new script

Author: Jeff Stokes

articles/hdinsight/hdinsight-hadoop-r-server-get-started.md

@@ -376,29 +376,39 @@ The following provides some sample code on use of the new functions:
 myHadoopCluster <- rxSparkConnect(reset = TRUE)
 
 
+```
+#..create a Spark compute context
+
+myHadoopCluster <- rxSparkConnect(reset = TRUE)
+rxSetComputeContext(myHadoopCluster)
+```
+
+
 ```
 #..retrieve some sample data from Hive and run a model 
 
 hiveData <- RxHiveData("select * from hivesampletable", 
                  colInfo = list(devicemake = list(type = "factor")))
 rxGetInfo(hiveData, getVarInfo = TRUE)
 
-rxSetComputeContext(myHadoopCluster)
 rxLinMod(querydwelltime ~ devicemake, data=hiveData)
 ```
 
+
 ```
-#..retrieve some sample data from Parquet and run a model
-pqData <- RxParquetData("/share/SampleData/AirlineDemoSmallParquet")
-rxGetInfo(pqData, getVarInfo = TRUE)
+#..retrieve some sample data from Parquet and run a model 
 
-rxSetComputeContext(myHadoopCluster)
-rxLinMod(ArrDelay~CRSDepTime + DayOfWeek, data = pqData)
-```   
- 
+pqData <- RxParquetData("/share/SampleData/AirlineDemoSmallParquet",
+         colInfo = list(DayOfWeek = list(type = "factor")))
+rxGetInfo(pqData, getVarInfo = TRUE)
 
+rxLinMod(ArrDelay~CRSDepTime + DayOfWeek, data = pqData)   
 ```
-#..check on Spark data objects, cleanup, and close the Spark session
+
+
+``` 
+#..check on Spark data objects, cleanup, and close the Spark session 
+
 ls <- rxSparkListData() # two data objs are cached
 rxSparkRemoveData(ls)
 rxSparkListData() # it should show empty list