welsh

mutis.lib.welsh(t1, d1, t2, d2, t, dt)[source]

Welsh (1999) correlation with adaptative binning.

This function implements the correlation function proposed by Welsh (1999), which allows for the computation of the correlation for -discrete- signals non-uniformly sampled in time.

Parameters

t1ndarray: Times corresponding to the first signal.
d1ndarray: Values of the first signal.
t2ndarray: Times corresponding to the second signal.
d2ndarray: Values of the second signal.
tndarray: Times on which to compute the correlation (binning).
dtndarray: Size of the bins on which to compute the correlation.

Returns

resndarray (size len(t)): Values of the correlation at the times t.

Examples

An example of raw usage would be:

>>> import numpy as np
>>> from mutis.lib.correlation import welsh_ab
>>> t1 = np.linspace(1, 10, 100); s1 = np.sin(t1)
>>> t2 = np.linspace(1, 10, 100); s2 = np.cos(t2)
>>> t = np.linspace(1, 10, 100);  dt = np.full(t.shape, 0.1)
>>> welsh(t1, d1, t2, d2, t, dt)

However, it is recommended to be used as expalined in the standard MUTIS’ workflow notebook.