skdh.preprocessing.CountWearDetection

class skdh.preprocessing.CountWearDetection(nonwear_window_min=90, epoch_seconds=60, use_actigraph_package=False)

Detect periods of wear/non-wear from acceleromter data using an implementation similar to the ActiGraph counts metric. Consecutive periods of zero activity counts are classified as non-wear.

Parameters:

nonwear_window_minint, optional

Minutes of zero count to consider nonwear. Default is 90 [2].

epoch_secondsint, optional

Number of seconds to accumulate counts for. Default is 60 seconds.

use_actigraph_packagebool, optional

Use the internal calculation of activity counts (skdh.utility.get_activity_counts()), or the Python package published by ActiGraph.

Methods

convert_timestamps(t)	Convert a timestamp/array of timestamps to a datetime object
predict([time, accel, fs])	Detect periods of non-wear.
save_results(results, file_name)	Save the results of the processing pipeline to a csv file

See also

utility.get_activity_counts

activity count calculation

Notes

Note that the internal method for calculating activity counts will give slightly different results than the package by ActiGraph, due to handling down-sampling differently to handle different devices better.

References

[1]

C. E. Matthews et al., “Amount of Time Spent in Sedentary Behaviors in the United States, 2003–2004,” American Journal of Epidemiology, vol. 167, no. 7, pp. 875–881, Apr. 2008, doi: 10.1093/aje/kwm390.

[2]

L. Choi, Z. Liu, C. E. Matthews, and M. S. Buchowski, “Validation of Accelerometer Wear and Nonwear Time Classification Algorithm,” Medicine & Science in Sports & Exercise, vol. 43, no. 2, pp. 357–364, Feb. 2011, doi: 10.1249/MSS.0b013e3181ed61a3.

predict(time=None, accel=None, *, fs=None, **kwargs)

Detect periods of non-wear.

Parameters:

timenumpy.ndarray

(N, ) array of unix timestamps (in seconds) since 1970-01-01.

accelnumpy.ndarray

(N, 3) array of measured acceleration values in units of g.

fsfloat, optional

Sampling frequency, in Hz. If not provided, will be computed from time.

Returns:

resultsdict

Dictionary of inputs, plus the key wear which is an array (N, 2) indicating the start and stop indices of wear.