There is only one script that indicates how to dowload the data and how to work with them. The script produces at the end a txt file written from a dataframe "run_data_summ"; this dataframe shows means of variables (features involving means and stds) grouped by subject and activity labels. When running the code one has acces to dataframes that include all data: run_data, and also dataframe with selected data (all means and standard deviations): run_data_selected. Within those two dataframes there is a variable "Source" that indicates whether the records were part of testing or training.