The R script does the following steps
Download the dataset from the web Loads the activity and feature info after unzipping the data Loads both the training and test datasets, keeping only those columns which can provide a mean or standard deviation Loads the activity and subject data for each dataset, and merge those columns with the dataset Merge the two datasets and convert the activity and subject columns into factors Create a tidy dataset that consists of the average (mean) value of each variable for each subject and activity pair. The end result is shown in the file tidy.txt.