window_rms

typhoon.test.rms.window_rms(signal, window_length=None, max_frequency=1000, match_num_of_cycles_with_samples=True, assume_previous_data=False)

Calculates RMS value of the signal using sliding window technique. User can specify “window_length” in seconds; the window length should represent the whole number of signal periods. The longer the “window_length” is, the longer will be initial transient during which the function doesn’t provide valid rms value. If “window length” is not specified, the function will try to determine window_length by using Fast Fourier Transform to find dominant frequency - “window_length” will have length of the one signal period. There is also additional option to optimize window_length calculation; if “match_num_of_cycles_with_samples” is set, window_length will be calculated to cover the smallest number of signal periods which will ensure that calculated number of samples matches exactly the whole number of signal periods. This will make rms calculation more precize, but will also increase transient time at the beginning of the rms result. If “assume_previous_data” is set to True, the function will assume signal values before the provided ones; In this way, rms result will not have transient period during the first “window_length” seconds; rms result will have expected values from beginning.

Parameters:

  • signal (Series) – A Pandas Series with the signal the RMS value is to be calculated.

  • window_length (float/int or Timedelta) – Duration in seconds (if float/int) of the window used in the calculation. Defaults to None, in which case the period of the strongest frequency found on the signal will be used for window period.

  • max_frequency (float/int or Timedelta) – If ‘window_length’ is not defined, sets the maximum frequency to automatically detect the ‘window_length’ as one period of the signal’s strongest frequency. Default: 1000

  • match_num_of_cycles_with_samples (bool) – If ‘window_length’ is not defined, tries to match the ‘window_length’ with an integer number of cycles of the strongest frequency. Default: True

  • assume_previous_data (bool) – If set, the return rms array will have valid values from beginning of function analysis. This is achieved by assuming that the signal in previous “window_length” seconds behaved exactly the same as in the first provided “window_length” seconds.

Returns:

rms – Time series with the same length as signal with the calculated RMS value.

Return type:

Series

Examples

>>> from typhoon.test.signals import assert_is_constant, pandas_sine
>>> from typhoon.test.rms import window_rms
>>> from typhoon.test.ranges import around, after
>>> from typhoon.types.timedelta import Timedelta as td
>>> from numpy import sqrt

>>> sine = pandas_sine(amplitude=220*sqrt(2), frequency=60)
>>> period = td(1/60)
>>> rms = window_rms(sine, period)
>>> assert_is_constant(rms, at_value=around(220, tol_p=0.01), strictness=1, during=after(period))