This is the README file for the run_analysis.R script and it's output activity_averages.csv For complete details on source data see the README.txt file included with the original data For details on activity_averages.csv see CodeBook.md

Create a new dataset containing the average of each variable selected from the original data set for each activity and each subject.

This script assumes you are running OS X.

• An empty working directory.
• libraries: readr & tidyr

• main.site: the main site containing the data to download
• folder: the folder inside the zip file that contains the files needed
• extract.files: the files, with subfolders if necessary, to extract from the zip file
• data.files: all the files in the working directory after downloading and unzipping the source data
• features: 561 features from features.txt
• mead.std: identifiers for features containing mean or std (standard deviation) in their name
• measurements: measurement data from files X_test.txt & X_train.txt
• subject: subject ids from files subject_test.txt & subject_train.txt
• activity.id: activity ids from files y_test.txt & y_train.txt
• combined.data: subject + activity.id + measurements
• averages: the mean of each of the 66 features selected, calculated by activity by subject

Original data: accelerometer data collected from the Samsung Galaxy S. Variables selected: all mean and standard deviation variables identified with mean() & std().

Summary of features: Prefix 't' denotes time. Prefix 'f' denotes frequency. '-XYZ' is used to denote 3-axial signals in the X, Y and Z directions. Additional features obtained by averaging the signals in a sample window.

Examples: tBodyAcc-XYZ tBodyAccJerk-XYZ tBodyGyro-XYZ fBodyAcc-XYZ fBodyAccJerk-XYZ fBodyGyro-XYZ gravityMean tBodyAccMean tBodyAccJerkMean tBodyGyroMean tBodyGyroJerkMean

It is not necessary to download the source file in advance. The script will download and unzip the data. Details: http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones Source: https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip

See README.txt with the original dataset for license details