Data Science Specialization: Getting and Cleaning Data

Author: Daniel Valencia Email: dafevara[at]gmail[dot]com Date: 2015-08-22

Introduction

In this file it's described the way to execute the run_analysis.R Script to replicate the steps in order to get the tidy data.

The Script Methods Description

For this work it was only needed one script called run_analysis.R. In this file the main function is Serket.run Serket is the code_name for the project and the name of the Github Repo.

The run_analysis.R script use a DATA_DIR global variable which stores the Data Directory where the input data set is ready to be used.

From that main method other auxiliar methods are called. Following the auxiliar method list:

• LoadData: Responsible for validate and load the raw data. It returns a list of raw tables including train and test sets.
• ClipData: Responsible for put together the subject, activity and measurement raw data. Subject and Activity data is added to keep consistence and be ready for the extraction fase.
• ExtractMeanAndStd: Responsible for the Mean and Standard Deviation measurement extraction. In this step Subject and Activity data is added to keep consistency through the whole process
• ActivityLabels: Responsible to prepare the activity label before merging with Mean and Std Data.
• MergeData: Responsible for merge Mean, Std, Subject and Activity Data
• AddSelfExplainNames: Responsible for add Human readable and Self-explain Variable Names to the MergedData. This method calls an other helper method called GenerateHumanReadableName which is responsible for generate a easy to read and understand variable name using the keywords from the original variable name.
• CreateSummarized: Responsible for create the final tidy data set. This method groups and summarize (using the mean over each measures) the data based on SubjectId, ActivityLabelId and ActivityLabel Variables.

The instructions List

To execute the analysis, there's only needed to suply the DATA_DIR value which is a directory system path to the UCI HAR Dataset data.

After that, Please run the run_analysis.R script with Rscript or within a R console.

The tidy data will be print as output.

Dependencies

This analysis depends on the following packages:

• dplyr