Normalization of Wearable Device Data

This work normalizes raw available from experimentation with smartphones during several activities and published in the Machine Learning Repository provided by the University of California, Irvine.

In the author's words, these "experiments have been carried out with a group of 30 volunteers within an age bracket of 19-48 years. Each person performed six activities (WALKING, WALKING_UPSTAIRS, WALKING_DOWNSTAIRS, SITTING, STANDING, LAYING) wearing a smartphone (Samsung Galaxy S II) on the waist. Using its embedded accelerometer and gyroscope, we captured 3-axial linear acceleration and 3-axial angular velocity at a constant rate of 50Hz. The experiments have been video-recorded to label the data manually. The obtained dataset has been randomly partitioned into two sets, where 70% of the volunteers was selected for generating the training data and 30% the test data."[1]

How Does the Script Work

Pre-requisites

• Packages dplyr and tidyr must be installed
• R or RStudio on your OS of choice

Just execute Rscript run_analysis.R from a command line, or source('run_analysis.R') from R or RStudio console - and the script will go through all stages, from downloading raw data, to normalization, to generation of the final tidy data file. The specific steps are outlined in details below:

1. Downloading bulk zip file with all data: If the file cannot be found on the local directory, it is downloaded from UCI and saved locally

2. Unzipping data: If the local directory is not found, the zipped file is unzipped on the default location

3. Load phase-independent data

   1. Labels for each activity performed (WALKING, WALKING_UPSTAIRS, WALKING_DOWNSTAIRS, SITTING, STANDING, LAYING). Labels are needed for later enrichement of the tidy data set. Activity labels are collected from activity_labels.txt
   2. Labels for each of the features, represented on the feature vector in each of the experiments, loaded from features.txt
   3. Filter features relevant to this exercise, only features related to mean and standard deviation.

4. Load phase-dependent data: We decided to name each of the sections in the raw data as a phase. For each of the phases, training and test, an observation is related to an experiment. The handling of observations in each of the phases go through the following steps:

   1. Load the identifier of the subject in each experiment, provided in <phase>/subject_<phase>.txt
   2. Load the activity on each experiment, provided in <phase>/y_<phase>.txt
   3. Load the vector feature for each experiment, provided in <phase>/X_<phase>.txt
   4. Join phase-dependent data to phase-independent data: enrich experiment with activity labels and subset the vector feature to the ones we care about (mean and standard deviation)

5. Finally, builds the tidy data set:

   1. Gathers all colums tBodyAcc_mean_X:fBodyBodyGyroJerkMag_std into one temporary column, variable
   2. Select the columns we care about for the tidy data set: subject, activity, variable, value
   3. Group on variable, activity, subject and summarize by the mean
   4. Saves to 'tidy_data.txt'

Code Book

The final data set provided after all these steps is incredibly simple and can be sinthetized by just four columns:

• variable: One of the measurement variables, valued from tBodyAcc_mean_X:fBodyBodyGyroJerkMag_std
• activity: The label of the activity, one of WALKING, WALKING_UPSTAIRS, WALKING_DOWNSTAIRS, SITTING, STANDING, LAYING
• subject: An identifier of the subject in play during those phases and experiments
• mean: The mean of the variable across all phases and experiments. Unit is the same of the specific variable in the raw data.

The resut of overall computation is saved and made available in the final result file tidy_data.txt, under the current directory.

References

[1] Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. Human Activity Recognition on Smartphones using a Multiclass Hardware-Friendly Support Vector Machine. International Workshop of Ambient Assisted Living (IWAAL 2012). Vitoria-Gasteiz, Spain. Dec 2012