The repo includes a run_analysis.R file, which takes the given data sets, attaches names to the variables, selects only those variables which give mean or standard deviation values (plus the ID variables which are "subject" and "activity") and creates a tidy (4-column) data set with the averages of each variable for each subject-activity combination (meaning that for instance for subject 1 each of the 79 parameter mean values are displayed for walking, and the same goes for each of the 5 other activities the subject was monitored while engaging in). A code book is also available.