Skip to content

Latest commit

 

History

History
108 lines (87 loc) · 7.03 KB

CodeBook.md

File metadata and controls

108 lines (87 loc) · 7.03 KB
Raw

CodeBook

Input Data

The input data, Human Activity Recognition Using Smartphones Data Set, is obtained from UCI Machine Learning repository. There are 2 data layers Raw sensory data and post-processed data

Raw sensor data

The data as captured by accelerometer and gyroscope sensors of a Samsung smartphone at a constant speed of 50Hz by 30 different individuals (subjects) aged from 19 to 48 as they perform activities of daily living (ADL).

Post processing data

The data is first cleaned and normalized and bounded within [-1,1], using the median filter and and a 3rd order low pass Butterworth filter with a corner frequency of 20 Hz.

Subsequently, the acceleration signal was then separated into body and gravity acceleration signals components(tBodyAcc-XYZ and tGravityAcc-XYZ). and the following additional variables were derived:

  • The body linear acceleration and angular velocity were derived in time to obtain Jerk signal (tBodyAccJerk-XYZ and tBodyGyroJerk-XYZ)
  • The magnitude of the above signals were calculated using the Euclidean norm (tBodyAccMag, tGravityAccMag, tBodyAccJerkMag, tBodyGyroMag, tBodyGyroJerkMag)
  • Some frequency domain variables were synthesized via the Fast Fourier Transform (FFT)(fBodyAcc-XYZ, fBodyAccJerk-XYZ, fBodyGyro-XYZ, fBodyAccJerkMag, fBodyGyroMag, fBodyGyroJerkMag)

The set of variables that were estimated from these signals are:

  • mean() Mean value
  • std() Standard deviation
  • mad() Median absolute deviation
  • max() Largest value in array
  • min() Smallest value in array
  • sma() Signal magnitude area
  • energy() Energy measure. Sum of the squares divided by the number of values.
  • iqr() Interquartile range
  • entropy() Signal entropy
  • arCoeff() Autorregresion coefficients with Burg order equal to 4
  • correlation() correlation coefficient between two signals
  • maxInds() index of the frequency component with largest magnitude
  • meanFreq() Weighted average of the frequency components to obtain a mean frequency
  • skewness() skewness of the frequency domain signal
  • kurtosis() kurtosis of the frequency domain signal
  • bandsEnergy() Energy of a frequency interval within the 64 bins of the FFT of each window.
  • angle() Angle between to vectors.

Additional vectors obtained by averaging the signals in a signal window sample. These are used on the angle() variable:

gravityMean tBodyAccMean tBodyAccJerkMean tBodyGyroMean tBodyGyroJerkMean

Stored Data structure

Description of the stored Input data can be found in the README file

Output Data - Tidy

Transformation

The transformation analysis is detailed in the README file

Tidy Data description

There are 69 columns divided into 2 column types categories and sensor derived (Derived from the observation of the sensors)

Category columns

There are 3 category columns

Name Class Values Comment
type character train, test Origin of observation
activity character standing,sitting,laying,walking,walking_downstairs,walking_upstairs
subject integer 1-30 Subject id

Sensor Derived data

There are 66 sensor derivded variables, that are ordered in the following table by their description and the data domain (Time domain, or Frequency domain). They are all of type numeric and their values are normalized and bounded within [-1,1]

Description Time domain variable Frequency domain variable
Body Acceleration tBodyAcc-mean-XYZ fBodyAcc-mean-XYZ 
                                 |  tBodyAcc-std-XYZ               |  fBodyAcc-std-XYZ

Gravity Acceleration | tGravityAcc-mean-XYZ |
| tGravityAcc-std-XYZ | Body Acceleration Jerk | tBodyAccJerk-mean-XYZ | fBodyAccJerk-mean-XYZ | tBodyAccJerk-std-XYZ | fBodyAccJerk-std-XYZ Body Angular Speed | tBodyGyro-mean-XYZ | fBodyGyro-mean-XYZ | tBodyGyro-std-XYZ | fBodyGyro-std-XYZ Body Angular Acceleration | tBodyGyroJerk-mean-XYZ |
| tBodyGyroJerk-std-XYZ |
Body Acceleration Magnitude | tBodyAcckMag-mean | fBodyAccMag-mean | tBodyAcckMag-std | fBodyAccMag-std Gravity Acceleration Magnitude | tGravityAccMag-mean |
| tGravityAccMag-std | Body Acceleration Jerk Magnitude | tBodyAccJerkMag-mean | fBodyAccJerkMag-mean | tBodyAccJerkMag-std | fBodyAccJerkMag-std Body Angular Speed Magnitude | tBodyGyroMag-mean | fBodyGyroMag-mean | tBodyGyroMag-std | fBodyGyroMag-std Body Angular Acceleration Magnitude | tBodyGyroJerkMag-mean | fBodyGyroJerkMag-mean | tBodyGyroJerkMag-std | fBodyGyroJerkMag-std

Notes:

  • -XYZ is used to denote 3-axial signals in the X, Y and Z directions, i.e variable described as tBodyAcc-mean-XYZ, is actually a description of 3 variables tBodyAcc-mean-X, tBodyAcc-mean-Y and tBodyAcc-Z.
  • The value of a variable whose name contain mean string is the average of all the mean observations for a specific subject, activity pair.
  • The value of a variable whose name contain std string is the average of standard deviation of all observations for a specific subject, activity pair.
  • There are 180 observation rows in the tidy dataset.
  • prefixes 'f' and 't' denote frequency domain and time domain repectively.

Stored data structure

The data is output to a file named tidy.txt in the current working directory. An example code to read it in the R langauge is

> View(read.table('tidy.txt', header = TRUE))