mne_denoise.dss.GaussDenoiser

class mne_denoise.dss.GaussDenoiser(a: float = 1.0)

Gaussian nonlinearity for iterative DSS.

The transform is source * exp(-a * source**2 / 2).

Parameters:

afloat, default=1.0

Gaussian scale.

denoise(source: ndarray) → ndarray

Apply the Gaussian FastICA nonlinearity.

Parameters:

sourcendarray, shape (n_times,) or (n_times, n_epochs)

Source time series.

Returns:

denoisedndarray, same shape as source

source * exp(-a * source**2 / 2).