CodeBook.md

Code book for Coursera Getting and Cleaning Data course project

Please see the README.md for more information.

This file contains following section: Data which describe the data, Variables which list the varibles and Transformations which explain the transformations on the data.

Data

The tidy_data.txt data file contain space-separated values.

The first row contains the names of the Variables and the following rows contain the values of these variables.

Variables

Each row contains, for a given subject and activity, 79 averaged signal measurements.

Identifiers

• subject

  Subject identifier, integer, ranges from 1 to 30.

• activity

  Activity identifier, string with 6 possible values:

  • WALKING: subject was walking
  • WALKING_UPSTAIRS: subject was walking upstairs
  • WALKING_DOWNSTAIRS: subject was walking downstairs
  • SITTING: subject was sitting
  • STANDING: subject was standing
  • LAYING: subject was laying

Average of measurements

The measurements are classified in two domains:

• Time-domain signals (variables prefixed by timeDomain)

• Frequency-domain signals (variables prefixed by frequencyDomain)

Time-domain signals

- `timeDomainBodyAccelerometerMeanX`
- `timeDomainBodyAccelerometerMeanY`
- `timeDomainBodyAccelerometerMeanZ`
- `timeDomainBodyAccelerometerStandardDeviationX`
- `timeDomainBodyAccelerometerStandardDeviationY`
- `timeDomainBodyAccelerometerStandardDeviationZ`
- `timeDomainGravityAccelerometerMeanX`
- `timeDomainGravityAccelerometerMeanY`
- `timeDomainGravityAccelerometerMeanZ`
- `timeDomainGravityAccelerometerStandardDeviationX`
- `timeDomainGravityAccelerometerStandardDeviationY`
- `timeDomainGravityAccelerometerStandardDeviationZ`
- `timeDomainBodyAccelerometerJerkMeanX`
- `timeDomainBodyAccelerometerJerkMeanY`
- `timeDomainBodyAccelerometerJerkMeanZ`
- `timeDomainBodyAccelerometerJerkStandardDeviationX`
- `timeDomainBodyAccelerometerJerkStandardDeviationY`
- `timeDomainBodyAccelerometerJerkStandardDeviationZ`
- `timeDomainBodyGyroscopeMeanX`
- `timeDomainBodyGyroscopeMeanY`
- `timeDomainBodyGyroscopeMeanZ`
- `timeDomainBodyGyroscopeStandardDeviationX`
- `timeDomainBodyGyroscopeStandardDeviationY`
- `timeDomainBodyGyroscopeStandardDeviationZ`
- `timeDomainBodyGyroscopeJerkMeanX`
- `timeDomainBodyGyroscopeJerkMeanY`
- `timeDomainBodyGyroscopeJerkMeanZ`
- `timeDomainBodyGyroscopeJerkStandardDeviationX`
- `timeDomainBodyGyroscopeJerkStandardDeviationY`
- `timeDomainBodyGyroscopeJerkStandardDeviationZ`
- `timeDomainBodyAccelerometerMagnitudeMean`
- `timeDomainBodyAccelerometerMagnitudeStandardDeviation`
- `timeDomainGravityAccelerometerMagnitudeMean`
- `timeDomainGravityAccelerometerMagnitudeStandardDeviation`
- `timeDomainBodyAccelerometerJerkMagnitudeMean`
- `timeDomainBodyAccelerometerJerkMagnitudeStandardDeviation`
- `timeDomainBodyGyroscopeMagnitudeMean`
- `timeDomainBodyGyroscopeMagnitudeStandardDeviation`
- `timeDomainBodyGyroscopeJerkMagnitudeMean`
- `timeDomainBodyGyroscopeJerkMagnitudeStandardDeviation`

Frequency-domain signals

- `frequencyDomainBodyAccelerometerMeanX`
- `frequencyDomainBodyAccelerometerMeanY`
- `frequencyDomainBodyAccelerometerMeanZ`
- `frequencyDomainBodyAccelerometerStandardDeviationX`
- `frequencyDomainBodyAccelerometerStandardDeviationY`
- `frequencyDomainBodyAccelerometerStandardDeviationZ`
- `frequencyDomainBodyAccelerometerMeanFrequencyX`
- `frequencyDomainBodyAccelerometerMeanFrequencyY`
- `frequencyDomainBodyAccelerometerMeanFrequencyZ`
- `frequencyDomainBodyAccelerometerJerkMeanX`
- `frequencyDomainBodyAccelerometerJerkMeanY`
- `frequencyDomainBodyAccelerometerJerkMeanZ`
- `frequencyDomainBodyAccelerometerJerkStandardDeviationX`
- `frequencyDomainBodyAccelerometerJerkStandardDeviationY`
- `frequencyDomainBodyAccelerometerJerkStandardDeviationZ`
- `frequencyDomainBodyAccelerometerJerkMeanFrequencyX`
- `frequencyDomainBodyAccelerometerJerkMeanFrequencyY`
- `frequencyDomainBodyAccelerometerJerkMeanFrequencyZ`
- `frequencyDomainBodyGyroscopeMeanX`
- `frequencyDomainBodyGyroscopeMeanY`
- `frequencyDomainBodyGyroscopeMeanZ`
- `frequencyDomainBodyGyroscopeStandardDeviationX`
- `frequencyDomainBodyGyroscopeStandardDeviationY`
- `frequencyDomainBodyGyroscopeStandardDeviationZ`
- `frequencyDomainBodyGyroscopeMeanFrequencyX`
- `frequencyDomainBodyGyroscopeMeanFrequencyY`
- `frequencyDomainBodyGyroscopeMeanFrequencyZ`
- `frequencyDomainBodyAccelerometerMagnitudeMean`
- `frequencyDomainBodyAccelerometerMagnitudeStandardDeviation`
- `frequencyDomainBodyAccelerometerMagnitudeMeanFrequency`
- `frequencyDomainBodyAccelerometerJerkMagnitudeMean`
- `frequencyDomainBodyAccelerometerJerkMagnitudeStandardDeviation`
- `frequencyDomainBodyAccelerometerJerkMagnitudeMeanFrequency`
- `frequencyDomainBodyGyroscopeMagnitudeMean`
- `frequencyDomainBodyGyroscopeMagnitudeStandardDeviation`
- `frequencyDomainBodyGyroscopeMagnitudeMeanFrequency`
- `frequencyDomainBodyGyroscopeJerkMagnitudeMean`
- `frequencyDomainBodyGyroscopeJerkMagnitudeStandardDeviation`
- `frequencyDomainBodyGyroscopeJerkMagnitudeMeanFrequency`

Transformations

The following transformations were applied to the source data:

1. The training and test sets were merged to create one data set.
2. The measurements on the mean and standard deviation (i.e. signals containing the strings mean and std) were extracted for each measurement, and the others were discarded.
3. The activity identifiers (originally coded as integers between 1 and 6) were replaced with descriptive activity names (see Identifiers section).
4. The variable names were replaced with descriptive variable names (e.g. tBodyAcc-mean()-X was expanded to timeDomainBodyAccelerometerMeanX), using the following set of rules:
   • Special characters (i.e. (, ), and -) were removed
   • The initial f and t were expanded to frequencyDomain and timeDomain respectively.
   • Acc, Gyro, Mag, Freq, mean, and std were replaced with Accelerometer, Gyroscope, Magnitude, Frequency, Mean, and StandardDeviation respectively.
   • Replaced (supposedly incorrect as per source's features_info.txt file) BodyBody with Body.
5. From the data set in step 4, the final data set was created with the average of each variable for each activity and each subject.

The collection of the source data and the transformations listed above were implemented by the run_analysis.R R script (see README.md file for usage instructions).