Getting and Cleaning Data Course Project

Requirements

• Unzipped UCI dataset, (a full description is available at the main Website )
• R (has only been tested with R 3.1.2)
• dplyr package

Overview

1. Put run_analysis.R and the unzipped UCI dataset folder in the same working directory
2. In R, enter the following commands

> source('run_analysis.R')
> result <- make_tidy()

• To observe the dataset of interest enter the following command

> result$d2 

Function Description

• make_tidy(dir = "UCI HAR Dataset/") assumes the directory name (dir) is unchanged. If an absolute path is provided, the function can be run from anywhere on the filesystem.
• make_tidy combines the data and test set X values (measurements), y values (activity labels), subject identifiers into a single table keeping only those X values that correspond to a mean or standard deviation. The y values are replaced with the character representations in activity_labels.txt This data frame is stored result$d1.
• make_tidy for each valid (activity label, subject id) pair computes the mean for each measurement still in result$d1. The names of these columns have been modified to be prepended with MEAN_ to indicate they correspond to means. The result of this computation is stored in result$d2.

Fields in New Dataset (result$d2)

• activity corresponds to the verbal description of the activity label
• subject corresponds to the subject responsible for producing the data in a given row
• MEAN_* corresponds to the mean of measurements from the original UCI that corresponded to means or standard deviations. (see UCI HAR Dataset/features.txt for the features containing case insensitive mean and std strings in the name)