Created: 2014-05-19 02:37
Updated: 2014-07-20 21:13

Getting and Cleaning Data: Project Assignment


Pre-requisites to running the program:

The run-analysis.R script in this repo requires the user to first download and unzip the "Human Activity Recognition Using Smartphones Dataset" from UCI. This dataset can be found at:

Make sure to set your R working directory to the top folder ("./UCI HAR Dataset/"") that your files are unzipped into.

About the data

A detailed description of the study design can be found in the README.txt file in the unzipped data. Briefly, 30 subjects each performed a series of 6 activities (standing, sitting, laying down, walking, walking downstairs and upstiars) while wearing a Samsung Galaxy II smartphone. The subject data was broken in to two groups, a training set of 21 subjects, and a test set of 9 subjects. The resulting data fell into 561 different 'features' as outlined in the features.txt in the dataset and includes some mean and standard deviation calculations.

About the script

The run_analysis.R script does the following:

*Merges the training and the test sets to create one data set.

*Extracts only the measurements on the mean and standard deviation for each measurement.

*Uses descriptive activity names to name the activities in the data set

*Appropriately labels the data set with descriptive variable names.

*Creates a second, independent tidy data set with the average of each variable for each activity and each subject. NOTE: for this last step, I have chosen to make a "tall" data set, that is 180 columns representing means for each variable (those involving mean or standard devitaion) for each of 30 subjects times 6 activities. This data set had 66 feature variables plus 2 variables for SubjectID and Activity (total of 68 variables).

The script itself is fully annotated with comments, but a couple of things are worth noting:

The variables chosen that measure mean and standard deviation excludes measure of meanFreq. This was done because the meanFreq variables had no corresponding measures of standard deviation. I felt leaving these in would muddy the content of the data set.

The naming of the variables was altered from the original features.txt. This was done to remove characters that may cause problems such as "-" and "()", as well as to make the variable names more readable by capitalizing the first letter of each abbreviation in the name (known as "CamelCase"" or "medial capitals""). So, for instance, the variable name tBodyAcc-mean()-X is changed to tBodyAccMeanX. There were also a couple of variables that were incorrectly coded with the prefix "fBodyBody..." instead of just "fBody...". These are also corrected in the script.

More details about the data, variables and transformations used can be found in the in this repo.

Cookies help us deliver our services. By using our services, you agree to our use of cookies Learn more