The run_analysis.R script reads datasets which contain accelerometer measurements from a Samsung Galaxy S smartphone, and outputs an independent tidy data set with the average of each variable for each activity and each subject from the original data.
It requires the input datafiles:
- X_train.txt
- subject_train.txt
- y_train.txt
- features.txt
- X_test.txt
- subject_test.txt
- y_test.txt
to be in the same working directory where the script is. The files can be downloaded in an archived format from the following link: https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip
- Reads in the relevant training and testing data.
- Adds the subjects and acitivities and varibale names to the training and test measurements.
- Binds together the training and test dataset.
- Selects the variables which contiain the mean and std_dev of the measurements.
- Converts the activities to factor type and decodes the activity ids into meaningful descriptions.
- Creates a dataset with the average of each variable for each activity and each subject.
- Formatting the final dataset to make it comply with the rules required for tidy data.
- Writing the resulting dataset to a file.