pyhctsa.operations.nonlinearity.delay_time¶

pyhctsa.operations.nonlinearity.delay_time(y, max_delay=0.2, past=1, random_seed=0)¶

Optimal delay time using the method of Parlitz and Wichard.

Parameters:¶

y : array-like¶

Input time series.

max_delay : int or float, optional¶

Maximum value of the delay to consider. Values in (0, 1) are interpreted as a proportion of the time-series length. Delays below 10 are raised to 10. Default is 0.2.

past : int, optional¶

Number of time-correlated points to discard (samples) when searching for value-neighbours, i.e., the Theiler window. Default is 1.

random_seed : int or None, optional¶

Seed for the Mersenne Twister used to draw the reference points.

Returns:¶

The first three values of tau, the differences between them, and the mean, standard deviation, minimum and maximum of tau. Returns NaN if max_delay is too long for the given time series, or if no reference point clears the Theiler window.

Return type:¶

dict