moabb.pipelines.features.ExtendedSSVEPSignal#

class moabb.pipelines.features.ExtendedSSVEPSignal[source]#

Prepare FilterBank SSVEP EEG signal for estimating extended covariances.

Riemannian approaches on SSVEP rely on extended covariances matrices, where the filtered signals are contenated to estimate a large covariance matrice.

FilterBank SSVEP EEG are of shape (n_trials, n_channels, n_times, n_freqs) and should be convert in (n_trials, n_channels*n_freqs, n_times) to estimate covariance matrices of (n_channels*n_freqs, n_channels*n_freqs).

fit(X, y)[source]#

No need to fit for ExtendedSSVEPSignal.

transform(X)[source]#

Transpose and reshape EEG for extended covmat estimation.