Description of run_analysis.R

Purpose:

The run_analysis.R script aims at combining the train and test datasets from the Samsung dataset linked bellow.

Files:

• run_analysis.r > the script to process the raw data.
• CodeBook.md > explains the transformations the data went through and the variables
• Readme.md > this document explaing how run_analysis work
• tidy_mean_dataset.txt > the required output from the assignment containg the average of all variables grouped by activity and subject
• tidy_dataset.txt > the breakdown by activity prior to aggregating the data in tidy_mean_dataset.txt
• dataset > a folder containg the raw data

The Output of run_analysis is two datasets:

• (optional) tidy_dataset.txt : 1 tidy dataset containing the extracted from the raw dataset mean and std variables across the 30 subjects and 6 activities. This dataset contains 10299 rows and 68 variables
• (required as per assignment) tidy_mean_dataset.txt : 1 tidy dataset calculating the average of each variable. This dataset contains 180 rows (30 subject x 6 activities) and 68 variables (subject, activity and 66 measures).

Dependencies

• data.table
• dplyr

How to run the script

run_analysis.R is self contained. You need to:

1. Download run_analysis.R
2. Set the working directory to the dir containing the script
3. Run the script in R Studio or R.
4. Alternatively save run_analysis.r and then. - 4.1 run source('run_analysis.R') from the working directory - 4.2 run read.table('tidy_mean_dataset.txt', header=TRUE)

Troubleshooting

The script was created in Version 0.98.1103 of RStudio on Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_1) AppleWebKit/600.2.5 (KHTML, like Gecko) I have tried to make it as easy as possible for you but different systems behave differently. Here is a guideline on how to resolve possible issues:

1. If data is not downloading - If your system is different and you experience issues you may need to download the data manually from: https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip - Please unzip the file and rename the folder to 'dataset'
2. If installed packages fails - Please run install.packages('data.table') and install.packages('dplyr')
3. The rest of the script should work as specified in the assignment once you have the data in the dataset folder under the working directory and the installed packages

Raw Data source:

Description and link: http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones Direct download link: https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip

Steps the script run_analysis.r executes to clean the data

1. Getting the data

• 1.1 Check if dataset exists and if not
  • 1.1.1 Download the dataset
  • 1.1.2 Unzip the dataset
  • 1.1.3 Rename the dataset

2. Check if packages exists and if not download them

• required: data.table & dplyr

3. Execute the steps as outlined in the assignment:

• 3.1 Merges the training and the test sets to create one data set. - 3.1.1 load test & train subject data / x data / load y data - 3.1.2 Merge on rows test & train subject data / x data / y data - 3.1.3 Merge subject, x and y data by column. Produces a data.frame of 10299 rows x 563 varaibles. - 3.1.4 Sort the dataset to make inspection easier
• 3.2 Extracts only the measurements on the mean and standard deviation for each measurement. - 3.2.1 Change names of merged dataset to ids v1 = 1 - 3.2.2 Load features.txt - 3.2.3 Find the mean and std variables with grep. Results in extract_id which can be used directly to subset the main dataset - 3.2.4 Subset the merged dataset. Produces a data.frame of 10299 rows x 68 varaibles.
• 3.3 Uses descriptive activity names to name the activities in the data set - 3.3.1 Load activities - 3.3.2 Match the loaded activities id to subset activityId - 3.3.3 Change to descriptive names and check result

summary(subseted$activityId) laying sitting standing walking walking-downstairs walking-upstairs 1944 1777 1906 1722 1406 1544

  - 3.4 Appropriately labels the data set with descriptive variable names. 
    - 3.4.1 Create format_names which uses gsub to change abbreviations in variables to actual labels.
    **I removed the brackets and dashes and spelled out time and frequency. Converted to lowercase.**

    ``` r
format_names <- function(variable) {
        process_names <- gsub('^t', 'time', variable)
        process_names <- gsub('^f', 'frequency', process_names)
        process_names <- gsub('\\(\\)', '', process_names)
        process_names <- gsub('-', '', process_names)        
        tolower(process_names)
        }
    ```
    
        - 3.4.2 Rename ids to descriptive names, where format_names substitutes abbreviated names with descriptive names

        ``` r
names(subseted) <- c('subject', 
                     'activity', 
                     as.character(format_names( features[extract_id, 'measure'] ))
)        
        ```
  - 3.5 From the data set in step 4, creates a second, independent tidy data set with the average of each variable for each activity and each subject. 
    - 3.5.1 Use dplyr package to group the variables and calucalte the mean on each group. **Produces a data.frame of 180 rows x 68 varaibles.**

  4. There is an option to generate csvs as well for easier reference. If you desire to do so, please uncomment the following lines at the very end of run_analysis.R

    ```r
            # write.csv(subseted, 'tidy_dataset.csv')
            # write.csv(tidy_mean_dataset, 'tidy_mean_dataset.csv')
    ```