run_analysis R The script run_analysis.R reconstructs the data from the Samsung galaxy expiriment. The files were loaded to a local machine; unziped and extracted to a local directory. Since these files reside in several sub-files a link was created to the Main file " UCI HAR Dataset" . From this point the appropiate path, pointing to the particular needed file, is added.
## create master data file
* load and combine X_test and X_train into one table [ a total of 561 attributes and 10,299 attributes ]
* load, combine and add to the table; names for the attributes previously loaded. ["features.txt"]
* load and combine activity and activity id. Maintain the original order of activities performed.
* y_train.txt and y_test.txt contains activity_id and order of performance.
* The activity labels were " left inner joined " in a separate column.
* name the two columns "activity_id" and "activity".
* Subject_test and subject_train.txt loaded combined.These are the subjects who were tested and trained.
* name the subject column "subject_id"
* the above procedures have created a master table consisting of 561 Columns and 10,299 rows.[ see script:"Dataset"]
* three columns are still to be set . [Subject_id, activity_id, activity]
## prepare tidy data set.
* Mean and Standard Devitation(std) are extrapolated from our "Dataset" and overriding existing ["Dataset].
* These are the attributtes required by our project.
* Names for our subjects, activity_id and activity are created and set and the three rows are now added to our table.
* tidydata now contains 89 columns 86 of which are a std or mean of features.
* tidy data is temporarily reshaped and the average for each mean or std is calculated by activity for
each test and train participant. we now have a table consisting of 180 observations and 89 attributes
( 86 of which are calculated by the script).
* descripitive data is made by editing the mean or standard devitation column names see code book .
* tidy data is now saved to local disk and also uploaded as "tidy_dataset.txt"