Getting and Cleaning Data Course Project

This project is delivered as the final assignment to the Getting and Cleaning Data course and aims to transform and tidy Samsung Galaxy S sensor measurement datasets for several subjects performing activities.

Refer to the CodeBook.md file for a description of the data and the transformations performed by this projects's code.

The working directory should be set to the directory containing the data as specified in the CodeBook file in a subdirectory UCI HAR Dataset.

The script to perform the requested transformations is in run_analysis.R. The file can be run as follows:

• source("run_analysis.R")
• run.analysis()

From the main script the following actions are performed:

• initialize the base path to UCI HAR Dataset
• read the activity_labels.txt file to be able to translate the activity IDs
• read the features.txt file to name the columns later on
• make a subset of the feature names containing only mean and std features
• get the merged Dataset getData()
  • get the training data getDataSet()
    • get the observation data (readAndFilterObservations)
    • get the subject data
    • get the activity data
    • combine the data using cbind() and the factor level/label lookup
  • get the test data getDataSet()
    • (see training data)
  • merge the data using rbind()
  • write result to mergedData.csv
• create the averaged and tidy Dataset averageAndMakeTidy()
  • average features for SubjectId and Activity
  • tidy the feature names (see the CodeBook.md file and R comments for description)
  • write result to tidyData.csv and tidyData.txt
• return the tidy Dataset