braindecode.preprocessing.RegressArtifact#

class braindecode.preprocessing.RegressArtifact(picks=None, *, exclude='bads', picks_artifact='eog', betas=None, proj=True, copy=True, verbose=None)[source]#

Braindecode preprocessor wrapper for regress_artifact().

Remove artifacts using regression based on reference channels.

Parameters:

instinstance of Epochs | Raw: The instance to process.
picksstr | array-like | slice | None: Channels to include. Slices and lists of integers will be interpreted as channel indices. In lists, channel type strings (e.g., ['meg', 'eeg']) will pick channels of those types, channel name strings (e.g., ['MEG0111', 'MEG2623'] will pick the given channels. Can also be the string values 'all' to pick all channels, or 'data' to pick data channels. None (default) will pick good data channels. Note that channels in info['bads'] will be included if their names or indices are explicitly provided.
excludelist | ‘bads’: List of channels to exclude from the regression, only used when picking based on types (e.g., exclude=”bads” when picks=”meg”). Specify 'bads' (the default) to exclude all channels marked as bad.

Added in version 1.2.
picks_artifactarray-like | str: Channel picks to use as predictor/explanatory variables capturing the artifact of interest (default is “eog”).
betasndarray, shape (n_picks, n_picks_ref) | None: The regression coefficients to use. If None (default), they will be estimated from the data.
projbool: Whether to automatically apply SSP projection vectors before performing the regression. Default is True.
copybool: If True (default), copy the instance before modifying it.
verbosebool | str | int | None: Control verbosity of the logging output. If None, use the default verbosity level. See the logging documentation and mne.verbose() for details. Should only be passed as a keyword argument.

Returns:

instinstance of Epochs | Raw: The processed data.
betasndarray, shape (n_picks, n_picks_ref): The betas used during regression.

Notes

To implement the method outlined in [1], remove the evoked response from epochs before estimating the regression coefficients, then apply those regression coefficients to the original data in two calls like (here for a single-condition epochs only):

>>> epochs_no_ave = epochs.copy().subtract_evoked()
>>> _, betas = mne.preprocessing.regress(epochs_no_ave)
>>> epochs_clean, _ = mne.preprocessing.regress(epochs, betas=betas)

Methods

fn(picks=None, *, exclude='bads', picks_artifact='eog', betas=None, proj=True, copy=True, verbose=None)[source]#