The script run_analysis.R must run in the directory that contains the train and test data sets. It reads in the two directories (xxx = test, train) the subject_xxx.txt, y_xxx.txt, and X_xxx.txt files and create two data sets: train_set and test_set. Then, it merges the sets.
Using the file features.txt, the script selects only the columns of the merged data set that correspond to a mean or a standard deviation of a signal.
The script also changes the 66-feature names according to the rules described in CodeBook.md and replaces the numeric identifier of each activity for its label.
Then, the mean of the features is calculated after aggregating all the records with the same Subject_ID and activity.
Finally, the tidy data set was saved in the tidy_dataset.txt file.