Creates a test and training dataframe selecting only the features that are concerned with averages and standard deviations. It uses dpylr and regular expressions to accomplish this.
Creates a new data.table named step5 which collapses the clean dataset into the average for all the varaibles by subject and activitity
##Be aware that the varaibles "subjects" and current_activ" are the last two variables in the table, and may not be obvious at first glance.