mne_features.bivariate.compute_spect_corr¶

mne_features.bivariate.compute_spect_corr(sfreq, data, with_eigenvalues=True, include_diag=False, psd_method='welch', psd_params=None)¶

Correlation Coefficients (computed from the power spectrum).

Parameters

sfreqfloat

Sampling rate of the data.

datandarray, shape (n_channels, n_times)

The signals.

with_eigenvaluesbool (default: True)

If True, the function also returns the eigenvalues of the correlation matrix.

include_diagbool (default: False)

If False, features corresponding to pairs of identical electrodes are not computed. In other words, features are not computed from pairs of electrodes of the form (ch[i], ch[i]).

psd_methodstr (default: ‘welch’)

Method used for the estimation of the Power Spectral Density (PSD). Valid methods are: 'welch', 'multitaper' or 'fft'.

psd_paramsdict or None (default: None)

If not None, dict with optional parameters (welch_n_fft, welch_n_per_seg, welch_n_overlap) to be passed to mne_features.utils.power_spectrum(). If None, default parameters are used (see doc for mne_features.utils.power_spectrum()).

Returns

outputndarray, shape (n_output,)

Where n_output = n_coefs + n_channels if with_eigenvalues is True and n_output = n_coefs if with_eigenvalues is False. If include_diag is True, then n_coefs = n_channels * (n_channels + 1) // 2 and n_coefs = n_channels * (n_channels - 1) // 2 otherwise.

Notes

Alias of the feature function: spect_corr. See [1].

References

1

https://kaggle2.blob.core.windows.net/forum-message-attachments/ 134445/4803/seizure-detection.pdf