skdh.preprocessing.GetDayWindowIndices#

class skdh.preprocessing.GetDayWindowIndices(bases=None, periods=None)#

Get the indices corresponding to days.

Parameters:

bases{None, int, list-like}, optional: Base hours [0, 23] in which to start a window of time. Default is None, which will not do any windowing. Both base and period must be defined in order to window. Can use multiple, but the number of bases must match the number of periods.
periods{None, int, list-like}, optional: Periods for each window, in [1, 24]. Defines the number of hours per window. Default is None, which will do no windowing. Both period and base must be defined to window. Can use multiple but the number of periods must match the number of bases.

Methods

`convert_timestamps`(t)	Convert a timestamp/array of timestamps to a datetime object
`predict`(*, time[, fs])	Compute the indices for days.
`save_results`(results, file_name)	Save the results of the processing pipeline to a csv file
`window_days`(time, fs)	Get the indices for days based on the windowing parameters.

predict(*, time, fs=None)#

Compute the indices for days.

Parameters:

timenumpy.ndarray: (N, ) array of unix timestamps, in seconds.
fsfloat, optional: Sampling frequency in Hz. If not provided, it is calculated from the timestamps.
tz_name{None, str}, optional: Timezone name. If none provided, timestamps are assumed to be naive in local time.

Returns:

datadict: Dictionary with the key day_ends, which itself is a dictionary of (N, 2) indices corresponding to [start, stop] of days. Keys are tuples of (base, period).

window_days(time, fs)#

Get the indices for days based on the windowing parameters.

Parameters:

timenumpy.ndarray: (N, ) array of unix timestamps, in seconds.
fsfloat, optional: Sampling frequency in Hz. If not provided, it is calculated from the timestamps.

Returns: