Alison MacLellan Created 2014-05-25

Human Activity Recognition Using Smartphones Dataset Version 1.0

Jorge L. Reyes-Ortiz, Davide Anguita, Alessandro Ghio, Luca Oneto. Smartlab - Non Linear Complex Systems Laboratory DITEN - Università degli Studi di Genova. Via Opera Pia 11A, I-16145, Genoa, Italy. [email protected] www.smartlab.ws

The experiments have been carried out with a group of 30 volunteers within an age bracket of 19-48 years. Each person performed six activities (WALKING, WALKING_UPSTAIRS, WALKING_DOWNSTAIRS, SITTING, STANDING, LAYING) wearing a smartphone (Samsung Galaxy S II) on the waist. Using its embedded accelerometer and gyroscope, we captured 3-axial linear acceleration and 3-axial angular velocity at a constant rate of 50Hz. The experiments have been video-recorded to label the data manually. The obtained dataset has been randomly partitioned into two sets, where 70% of the volunteers was selected for generating the training data and 30% the test data.

The sensor signals (accelerometer and gyroscope) were pre-processed by applying noise filters and then sampled in fixed-width sliding windows of 2.56 sec and 50% overlap (128 readings/window). The sensor acceleration signal, which has gravitational and body motion components, was separated using a Butterworth low-pass filter into body acceleration and gravity. The gravitational force is assumed to have only low frequency components, therefore a filter with 0.3 Hz cutoff frequency was used. From each window, a vector of features was obtained by calculating variables from the time and frequency domain. See 'features_info.txt' for more details.

• Triaxial acceleration from the accelerometer (total acceleration) and the estimated body acceleration.
• Triaxial Angular velocity from the gyroscope.
• A 561-feature vector with time and frequency domain variables.
• Its activity label.
• An identifier of the subject who carried out the experiment.

• 'README.txt'

• 'features_info.txt': Shows information about the variables used on the feature vector.

• 'features.txt': List of all features.

• 'activity_labels.txt': Links the class labels with their activity name.

• 'train/X_train.txt': Training set.

• 'train/y_train.txt': Training labels.

• 'test/X_test.txt': Test set.

• 'test/y_test.txt': Test labels.

• 'train/subject_train.txt': Each row identifies the subject who performed the activity for each window sample. Its range is from 1 to 30.

• 'train/Inertial Signals/total_acc_x_train.txt': The acceleration signal from the smartphone accelerometer X axis in standard gravity units 'g'. Every row shows a 128 element vector. The same description applies for the 'total_acc_x_train.txt' and 'total_acc_z_train.txt' files for the Y and Z axis.

• 'train/Inertial Signals/body_acc_x_train.txt': The body acceleration signal obtained by subtracting the gravity from the total acceleration.

• 'train/Inertial Signals/body_gyro_x_train.txt': The angular velocity vector measured by the gyroscope for each window sample. The units are radians/second.

• Features are normalized and bounded within [-1,1].
• Each feature vector is a row on the text file.

For more information about this dataset contact: [email protected]

Use of this dataset in publications must be acknowledged by referencing the following publication [1]

[1] Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. Human Activity Recognition on Smartphones using a Multiclass Hardware-Friendly Support Vector Machine. International Workshop of Ambient Assisted Living (IWAAL 2012). Vitoria-Gasteiz, Spain. Dec 2012

This dataset is distributed AS-IS and no responsibility implied or explicit can be addressed to the authors or their institutions for its use or misuse. Any commercial use is prohibited.

Jorge L. Reyes-Ortiz, Alessandro Ghio, Luca Oneto, Davide Anguita. November 2012.

The features selected for this database come from the accelerometer and gyroscope 3-axial raw signals tAcc-XYZ and tGyro-XYZ. These time domain signals (prefix 't' to denote time) were captured at a constant rate of 50 Hz. Then they were filtered using a median filter and a 3rd order low pass Butterworth filter with a corner frequency of 20 Hz to remove noise. Similarly, the acceleration signal was then separated into body and gravity acceleration signals (tBodyAcc-XYZ and tGravityAcc-XYZ) using another low pass Butterworth filter with a corner frequency of 0.3 Hz.

Subsequently, the body linear acceleration and angular velocity were derived in time to obtain Jerk signals (tBodyAccJerk-XYZ and tBodyGyroJerk-XYZ). Also the magnitude of these three-dimensional signals were calculated using the Euclidean norm (tBodyAccMag, tGravityAccMag, tBodyAccJerkMag, tBodyGyroMag, tBodyGyroJerkMag).

Finally a Fast Fourier Transform (FFT) was applied to some of these signals producing fBodyAcc-XYZ, fBodyAccJerk-XYZ, fBodyGyro-XYZ, fBodyAccJerkMag, fBodyGyroMag, fBodyGyroJerkMag. (Note the 'f' to indicate frequency domain signals).

'-XYZ' is used to denote 3-axial signals in the X, Y and Z directions.

• tBodyAcc-XYZ
• tGravityAcc-XYZ
• tBodyAccJerk-XYZ
• tBodyGyro-XYZ
• tBodyGyroJerk-XYZ
• tBodyAccMag
• tGravityAccMag
• tBodyAccJerkMag
• tBodyGyroMag
• tBodyGyroJerkMag
• fBodyAcc-XYZ
• fBodyAccJerk-XYZ
• fBodyGyro-XYZ
• fBodyAccMag
• fBodyAccJerkMag
• fBodyGyroMag
• fBodyGyroJerkMag

• mean(): Mean value
• std(): Standard deviation
• mad(): Median absolute deviation
• max(): Largest value in array
• min(): Smallest value in array
• sma(): Signal magnitude area
• energy(): Energy measure. Sum of the squares divided by the number of values.
• iqr(): Interquartile range
• entropy(): Signal entropy
• arCoeff(): Autorregresion coefficients with Burg order equal to 4
• correlation(): correlation coefficient between two signals
• maxInds(): index of the frequency component with largest magnitude
• meanFreq(): Weighted average of the frequency components to obtain a mean frequency
• skewness(): skewness of the frequency domain signal
• kurtosis(): kurtosis of the frequency domain signal
• bandsEnergy(): Energy of a frequency interval within the 64 bins of the FFT of each window.
• angle(): Angle between to vectors.

• gravityMean
• tBodyAccMean
• tBodyAccJerkMean
• tBodyGyroMean
• tBodyGyroJerkMean

The complete list of variables of each feature vector is available in 'features.txt'

Read in the appropriate reference files

features <- read.table("UCI HAR Dataset/features.txt") # get features

activityLabels <- read.table("UCI HAR Dataset/activity_labels.txt") # get activity labels

Process the Training (Train) Data adding column names from the features, the subject, and activity, merging it all into the table Train.

xTrain <- read.table("UCI HAR Dataset/train/X_train.txt") #7352
yTrain <- read.table("UCI HAR Dataset/train/y_train.txt") #7352 Activity
colnames(xTrain) <-features$V2 # features
subjectTrain <- read.table("UCI HAR Dataset/train/subject_train.txt")
colnames(subjectTrain)<-"subject"
colnames(yTrain)<-"activity"
Train <- cbind(xTrain, subjectTrain, yTrain)

Do the same with the Testing (Test) data.

xTest <- read.table("UCI HAR Dataset/test/X_test.txt") #2947
yTest <- read.table("UCI HAR Dataset/test/y_test.txt") #2947 Activity
colnames(xTest) <- features$V2 #features 
subjectTest <- read.table("UCI HAR Dataset/test/subject_test.txt")
colnames(subjectTest)<-"subject"
colnames(yTest)<-"activity"
Test <- cbind(xTest, subjectTest, yTest)

Merge the Test and Train into one data table called oneSet and add the appropriate column names.

oneSet <- rbind(Train, Test)
colnames(oneSet) <-colnames(Train)

We want to only use the standard deviation and mean values. I pulled out the column number and name with the std() and mean() regular expressions

meanVectorN<-grep("mean\\(\\)", features$V2) # get the column numbers of the means
meanVector<-grep("mean\\(\\)", features$V2, value=TRUE) # get the values of the means
stdVectorN<-grep("std\\(\\)", features$V2) # get the column numbers of the std deviation
stdVector<-grep("std\\(\\)", features$V2, value=TRUE) # get the columna names values of the std deviation
extraColumns<-c("subject","activity")

These vectors are then used to pull the selected columns into a new data table.

selectedColumns <- oneSet[c(meanVector,stdVector,extraColumns)]
selectedColumnsNames <-colnames(selectedColumns)

Now I expanded the column names to a more easily read value using gsub

selectedColumnsNames <- gsub("^f", "fastfouriertransform", selectedColumnsNames)
selectedColumnsNames <- gsub("^t", "total", selectedColumnsNames)
selectedColumnsNames <- gsub("Body", "body", selectedColumnsNames)
selectedColumnsNames <- gsub("Gyro", "gyroscope", selectedColumnsNames)
selectedColumnsNames <- gsub("Acc", "accelerometer", selectedColumnsNames)
selectedColumnsNames <- gsub("Jerk", "jerk", selectedColumnsNames)
selectedColumnsNames <- gsub("Mag", "mag", selectedColumnsNames)
selectedColumnsNames <- gsub("Gravity", "gravityaccelerometer", selectedColumnsNames)
selectedColumnsNames <- gsub("\\-X", "\\-x", selectedColumnsNames)
selectedColumnsNames <- gsub("\\-Y", "\\-y", selectedColumnsNames)
selectedColumnsNames <- gsub("\\-Z", "\\-z", selectedColumnsNames)
selectedColumnsNames <- gsub("\\-", "", selectedColumnsNames)
selectedColumnsNames <- gsub("std\\(\\)", "standarddeviation", selectedColumnsNames)
selectedColumnsNames <- gsub("mean\\(\\)", "mean", selectedColumnsNames)
colnames(selectedColumns) <- selectedColumnsNames

To make processing simpler, I put the data into a data.frame

dfSelectedColumns <-data.frame(selectedColumns)

Now I compute the averages for the values by subject and activity

avgs <- aggregate(dfSelectedColumns, list(dfSelectedColumns$subject, dfSelectedColumns$activity), mean, na.rm=TRUE)

I then merge the activity with the descritive value of the activity, example 1 = Walking

key <- data.frame(activityLabels)
colnames(key) <- c("activity","activityvalue")
avgs <- merge(avgs, key, by="activity")

Before writing to the file, I removed the columns added from the aggregate.

• totalbodyaccelerometermeanx
• totalbodyaccelerometermeany
• totalbodyaccelerometermeanz
• totalgravityaccelerometeraccelerometermeanx
• totalgravityaccelerometeraccelerometermeany
• totalgravityaccelerometeraccelerometermeanz
• totalbodyaccelerometerjerkmeanx
• totalbodyaccelerometerjerkmeany
• totalbodyaccelerometerjerkmeanz
• totalbodygyroscopemeanx
• totalbodygyroscopemeany
• totalbodygyroscopemeanz
• totalbodygyroscopejerkmeanx
• totalbodygyroscopejerkmeany
• totalbodygyroscopejerkmeanz
• totalbodyaccelerometermagmean
• totalgravityaccelerometeraccelerometermagmean
• totalbodyaccelerometerjerkmagmean
• totalbodygyroscopemagmean
• totalbodygyroscopejerkmagmean
• fastfouriertransformbodyaccelerometermeanx
• fastfouriertransformbodyaccelerometermeany
• fastfouriertransformbodyaccelerometermeanz
• fastfouriertransformbodyaccelerometerjerkmeanx
• fastfouriertransformbodyaccelerometerjerkmeany
• fastfouriertransformbodyaccelerometerjerkmeanz
• fastfouriertransformbodygyroscopemeanx
• fastfouriertransformbodygyroscopemeany
• fastfouriertransformbodygyroscopemeanz
• fastfouriertransformbodyaccelerometermagmean
• fastfouriertransformbodybodyaccelerometerjerkmagmean
• fastfouriertransformbodybodygyroscopemagmean
• fastfouriertransformbodybodygyroscopejerkmagmean
• totalbodyaccelerometerstandarddeviationx
• totalbodyaccelerometerstandarddeviationy
• totalbodyaccelerometerstandarddeviationz
• totalgravityaccelerometeraccelerometerstandarddeviationx
• totalgravityaccelerometeraccelerometerstandarddeviationy
• totalgravityaccelerometeraccelerometerstandarddeviationz
• totalbodyaccelerometerjerkstandarddeviationx
• totalbodyaccelerometerjerkstandarddeviationy
• totalbodyaccelerometerjerkstandarddeviationz
• totalbodygyroscopestandarddeviationx
• totalbodygyroscopestandarddeviationy
• totalbodygyroscopestandarddeviationz
• totalbodygyroscopejerkstandarddeviationx
• totalbodygyroscopejerkstandarddeviationy
• totalbodygyroscopejerkstandarddeviationz
• totalbodyaccelerometermagstandarddeviation
• totalgravityaccelerometeraccelerometermagstandarddeviation
• totalbodyaccelerometerjerkmagstandarddeviation
• totalbodygyroscopemagstandarddeviation
• totalbodygyroscopejerkmagstandarddeviation
• fastfouriertransformbodyaccelerometerstandarddeviationx
• fastfouriertransformbodyaccelerometerstandarddeviationy
• fastfouriertransformbodyaccelerometerstandarddeviationz
• fastfouriertransformbodyaccelerometerjerkstandarddeviationx
• fastfouriertransformbodyaccelerometerjerkstandarddeviationy
• fastfouriertransformbodyaccelerometerjerkstandarddeviationz
• fastfouriertransformbodygyroscopestandarddeviationx
• fastfouriertransformbodygyroscopestandarddeviationy
• fastfouriertransformbodygyroscopestandarddeviationz
• fastfouriertransformbodyaccelerometermagstandarddeviation
• fastfouriertransformbodybodyaccelerometerjerkmagstandarddeviation
• fastfouriertransformbodybodygyroscopemagstandarddeviation
• fastfouriertransformbodybodygyroscopejerkmagstandarddeviation

The final two columns is the subject and the activity