cvalenzu / anomaly_detection Goto Github PK

Using the labeling script create the labeling needed for the FFT approach.

The script does the following:

• Reads all the raw data and FFT
• Filter the files where the FFT mean is less than 1 (All the files where the motor is off)
• Then shows the plots in the time and Fourier domain, and the statistics (mean,std, etc.)
• And waits for the label (0: for offline, 1:normal and 2: anomaly)

then the file is stored in "labels.csv"