Skip to content

Latest commit

 

History

History
333 lines (227 loc) · 14.1 KB

CodeBook.md

File metadata and controls

333 lines (227 loc) · 14.1 KB

Summarized dataset based on the 'Human Activity Recognition Using Smartphones Dataset' from UCI

The original dataset is fully documented at http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones

It consists of several files (organized in multiple folders) in a single ZIP archive file.
After downloading and unziping the archive one can refer to the README.txt to get a detailed explanation of how the original data were generated.

This CodeBook focuses on describing the summarized dataset that we generated from the original dataset, as per R pre-processing instructions in run_analysis.R (refer to README.md for more details about how to generate the summarized data).

The summarized dataset consists in a single file - uci_har_summarized_data.txt - in table format, with a header line and then one line per subject and activity (executed by said subject) covered by the original dataset.

The original dataset provided 518 'features' for each 'time window' (a subset of the original sensor recording) applicable to a given subject and activity. We selected a subset of those original features (whatever referred to the mean and standard deviation of a particular underlying measure across the time window), and then we summarized them across time windows (for each given subject and activity) by considering their means.

The variables provided in the summarized dataset (each column in the file) are:

  1. subject

Identifier of the subject who carried out the experiment. An integer ranging from 1 to 30.

  1. activity

Label of the activity performed by the subject. Either of:

  • WALKING
  • WALKING_UPSTAIRS
  • WALKING_DOWNSTAIRS
  • SITTING
  • STANDING
  • LAYING
  1. tBodyAcc-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body acceleration derived from the accelerometer sensor recording on X-axis.

  1. tBodyAcc-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body acceleration derived from the accelerometer sensor recording on Y-axis.

  1. tBodyAcc-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body acceleration derived from the accelerometer sensor recording on Z-axis.

  1. tBodyAcc-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body acceleration derived from the accelerometer sensor recording on X-axis.

  1. tBodyAcc-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body acceleration derived from the accelerometer sensor recording on Y-axis.

  1. tBodyAcc-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body acceleration derived from the accelerometer sensor recording on Z-axis.

  1. tGravityAcc-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the gravity acceleration derived from the accelerometer sensor recording on X-axis.

  1. tGravityAcc-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the gravity acceleration derived from the accelerometer sensor recording on Y-axis.

  1. tGravityAcc-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the gravity acceleration derived from the accelerometer sensor recording on Z-axis.

  1. tGravityAcc-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the gravity acceleration derived from the accelerometer sensor recording on X-axis.

  1. tGravityAcc-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the gravity acceleration derived from the accelerometer sensor recording on Y-axis.

  1. tGravityAcc-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the gravity acceleration derived from the accelerometer sensor recording on Z-axis.

  1. tBodyAccJerk-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on X-axis.

  1. tBodyAccJerk-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Y-axis.

  1. tBodyAccJerk-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Z-axis.

  1. tBodyAccJerk-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on X-axis.

  1. tBodyAccJerk-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Y-axis.

  1. tBodyAccJerk-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Z-axis.

  1. tBodyGyro-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body angular velocity derived from the gyroscope sensor recording on X-axis.

  1. tBodyGyro-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body angular velocity derived from the gyroscope sensor recording on Y-axis.

  1. tBodyGyro-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body angular velocity derived from the gyroscope sensor recording on Z-axis.

  1. tBodyGyro-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body angular velocity derived from the gyroscope sensor recording on X-axis.

  1. tBodyGyro-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body angular velocity derived from the gyroscope sensor recording on Y-axis.

  1. tBodyGyro-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body angular velocity derived from the gyroscope sensor recording on Z-axis.

  1. tBodyGyroJerk-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body angular velocity jerk (a.k.a. time derivative of the velocity) derived from the gyroscope sensor recording on X-axis.

  1. tBodyGyroJerk-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body angular velocity jerk (a.k.a. time derivative of the velocity) derived from the gyroscope sensor recording on Y-axis.

  1. tBodyGyroJerk-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the body angular velocity jerk (a.k.a. time derivative of the velocity) derived from the gyroscope sensor recording on Z-axis.

  1. tBodyGyroJerk-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body angular velocity jerk (a.k.a. time derivative of the velocity) derived from the gyroscope sensor recording on X-axis.

  1. tBodyGyroJerk-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body angular velocity jerk (a.k.a. time derivative of the velocity) derived from the gyroscope sensor recording on Y-axis.

  1. tBodyGyroJerk-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the body angular velocity jerk (a.k.a. time derivative of the velocity) derived from the gyroscope sensor recording on Z-axis.

  1. fBodyAcc-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration derived from the accelerometer sensor recording on X-axis.

  1. fBodyAcc-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration derived from the accelerometer sensor recording on Y-axis.

  1. fBodyAcc-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration derived from the accelerometer sensor recording on Z-axis.

  1. fBodyAcc-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration derived from the accelerometer sensor recording on X-axis.

  1. fBodyAcc-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration derived from the accelerometer sensor recording on Y-axis.

  1. fBodyAcc-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration derived from the accelerometer sensor recording on Z-axis.

  1. fBodyAccJerk-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on X-axis.

  1. fBodyAccJerk-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Y-axis.

  1. fBodyAccJerk-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Z-axis.

  1. fBodyAccJerk-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on X-axis.

  1. fBodyAccJerk-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Y-axis.

  1. fBodyAccJerk-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body acceleration jerk (a.k.a. time derivative of the acceleration) derived from the accelerometer sensor recording on Z-axis.

  1. fBodyGyro-mean-X-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body angular velocity derived from the gyroscope sensor recording on X-axis.

  1. fBodyGyro-mean-Y-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body angular velocity derived from the gyroscope sensor recording on Y-axis.

  1. fBodyGyro-mean-Z-mean

The mean across sampling time windows (for a given subject and activity) of the mean (within each time window) of the Fast Fourier Transform (FFT) of the body angular velocity derived from the gyroscope sensor recording on Z-axis.

  1. fBodyGyro-std-X-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body angular velocity derived from the gyroscope sensor recording on X-axis.

  1. fBodyGyro-std-Y-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body angular velocity derived from the gyroscope sensor recording on Y-axis.

  1. fBodyGyro-std-Z-mean

The mean across sampling time windows (for a given subject and activity) of the standard deviation (within each time window) of the Fast Fourier Transform (FFT) of the body angular velocity derived from the gyroscope sensor recording on Z-axis.