This 'sleep_processing_pipeline.predict_pipeline_acceleration()' function shows the legacy algorithm based on Cole/Kripke as the default. Right?
And look at the tutorial right after.
In 'Cut Data to Wear Period' part, the wear_detection.WearDetection() function is used, and the internal logic shows that it is based on van hees heuristic. So the plot underneath is a picture obtained by van hees?
In some papers, applied a band pass filter to 50hz acceleration sensor data and processed the signal. The filtering code is not visible in the internal logic, but I wonder if it doesn't matter because it's heuristic.
This 'get_major_wear_block(data)' function is analyzed by wearing the imu for the longest time, and the sample data is only at night. If it is worn long during the day, how can I judge it?
I have additional questions on this site. https://biopsykit.readthedocs.io/en/latest/examples/_notebooks/Sleep_IMU_Example.html
And look at the tutorial right after.
In 'Cut Data to Wear Period' part, the wear_detection.WearDetection() function is used, and the internal logic shows that it is based on van hees heuristic. So the plot underneath is a picture obtained by van hees?
In some papers, applied a band pass filter to 50hz acceleration sensor data and processed the signal. The filtering code is not visible in the internal logic, but I wonder if it doesn't matter because it's heuristic.
This 'get_major_wear_block(data)' function is analyzed by wearing the imu for the longest time, and the sample data is only at night. If it is worn long during the day, how can I judge it?
Thank you!