mne.channels.compute_interpolation_data

mne.channels.compute_interpolation_data(info, *, origin='auto', mode='accurate', verbose=None)

Precompute data to speed up bad channel interpolation.

Parameters:

infoinstance of Info

The measurement info containing the channel positions.

originarray_like, shape (3,) | str

Origin of the sphere in the head coordinate frame and in meters. Can be 'auto' (default), which means a head-digitization-based origin fit.

modestr

Either 'accurate' (default) or 'fast', determines the quality of the Legendre polynomial expansion used for MEG.

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:

interpolationdict

Maps 'eeg' and/or 'meg' (whichever channel types are present) to InterpolationData instances, for use as the interpolation argument of mne.io.Raw.interpolate_bads(), mne.Epochs.interpolate_bads(), or mne.Evoked.interpolate_bads().

Notes

The precomputed data does not depend on which channels are bad: for EEG it holds the spherical spline Legendre series evaluated between all pairs of sensors, and for MEG the lead field dot products between all pairs of sensors. It can thus be reused for any set of bad channels, e.g., when interpolating different channels in different epochs, as long as the sensor positions (and for MEG, info['dev_head_t']) are unchanged.

New in v1.14.