The original data set was taken from http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones . Using instructions from Coursera Project, the dataset has been modified.

• 'README.MD'
• 'ProjectCodeBook.pdf'
• 'run_analysis.R'

The output file MyTidyDataset.txt has been uploaded to Coursera Project submission site.

1. Copy run_analysis.R to working directory. Optionally copy Readme.MD and ProjectCodebook.MD to working directory.
2. Create a folder called "UCI" in the working directory
3. Unzip and Copy the input dataset provided by Coursera project instructions

The resulting file structure will be as shown below.

Working directory

--run_analysis.R

--README.MD

--ProjectCodeBook.pdf

--UCI

-------test folder and its content

-------train folder and its contents

-------activity_labels.txt

-------features.txt

-------features_info.txt

-------README.txt

Step 1:

• Read Training dataset and test dataset and cobine them with rbind

Step 2:

• Read features and remove characters such as (, ), hyphen, comma, and other duplicate data. Note: removed paranthesis with \(\), Using gsub to replace strings
• Select only features with mean or std and drop all other columns. Used of grep to accomplish finding columns

Step 3:

• Read Subject training set and test set, cbind the subjects to the main dataset
• Read activity training set and test set, combine them to one data frame and add it to the main dataset
• Read descriptive activity labels and merge it based on V1 column
• After joining column V1 becomes duplicate so drop it

Step 4:

• Change column names to be more readable

Step 5:

• Summarize data by subject and activity, used ddply to make a tidy dataset