Skip to content

Latest commit

 

History

History
27 lines (21 loc) · 1.72 KB

README.md

File metadata and controls

27 lines (21 loc) · 1.72 KB
Raw

Getting and Cleaning Data Course Project.

An analysis that tidies a data set.

The files in the repo are:

  • data/UCI HAR Dataset - contains copies of the original readme and features file.
  • CodeBook.md - a thorough description of the analysis and a list of the variables.
  • tidySet.txt - a copy of the independent set output from run_analysis.R
  • README.md - this file
  • run_analysis.R - The R script that completes the data transformation

To simplify analysis, this repo contains only one R script to complete the analysis. The details are laid out in CodeBook.md; but the steps to transform the data are as follows:

  1. Import the test/training data, labels(activity codes) and subject.
  2. Import the activity and feature labels.
  3. Attach the subject and activity codes to the data sets.
  4. Merge the data sets with rbind()
  5. Attach the feature labels to the merged data set.
  6. Map the activity names to the activity code using plyr::mapvalues().
  7. Filter and keep only data columns that match subject, activity, mean() and std(). Columns that do not match these expressions are discarded.
  8. Clean the activity and feature labels with gsub and grep. Clean variable/label names are free of spaces; and special characers such as -, () and _. In addition, clean names will follow the format firstSecondThird.
  9. The data is reduced to molten form via melt(), with id variables subject and activity. All remaining variables are considered measured variables.
  10. The molten/tall data is then recast - and aggregated with mean - into its wide form via dcast().
  11. Finally this separate, aggregated data is written to a file called tidySet.txt.

Questions or comments can be directed to [email protected]