coffeine.power_features.compute_features#
- coffeine.power_features.compute_features(inst: ~mne.epochs.BaseEpochs | ~mne.io.base.BaseRaw, features: tuple[str] | list[str] = ('psds', 'covs'), duration: float = 60.0, shift: float = 10.0, n_fft: int = 512, n_overlap: int = 256, fs: float = 63.0, fmin: float = 0.0, fmax: float = 30.0, frequency_bands: dict[str, tuple[float, float]] | None = None, clean_func: callable = <function <lambda>>, cov_method: str = 'oas') tuple[dict, dict]#
Compute features from raw data or clean epochs.
- Parameters:
- instRaw object | Epochs object
An instance of Raw or Epochs.
- featuresstr | list of str
The features to be computed. It can be ‘psds’, ‘covs’, ‘cross_frequency_covs’, ‘cross_frequency_corrs’ or ‘cospectral_covs’. If nothing is provided, defaults to (‘psds’, ‘covs’).
- durationfloat
The length of the epochs. If nothing is provided, defaults to 60.
- shiftfloat
The duration to separate events by (sliding shift of the epochs). If nothing is provided, defaults to 10.
- n_fftint
The length of FFT used for computing power spectral density (PSD) using Welch’s method and the cospectral covariance. If nothing is provided, defaults to 512.
- n_overlapint
The number of points of overlap between segments for PSD computation and for the estimation of cospectral covariance matrix. If nothing is provided, defaults to 256.
- fsfloat
The sampling frequency of the signal for the estimation of cospectral covariance matrix. If nothing is provided, defaults to 63.0.
- fminint
The minimal frequency to be returned for the estimation of cospectral covariance matrix and for PSD computation. If nothing is provided, defaults to 0.
- fmaxint
The maximal frequency to be returned for the estimation of cospectral covariance matrix and for PSD computation. If nothing is provided, defaults to 30.
- frequency_bandsdict
The frequency bands with which inst is filtered. If nothing is provided, defaults to {‘alpha’: (8.0, 12.0)}.
- clean_funclambda function
If nothing is provided, defaults to lambda x: x.
- cov_methodstr (default ‘oas’)
The covariance estimator to be used. Ignored for feature types not not related to covariances. Must be a method accepted by MNE’s covariance functions.
- Returns:
- computed_featuresdict
The features extracted.
- resdict
The number of epochs, good epochs, clean epochs and frequencies.