Skip to content

Codebook.md

Latest commit

 

History

History
36 lines (23 loc) · 1.63 KB

Codebook.md

File metadata and controls

36 lines (23 loc) · 1.63 KB

Codebook for the UCI data on Human Activity Recognition Using Smartphones

This will describe the Tidy Data set in detail along with the process used to create the data. This R code will transform the UCI Data on Human Activity Recognition Using Smartphones into a Tidy Data Set. More information on the data set can be obtained from http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones and

The Study Design and Input Data

The study design for the original data cab be found at:

http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones

The input data set used can be found at:

https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip

The Transformation Process

Step 1 - Preparation

  • Read all test and train observations and identifier files (subjects, activities, column names)
  • Add the identifiers to the observations
  • Select the columns that will be used in next steps
  • Combine subjects, activities and observations into one data frame

Step 2 - Combine all observations

  • Combine training and test data sets into one large data set
  • Merge in the Activity Label to provide a proper header name and remove the ID column

Step 3 - Regroup and calculate means

  • Change the data grouping of the observations to compute the mean for each combination of subject and activity

Step 4 - Clean up column names

  • Perform some edits on the column names to improve readability

Step 5 - Write the new data into a file

  • The new data is written into the file "tidydata.csv"