smartgamer / getcleandataproject

code book

=================

The data comes from the accelerometer and gyroscope 3-axial raw signals timeAcc-XYZ and timeGyro-XYZ. 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 (timeBodyAcc-XYZ and timeGravityAcc-XYZ) 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 (timeBodyAccJerk-XYZ and timeBodyGyroJerk-XYZ). Also the magnitude of these three-dimensional signals were calculated using the Euclidean norm (timeBodyAccMag, timeGravityAccMag, timeBodyAccJerkMag, timeBodyGyroMag, timeBodyGyroJerkMag).

Finally a Fast Fourier Transform (FFT) was applied to some of these signals producing freqBodyAcc-XYZ, freqBodyAccJerk-XYZ, freqBodyGyro-XYZ, freqBodyAccJerkMag, freqBodyGyroMag, freqBodyGyroJerkMag.

These signals were used to estimate variables of the feature vector for each pattern:
'-XYZ' is used to denote 3-axial signals in the X, Y and Z directions.

• timeBodyAcc-XYZ
• timeGravityAcc-XYZ
• timeBodyAccJerk-XYZ
• timeBodyGyro-XYZ
• timeBodyGyroJerk-XYZ
• timeBodyAccMag
• timeGravityAccMag
• timeBodyAccJerkMag
• timeBodyGyroMag
• timeBodyGyroJerkMag
• freqBodyAcc-XYZ
• freqBodyAccJerk-XYZ
• freqBodyGyro-XYZ
• freqBodyAccMag
• freqBodyAccJerkMag
• freqBodyGyroMag
• freqBodyGyroJerkMag

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

• Mean: Mean value

• Std: Standard deviation

activity: 6 modes: WALKING,WALKING_UPSTAIRS,WALKING_DOWNSTAIRS,SITTING,STANDING,LAYING .

subject: number 1:30, representing each of 30 subjects tested.