The features selected for this database come from the accelerometer and gyroscope 3-axial raw signals. These time domain signals were captured at a constant rate of 50 Hz. Then they were filtered using a median filter and a 3rd order low pass Butterworth filter with a corner frequency of 20 Hz to remove noise. Similarly, the acceleration signal was then separated into body and gravity acceleration signals using another low pass Butterworth filter with a corner frequency of 0.3 Hz.
Subsequently, the body linear acceleration and angular velocity were derived in time to obtain Jerk signals. Also the magnitude of these three-dimensional signals were calculated using the Euclidean norm.
Finally a Fast Fourier Transform (FFT) was applied to some of these signals producing, which are prefixed with 'frequencydomain' to indicate the FFT
These signals were used to estimate variables of the feature vector for each pattern:
'x', 'y', or 'z' is used to denote 3-axial signals in the X, Y and Z directions.
'mean' and 'std' are used to denote whether which summary statistic is being reported for that variable.
Each row has the subject's numerical identifier and activity (walking, walking upstairs, walking downstairs, sitting, standing, laying).
List of fields from left to right:
subject
activity
timedomainbodyaccelerationmeanx
timedomainbodyaccelerationmeany
timedomainbodyaccelerationmeanz
timedomainbodyaccelerationstdx
timedomainbodyaccelerationstdy
timedomainbodyaccelerationstdz
timedomaingravityaccelerationmeanx
timedomaingravityaccelerationmeany
timedomaingravityaccelerationmeanz
timedomaingravityaccelerationstdx
timedomaingravityaccelerationstdy
timedomaingravityaccelerationstdz
timedomainbodyaccelerationjerkmeanx
timedomainbodyaccelerationjerkmeany
timedomainbodyaccelerationjerkmeanz
timedomainbodyaccelerationjerkstdx
timedomainbodyaccelerationjerkstdy
timedomainbodyaccelerationjerkstdz
timedomainbodygyroscopemeanx
timedomainbodygyroscopemeany
timedomainbodygyroscopemeanz
timedomainbodygyroscopestdx
timedomainbodygyroscopestdy
timedomainbodygyroscopestdz
timedomainbodygyroscopejerkmeanx
timedomainbodygyroscopejerkmeany
timedomainbodygyroscopejerkmeanz
timedomainbodygyroscopejerkstdx
timedomainbodygyroscopejerkstdy
timedomainbodygyroscopejerkstdz
timedomainbodyaccelerationmagnitudemean
timedomainbodyaccelerationmagnitudestd
timedomaingravityaccelerationmagnitudemean
timedomaingravityaccelerationmagnitudestd
timedomainbodyaccelerationjerkmagnitudemean
timedomainbodyaccelerationjerkmagnitudestd
timedomainbodygyroscopemagnitudemean
timedomainbodygyroscopemagnitudestd
timedomainbodygyroscopejerkmagnitudemean
timedomainbodygyroscopejerkmagnitudestd
frequencydomainbodyaccelerationmeanx
frequencydomainbodyaccelerationmeany
frequencydomainbodyaccelerationmeanz
frequencydomainbodyaccelerationstdx
frequencydomainbodyaccelerationstdy
frequencydomainbodyaccelerationstdz
frequencydomainbodyaccelerationjerkmeanx
frequencydomainbodyaccelerationjerkmeany
frequencydomainbodyaccelerationjerkmeanz
frequencydomainbodyaccelerationjerkstdx
frequencydomainbodyaccelerationjerkstdy
frequencydomainbodyaccelerationjerkstdz
frequencydomainbodygyroscopemeanx
frequencydomainbodygyroscopemeany
frequencydomainbodygyroscopemeanz
frequencydomainbodygyroscopestdx
frequencydomainbodygyroscopestdy
frequencydomainbodygyroscopestdz
frequencydomainbodyaccelerationmagnitudemean
frequencydomainbodyaccelerationmagnitudestd
frequencydomainbodybodyaccelerationjerkmagnitudemean
frequencydomainbodybodyaccelerationjerkmagnitudestd
frequencydomainbodybodygyroscopemagnitudemean
frequencydomainbodybodygyroscopemagnitudestd
frequencydomainbodybodygyroscopejerkmagnitudemean
frequencydomainbodybodygyroscopejerkmagnitudestd
Description of data and how they were obtained adapted from 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.
summary_file.txt has the same variables as above, but summarizes by taking the average of each column across like activities for each subject.
in 'X_ combined.txt', each subject had between 35 and 95 separate clusters of observations with the same activity. 'summary_file.txt' takes the average mean and average standard deviation across each cluster.