Course Project - Getting and Cleaning Data

This repository contains the files relevant to the course project in Johns Hopkins' Getting and Cleaning Data course in Coursera.

Files

There are four files in the repository namely:

• README.md
• run_analysis.R
• Codebook.md
• tidy-HAR-data.txt

README.md

This is the README.md file. It contains descriptions of the other files and how the HAR data was processed.

run_analysis.R

The run_analysis.R file is the script that contains the code specified in the project. For the script to work, the following files from the UCI HAR Dataset must be in the same folder as the run_analysis.R script:

• subject_test.txt (from the UCI HAR Datatset/test folder)
• subject_train.txt (from the UCI HAR Dataset/train folder)
• y_test.txt (from the UCI HAR Dataset/test folder)
• y_train.txt (from the UCI HAR Dataset/train folder)
• X_test.txt (from the UCI HAR Dataset/test folder)
• X_train.txt (from the UCI HAR Dataset/train folder)
• features.txt (from the UCI HAR Dataset folder)

Processing done by run_analysis.R

1. Merging the "train" and "test" datasets
   The script first reads the subject_test.txt file into a data frame using read.table. Contents of the subject_train.txt file is added to this data frame using rbind. The same procedure is done to create dataframes for y_test.txt and y_train.txt files as well as the X_test.txt and X_train.txt files. This results to three data frames containing the merged train and test datasets for subject ("subject"), y ("activity"), and X ("feature").

2. Column names for the 561-feature vector
   The features.txt is read into a data frame to be used as column names for the data frame containing the 561-feature vector (i.e. X or the "feature" data frame). Parentheses are removed and the data frame is applied as column name to the "feature" data frame created in step 1.

3. Extracting features with mean() and std()
   A vector named toExtract is created. This vector contains the names of the features in the 561-feature vector that has the "mean()" or "std()" in their names. This will be used in the subsequent step to extract only the variables specified in the assignment.

4. Combining dataframes and extracting only the needed data
   The "subject", "activity", and "feature" data frames created in step 1 is merged into one dataframe. It should be noted that not all columns of the "feature" data frame is included in the merged data frame. Only the columns specified in the toExtract vector created in step 3 are included.

5. Labeling activities
   The data frame created from step 4 ("harData") is processed by changing the numeric labels used for activities into the respective character representation. That is activity "1" is renamed to "walking", activity "2" is renamed to "walking upstairs" and so on.

6. Grouping of data
   As the assignment calls for data to be analyzed for each activity and each subject, the group_by is called on the "harData" data frame using subject and activity as categories for grouping. The result is stored in a new data frame ("by_sa").

7. Summarize
   The function summarize_each is called on the "by_sa" data frame. In the options for summarize_each, it is specified that the mean for each column be computed. This will give us results the results specified in the assignment. That is, the "mean of means" or "mean of SDs" are computed for each activity and each subject. The result of this summarization is stored in another dataframe ("tidy").

8. Write out results to file Appropriate column names are assigned to the "tidy" data frame. This data frame is then saved to a file (named "tidy-HAR-Data.txt") using write.table. The file will be in the same folder as the script.

9. Clean up The script then cleans up unsused data frames in memory. It does leave a the data frame named "tidy" that contains the data written out to the "tidy-HAR-Data.txt" file.

tidy-HAR-Data.txt

The tidy-HAR-Data.txt is a sample output of the run_analysis.R script. To read the data into R, use read.table with the header option set to true.

tidyData <- read.table("./path/to/tidy-HAR-Data.txt", header = TRUE)

Codebook.md

The Codebook.md files contains descriptions of the variables in the resulting "tidy-HAR-Data.txt" file.