQ statement, 77
PARTIAL statement
CORR procedure, 13
PCHI option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 83, 173
PCORR option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 83
TEST statement (FREQ), 97

394

Syntax Index

PCTLAXIS option
QQPLOT statement (UNIVARIATE), 263, 305,
372
PCTLDEF= option
PROC UNIVARIATE statement, 207, 273
PCTLMINOR option
PROBPLOT statement (UNIVARIATE), 250
QQPLOT statement (UNIVARIATE), 263
PCTLNAME= option
OUTPUT statement (UNIVARIATE), 240
PCTLORDER= option
PROBPLOT statement (UNIVARIATE), 250
PCTLPRE= option
OUTPUT statement (UNIVARIATE), 239
PCTLPTS= option
OUTPUT statement (UNIVARIATE), 239
PCTLSCALE option
QQPLOT statement (UNIVARIATE), 263, 305
PEARSON option
PROC CORR statement, 11
PERCENTS= option
HISTOGRAM statement (UNIVARIATE), 227
PFILL= option
HISTOGRAM statement (UNIVARIATE), 227
PHI option
OUTPUT statement (FREQ), 83
PLCORR option
OUTPUT statement (FREQ), 83
TABLES statement (FREQ), 94
PLOT option
PROC UNIVARIATE statement, 319
PLOTS option
PROC UNIVARIATE statement, 207
PLOTSIZE= option
PROC UNIVARIATE statement, 207
POINT option
EXACT statement (FREQ), 79
POSITION= option
INSET statement (UNIVARIATE), 236
PRINTKWT option
TABLES statement (FREQ), 94
PROBPLOT statement
UNIVARIATE procedure, 241
PROC CORR statement, 7,
See CORR procedure
CORR procedure, 7
PROC FREQ statement,
See FREQ procedure
PROC UNIVARIATE statement, 203,
See UNIVARIATE procedure

Q
QQPLOT statement
UNIVARIATE procedure, 253

R
RANK option
PROC CORR statement, 11
RANKADJ= option

PROBPLOT statement (UNIVARIATE), 250

QQPLOT statement (UNIVARIATE), 263, 298
RDIF1 option
OUTPUT statement (FREQ), 83
RDIF2 option
OUTPUT statement (FREQ), 83
REFPOINT= option
INSET statement (UNIVARIATE), 236
RELRISK option
OUTPUT statement (FREQ), 83
TABLES statement (FREQ), 94, 169
RISKDIFF option
OUTPUT statement (FREQ), 83
TABLES statement (FREQ), 94
RISKDIFF1 option
OUTPUT statement (FREQ), 83
RISKDIFF2 option
OUTPUT statement (FREQ), 83
RISKDIFFC option
TABLES statement (FREQ), 94
ROBUSTSCALE option
PROC UNIVARIATE statement, 207, 329
ROUND= option
PROC UNIVARIATE statement, 207
RRC1 option
OUTPUT statement (FREQ), 83
RRC2 option
OUTPUT statement (FREQ), 83
RSK1 option
OUTPUT statement (FREQ), 83
RSK11 option
OUTPUT statement (FREQ), 83
RSK12 option
OUTPUT statement (FREQ), 83
RSK2 option
OUTPUT statement (FREQ), 83
RSK21 option
OUTPUT statement (FREQ), 83
RSK22 option
OUTPUT statement (FREQ), 83
RTINCLUDE option
HISTOGRAM statement (UNIVARIATE), 227,
340

S
SCALE= option
HISTOGRAM statement (UNIVARIATE), 227,
289, 290, 346
PROBPLOT statement (UNIVARIATE), 250
QQPLOT statement (UNIVARIATE), 263
SCORES= option
TABLES statement (FREQ), 95, 181
SCOROUT option
TABLES statement (FREQ), 95
SCORR option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 83
TEST statement (FREQ), 97
SEED= option

Syntax Index
EXACT statement (FREQ), 79
SHAPE= option
HISTOGRAM statement (UNIVARIATE), 227
PROBPLOT statement (UNIVARIATE), 251
QQPLOT statement (UNIVARIATE), 264
SIGMA= option
HISTOGRAM statement (UNIVARIATE), 227,
289, 343
PROBPLOT statement (UNIVARIATE), 251,
360
QQPLOT statement (UNIVARIATE), 264, 366,
369
SINGULAR= option
PROC CORR statement, 11
SLOPE= option
PROBPLOT statement (UNIVARIATE), 251
QQPLOT statement (UNIVARIATE), 264
SMDCR option
OUTPUT statement (FREQ), 83
TEST statement (FREQ), 97, 177
SMDRC option
OUTPUT statement (FREQ), 83
TEST statement (FREQ), 97
SPARSE option
TABLES statement (FREQ), 95, 161
SPEARMAN option
PROC CORR statement, 11
SQUARE option
PROBPLOT statement (UNIVARIATE), 251,
360
QQPLOT statement, 366
QQPLOT statement (UNIVARIATE), 265
SSCP option
PROC CORR statement, 11
STUTC option
OUTPUT statement (FREQ), 83
TEST statement (FREQ), 97

T
TABLES statement
FREQ procedure, 84
TEST statement
FREQ procedure, 96
TESTF= option
TABLES statement (FREQ), 96, 104
TESTP= option
TABLES statement (FREQ), 96, 104, 164
THETA= option
HISTOGRAM statement (UNIVARIATE), 227,
289, 346, 354, 363
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
THRESHOLD= option
HISTOGRAM statement (UNIVARIATE), 228,
290
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
TOTPCT option
TABLES statement (FREQ), 96

395

TREND option
EXACT statement (FREQ), 78, 177
OUTPUT statement (FREQ), 83
TABLES statement (FREQ), 96, 177
TRIMMED= option
PROC UNIVARIATE statement, 207, 329
TSYMM option
OUTPUT statement (FREQ), 83
TURNVLABELS option
HISTOGRAM statement (UNIVARIATE), 228

U
U option
OUTPUT statement (FREQ), 83
UCR option
OUTPUT statement (FREQ), 83
UNIVARIATE procedure
syntax, 202
UNIVARIATE procedure, BY statement, 209
DESCENDING option, 209
NOTSORTED option, 209
UNIVARIATE procedure, CLASS statement, 210
KEYLEVEL= option, 211
MISSING option, 210
ORDER= option, 210
UNIVARIATE procedure, FREQ statement, 212
UNIVARIATE procedure, HISTOGRAM statement,
212
ALPHA= option, 217, 289
ANNOKEY option, 217
ANNOTATE= option, 218, 355
BARWIDTH= option, 218
BETA option, 218, 288, 346
BETA= option, 218, 289
C= option, 218, 297, 352
CAXIS= option, 219
CBARLINE= option, 219
CFILL= option, 219
CFRAME= option, 219
CFRAMESIDE= option, 219
CFRAMETOP= option, 219
CGRID= option, 219
CHREF= option, 219
COLOR= option, 219
CPROP= option, 220, 345
CTEXT= option, 220
CTEXTSIDE= option, 220
CTEXTTOP= option, 220
CVREF= option, 220
DESCRIPTION= option, 220
ENDPOINTS= option, 220, 340
EXPONENTIAL option, 221, 289
FILL option, 221
FONT= option, 221
FORCEHIST option, 222
FRONTREF option, 222
GAMMA option, 222, 290, 348
GRID option, 222
HEIGHT= option, 222

396

Syntax Index

HMINOR= option, 222

HOFFSET= option, 222
HREF= option, 222
HREFLABELS= option, 222
HREFLABPOS= option, 223
INFONT= option, 223
INHEIGHT= option, 223
INTERTILE= option, 223, 345
K= option, 223, 297
KERNEL option, 223, 297, 352
L= option, 224
LGRID= option, 224
LHREF= option, 224
LOGNORMAL option, 224, 291, 348, 354, 363
LOWER= option, 224
LVREF= option, 224
MAXNBIN= option, 224
MAXSIGMAS= option, 224
MIDPERCENTS option, 225, 343
MIDPOINTS= option, 225, 338, 340
MU= option, 226, 343
NAME= option, 226
NCOLS= option, 226
NOBARS option, 226
NOFRAME option, 226
NOHLABEL option, 226
NOPLOT option, 226
NOPRINT option, 226
NORMAL option, 226, 291, 343
NOVLABEL option, 226
NOVTICK option, 226
NROWS= option, 226, 334
OUTHISTOGRAM= option, 227, 308, 340
PERCENTS= option, 227
PFILL= option, 227
RTINCLUDE option, 227, 340
SCALE= option, 227, 289, 290, 346
SHAPE= option, 227
SIGMA= option, 227, 289, 343
THETA= option, 227, 289, 346, 354, 363
THRESHOLD= option, 228, 290
TURNVLABELS option, 228
UPPER= option, 228
VAXIS= option, 228
VAXISLABEL= option, 228
VMINOR= option, 228
VOFFSET= option, 228
VREF= option, 228
VREFLABELS= option, 228
VREFLABPOS= option, 229
VSCALE= option, 229
W= option, 229
WAXIS= option, 229
WBARLINE= option, 229
WEIBULL option, 229, 292, 348
WGRID= option, 229
ZETA= option, 229
UNIVARIATE procedure, ID statement, 230
UNIVARIATE procedure, INSET statement, 230

CFILL= option, 235

CFILLH= option, 235
CFRAME= option, 235
CHEADER= option, 235
CSHADOW= option, 235
CTEXT= option, 235
DATA option, 235
DATA= option, 235
FONT= option, 236
FORMAT= option, 236
HEADER= option, 236
HEIGHT= option, 236
NOFRAME option, 236
POSITION= option, 236
REFPOINT= option, 236
UNIVARIATE procedure, OUTPUT statement, 237,
267
OUT= option, 237
PCTLNAME= option, 240
PCTLPRE= option, 239
PCTLPTS= option, 239
UNIVARIATE procedure, PROBPLOT statement, 241
ALPHA= option, 245
ANNOKEY option, 245
ANNOTATE= option, 245
BETA option, 245, 301
BETA= option, 245
C= option, 246
CAXIS= option, 246
CFRAME= option, 246
CFRAMESIDE= option, 246
CFRAMETOP= option, 246
CGRID= option, 246
CHREF= option, 246
COLOR= option, 246
CTEXT= option, 247
CVREF= option, 247
DESCRIPTION= option, 247
EXPONENTIAL option, 247, 301
FONT= option, 247
GAMMA option, 247, 301
GRID option, 248
HEIGHT= option, 248
HMINOR= option, 248
HREF= option, 248
HREFLABELS= option, 248
HREFLABPOS= option, 248
INFONT= option, 248
INHEIGHT= option, 248
INTERTILE= option, 248
L= option, 248
LGRID= option, 248
LHREF= option, 249
LOGNORMAL option, 249, 302, 360
LVREF= option, 249
MU= option, 249
NADJ= option, 249
NAME= option, 249
NCOLS= option, 249

Syntax Index
NOFRAME option, 249
NOHLABEL option, 250
NORMAL option, 250, 302
NOVLABEL option, 250
NOVTICK option, 250
NROWS= option, 250
PCTLMINOR option, 250
PCTORDER= option, 250
RANKADJ= option, 250
SCALE= option, 250
SHAPE= option, 251
SIGMA= option, 251, 360
SLOPE= option, 251
SQUARE option, 251, 360
THETA= option, 252
THRESHOLD= option, 252
VAXISLABEL= option, 252
VMINOR= option, 252
VREF= option, 252
VREFLABELS= option, 252
VREFLABPOS= option, 252
W= option, 252
WAXIS= option, 252
WEIBULL option, 253, 302
WEIBULL2 option, 303
WEIBULL2 statement, 253
ZETA= option, 253
UNIVARIATE procedure, PROC UNIVARIATE
statement, 203
ALL option, 204
ALPHA= option, 204
ANNOTATE= option, 204, 305
CIBASIC option, 204, 326
CIPCTLDF option, 204
CIPCTLNORMAL option, 205
CIQUANTDF option, 328
CIQUANTNORMAL option, 205, 328
DATA= option, 205, 305
EXCLNPWGT option, 205
FREQ option, 205, 317
GOUT= option, 205
KEYLEVEL= option, 337
LOCCOUNT option, 205, 331
MODES option, 206, 314
MU0= option, 206
NEXTROBS= option, 206, 315
NEXTRVAL= option, 206, 315
NOBYPLOT option, 206
NOPRINT option, 206
NORMAL option, 206
NORMALTEST option, 206
PCTLDEF= option, 207, 273
PLOT option, 319
PLOTS option, 207
PLOTSIZE= option, 207
ROBUSTSCALE option, 207, 329
ROUND= option, 207
TRIMMED= option, 207, 329
VARDEF= option, 208

397

WINSORIZED= option, 209, 329

UNIVARIATE procedure, QQPLOT statement, 253
ALPHA= option, 258
ANNOKEY option, 258
ANNOTATE= option, 258
BETA option, 258, 301
BETA= option, 258
C= option, 259, 375
CAXIS= option, 259
CFRAME= option, 259
CFRAMESIDE= option, 259
CFRAMETOP= option, 259
CGRID= option, 259
CHREF= option, 259
COLOR= option, 259
CTEXT= option, 259
CVREF= option, 259
DESCRIPTION= option, 260
EXPONENTIAL option, 260, 301
FONT= option, 260
GAMMA option, 260, 301
GRID option, 260, 263, 372
GRIDCHAR= option, 263
HEIGHT= option, 260
HMINOR= option, 260
HREF= option, 261
HREFLABELS= option, 261
HREFLABPOS= option, 261
INFONT= option, 261
INHEIGHT= option, 261
INTERTILE= option, 261
L= option, 261
LABEL= option, 263
LGRID= option, 261, 263
LHREF= option, 261
LOGNORMAL option, 261, 302
LVREF= option, 262
MU= option, 262, 366
NADJ= option, 262, 298
NAME= option, 262
NCOLS= option, 262
NOFRAME option, 262
NOHLABEL option, 262
NORMAL option, 262, 302
NOVLABEL option, 263
NOVTICK option, 263
NROWS= option, 263
PCTLAXIS option, 263, 305, 372
PCTLMINOR option, 263
PCTLSCALE option, 263, 305
RANKADJ= option, 263, 298
SCALE= option, 263
SHAPE= option, 264
SIGMA= option, 264, 366, 369
SLOPE= option, 264
SQUARE option, 265, 366
THETA= option, 265
THRESHOLD= option, 265
VAXISLABEL= option, 265

398

Syntax Index

VMINOR= option, 265

VREF= option, 265
VREFLABELS= option, 265
VREFLABPOS= option, 265
W= option, 265
WAXIS= option, 265
WEIBULL option, 266, 302, 375
WEIBULL2 option, 303
WEIBULL2 statement, 266
ZETA= option, 266, 369
UNIVARIATE procedure, VAR statement, 266
UNIVARIATE procedure, WEIGHT statement, 267
UPPER= option
HISTOGRAM statement (UNIVARIATE), 228
URC option
OUTPUT statement (FREQ), 83

V
VAR statement
CORR procedure, 13
UNIVARIATE procedure, 266
VARDEF= option
PROC CORR statement, 11
PROC UNIVARIATE statement, 208
VAXIS= option
HISTOGRAM statement (UNIVARIATE), 228
VAXISLABEL= option
HISTOGRAM statement (UNIVARIATE), 228
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
VMINOR= option
HISTOGRAM statement (UNIVARIATE), 228
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
VOFFSET= option
HISTOGRAM statement (UNIVARIATE), 228
VREF= option
HISTOGRAM statement (UNIVARIATE), 228
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
VREFLABELS= option
HISTOGRAM statement (UNIVARIATE), 228
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
VREFLABPOS= option
HISTOGRAM statement (UNIVARIATE), 229
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
VSCALE= option
HISTOGRAM statement (UNIVARIATE), 229

W
W= option
HISTOGRAM statement (UNIVARIATE), 229
PROBPLOT statement (UNIVARIATE), 252
QQPLOT statement (UNIVARIATE), 265
WAXIS= option
HISTOGRAM statement (UNIVARIATE), 229
PROBPLOT statement (UNIVARIATE), 252

QQPLOT statement (UNIVARIATE), 265

WBARLINE= option
HISTOGRAM statement (UNIVARIATE), 229
WEIBULL option
HISTOGRAM statement (UNIVARIATE), 229,
292, 348
PROBPLOT statement (UNIVARIATE), 253,
302
QQPLOT statement (UNIVARIATE), 266, 302,
375
WEIBULL2 option
PROBPLOT statement (UNIVARIATE), 253,
303
QQPLOT statement (UNIVARIATE), 266, 303
WEIGHT statement
CORR procedure, 13
FREQ procedure, 97
UNIVARIATE procedure, 267
WGRID= option
HISTOGRAM statement (UNIVARIATE), 229
WINSORIZED= option
PROC UNIVARIATE statement, 209, 329
WITH statement
CORR procedure, 14
WTKAP option
EXACT statement (FREQ), 78
OUTPUT statement (FREQ), 83
TEST statement (FREQ), 97

Z
ZEROS option
WEIGHT statement (FREQ), 98
ZETA= option
HISTOGRAM statement (UNIVARIATE), 229
PROBPLOT statement (UNIVARIATE), 253
QQPLOT statement (UNIVARIATE), 266, 369

IBM I DB2 Web Query For I Version 2.1 Implementation Guide
No ratings yet
IBM I DB2 Web Query For I Version 2.1 Implementation Guide
880 pages
Sas Course Content Porak Technologies
No ratings yet
Sas Course Content Porak Technologies
8 pages
Sas Documentation
No ratings yet
Sas Documentation
742 pages
SAS Base Certification Preparatio Document - Part 1
No ratings yet
SAS Base Certification Preparatio Document - Part 1
50 pages
Windows Management Instrumentation
No ratings yet
Windows Management Instrumentation
10 pages
Base SAS Syllabus
No ratings yet
Base SAS Syllabus
2 pages
A+ Core Hardware
No ratings yet
A+ Core Hardware
12 pages
Catalyst Command Line Interface
No ratings yet
Catalyst Command Line Interface
8 pages
NTFS Permissions
No ratings yet
NTFS Permissions
21 pages
Embedded Operating Systems and Linux: Presented by Aaruni Giriraj Viiith Sem I.T
No ratings yet
Embedded Operating Systems and Linux: Presented by Aaruni Giriraj Viiith Sem I.T
24 pages
Windows Security: Fore Academy Security Essentials (V)
No ratings yet
Windows Security: Fore Academy Security Essentials (V)
17 pages
Manage Azure Ad Users and Groups Slides
No ratings yet
Manage Azure Ad Users and Groups Slides
18 pages
Pycharm Shortcuts
No ratings yet
Pycharm Shortcuts
4 pages
Creating Users and Groups
No ratings yet
Creating Users and Groups
23 pages
Network+ Lesson 2
No ratings yet
Network+ Lesson 2
70 pages
02 - XML - Fundamentals
No ratings yet
02 - XML - Fundamentals
26 pages
Chapter 1 - Introduction To Blockchain Technology
No ratings yet
Chapter 1 - Introduction To Blockchain Technology
10 pages
Topic 6.0 Network Troubleshooting
No ratings yet
Topic 6.0 Network Troubleshooting
22 pages
Pmad61 001 WM
No ratings yet
Pmad61 001 WM
244 pages
VMware PowerCli Get-Help
No ratings yet
VMware PowerCli Get-Help
63 pages
Microsoft Office 365 Plans
No ratings yet
Microsoft Office 365 Plans
13 pages
Lab DOS Command Line Assignment Instructions
No ratings yet
Lab DOS Command Line Assignment Instructions
11 pages
02 Java Fundamentals
No ratings yet
02 Java Fundamentals
79 pages
Network Security For Virutal Machine in Cloud Computing
No ratings yet
Network Security For Virutal Machine in Cloud Computing
4 pages
10RX HowTo Configure Syslog
100% (1)
10RX HowTo Configure Syslog
18 pages
Structured Query Language (SQL)
No ratings yet
Structured Query Language (SQL)
145 pages
MySQL Data Types Quick Reference Table
No ratings yet
MySQL Data Types Quick Reference Table
3 pages
SQLsecrets Revealed
No ratings yet
SQLsecrets Revealed
103 pages
DataStage Stages 12-Dec-2013 12PM
No ratings yet
DataStage Stages 12-Dec-2013 12PM
47 pages
These Interview Questions With Answers Will Test Your Basic Knowledge of PC and Networking
No ratings yet
These Interview Questions With Answers Will Test Your Basic Knowledge of PC and Networking
4 pages
Lect 08 - Configuring Domain Name Service
No ratings yet
Lect 08 - Configuring Domain Name Service
52 pages
Malware Hunting With The Sysinternals Tools !
100% (1)
Malware Hunting With The Sysinternals Tools !
23 pages
Application Development For CICS
No ratings yet
Application Development For CICS
384 pages
Domain Name System
No ratings yet
Domain Name System
5 pages
Lauren Haworth, Genentech, Inc., South San Francisco, CA: ODS RTF: The Basics and Beyond
No ratings yet
Lauren Haworth, Genentech, Inc., South San Francisco, CA: ODS RTF: The Basics and Beyond
19 pages
Mana Mohan R
No ratings yet
Mana Mohan R
147 pages
Structured Query Language
No ratings yet
Structured Query Language
26 pages
Lecture 4 - Loops and Files
No ratings yet
Lecture 4 - Loops and Files
51 pages
MySQL To DB2 Conversion Guide (sg247093) PDF
No ratings yet
MySQL To DB2 Conversion Guide (sg247093) PDF
476 pages
SQL Statements: - Select - Insert - Update - Delete - Create - Alter - Drop - Rename - Truncate - Commit - Rollback - Savepoint
100% (1)
SQL Statements: - Select - Insert - Update - Delete - Create - Alter - Drop - Rename - Truncate - Commit - Rollback - Savepoint
231 pages
Crystal Report Server PDF
No ratings yet
Crystal Report Server PDF
16 pages
Dates Format in Sas
No ratings yet
Dates Format in Sas
6 pages
Microsoft Office 365 Pro Plus Product Key
No ratings yet
Microsoft Office 365 Pro Plus Product Key
1 page
C++ Lab 01 - Basic Program Structure
No ratings yet
C++ Lab 01 - Basic Program Structure
9 pages
Vmware Vsphere 6.7.X On Dell Emc Poweredge Servers: Compatibility Matrix
No ratings yet
Vmware Vsphere 6.7.X On Dell Emc Poweredge Servers: Compatibility Matrix
26 pages
SQL
No ratings yet
SQL
88 pages
MYSQL
No ratings yet
MYSQL
69 pages
Presentation CHAPTER1
No ratings yet
Presentation CHAPTER1
28 pages
Windows Server 2012 First Domain Controller
No ratings yet
Windows Server 2012 First Domain Controller
15 pages
SQL Injection
100% (1)
SQL Injection
10 pages
Base Sas Programming 2019 10 04 13 47 36 122
No ratings yet
Base Sas Programming 2019 10 04 13 47 36 122
281 pages
Db2 9 Fundamentals Exam 730 Prep, Part 7:: Introducing Xquery
No ratings yet
Db2 9 Fundamentals Exam 730 Prep, Part 7:: Introducing Xquery
28 pages
Notes in LINUX
No ratings yet
Notes in LINUX
3 pages
Command-Line Basics - Deep Security
No ratings yet
Command-Line Basics - Deep Security
12 pages
Printer Troubleshooting Host
No ratings yet
Printer Troubleshooting Host
22 pages
Vsphere Esxi Vcenter Server 602 Installation Setup Guide PDF
No ratings yet
Vsphere Esxi Vcenter Server 602 Installation Setup Guide PDF
298 pages
Python Programming
No ratings yet
Python Programming
17 pages
Mastering Active Directory
From Everand
Mastering Active Directory
VICTOR P HENDERSON
No ratings yet
Mastering SaltStack
From Everand
Mastering SaltStack
Joseph Hall
No ratings yet
Oracle VM Manager 2.1.2
From Everand
Oracle VM Manager 2.1.2
Tarry Singh
No ratings yet
SAS#14-ACC 116
No ratings yet
SAS#14-ACC 116
13 pages
M - SC Garment Manufacturing Technology (GMT) (Scheme & Syllabus 1st To 4th Sem)
No ratings yet
M - SC Garment Manufacturing Technology (GMT) (Scheme & Syllabus 1st To 4th Sem)
105 pages
Art of Reabanks
100% (1)
Art of Reabanks
5 pages
Data Analytics Unit 1
No ratings yet
Data Analytics Unit 1
10 pages
7.ms excel
No ratings yet
7.ms excel
20 pages
AutoXlsTable Manual
No ratings yet
AutoXlsTable Manual
21 pages
MS Excel Lesson 2
No ratings yet
MS Excel Lesson 2
11 pages
Marketing Management For Engineering Managers PDF
No ratings yet
Marketing Management For Engineering Managers PDF
5 pages
Vertex42 Money Manager 2.1: INSTRUCTIONS - For Excel 2010 or Later
No ratings yet
Vertex42 Money Manager 2.1: INSTRUCTIONS - For Excel 2010 or Later
29 pages
The SAP Killer: Build or Migrate Any DBMS or MIS Up To A Full Blown ERP
No ratings yet
The SAP Killer: Build or Migrate Any DBMS or MIS Up To A Full Blown ERP
41 pages
Week 8 Assignment - Portfolio Analysis
No ratings yet
Week 8 Assignment - Portfolio Analysis
3 pages
Focus Areas 2020-21 Accountancy CA PDF
No ratings yet
Focus Areas 2020-21 Accountancy CA PDF
4 pages
IT-402-X-2024-25-SampleQP-1
No ratings yet
IT-402-X-2024-25-SampleQP-1
7 pages
The Python Advantage Python For Excel In 2024 Hayden Van Der Post instant download
No ratings yet
The Python Advantage Python For Excel In 2024 Hayden Van Der Post instant download
52 pages
Computer Applications That Support Nursing Research Literature
No ratings yet
Computer Applications That Support Nursing Research Literature
8 pages
Mediacharge Trajectories
No ratings yet
Mediacharge Trajectories
8 pages
Sales Admin Specialist Job Description
No ratings yet
Sales Admin Specialist Job Description
2 pages
Data Cleansing Guidelines v2
No ratings yet
Data Cleansing Guidelines v2
8 pages
73 220 Lecture05
No ratings yet
73 220 Lecture05
13 pages
The Anaplan Platform Explained
No ratings yet
The Anaplan Platform Explained
3 pages
Panduan STID1103 MS Excel
No ratings yet
Panduan STID1103 MS Excel
372 pages
Excel Notes PDF
No ratings yet
Excel Notes PDF
59 pages
DELTA_IA_AM_ESP_UM_EN_20230913
No ratings yet
DELTA_IA_AM_ESP_UM_EN_20230913
60 pages
Class 10 1st Summative Suggestion 2025
No ratings yet
Class 10 1st Summative Suggestion 2025
14 pages
Class VI, Holiday Homework R
No ratings yet
Class VI, Holiday Homework R
11 pages
MODULE 2_(Ask Questions to Make Data-Driven Decisions)
No ratings yet
MODULE 2_(Ask Questions to Make Data-Driven Decisions)
23 pages
JournalEntry-2
No ratings yet
JournalEntry-2
14 pages
DISC112 - Computer and Problem Solving - Mahira Ilyas Spring 2020
No ratings yet
DISC112 - Computer and Problem Solving - Mahira Ilyas Spring 2020
5 pages
CSEC-INFORMATION-TECHNOLOGY-SBA-2025 - new
No ratings yet
CSEC-INFORMATION-TECHNOLOGY-SBA-2025 - new
16 pages
Oracle PSB Setup Appsworld
No ratings yet
Oracle PSB Setup Appsworld
28 pages