This repository contains a script to produce a tidy data set summarizing

the data in the UCI Human Activity Recognition (HAR) dataset by subject

by activity. Four files are included:

README.md This file

CodeBook.md A codebook describing the data in the tidy data set

run_analysis.R The R script which generates the tidy data set and

saves to file

tidydataset.txt A file containing the tidy data set. The file

includes row headers and contains

one line for each subject and activity pair. This

can be read back into R with the command:

read.table ("tidydataset.txt")

The script groups the UCI HAR data by subject by activity (30 subjects x

6 activties each), then provides the average of the mean and standard

deviation of each data sample collected:

1) Body Acceleration and Jerk signals along each axis

2) Body Gyro (angular momentum) and Jerk signals along each axis

3) Gravity Acceleration along each axis

4) Magnitudes of each of the above signals

5) Both time and frequency domain signals are provided

For the standard deviation of each sample provided, the average across

samples provided is the simple average, not the sum of squares average.