mne.combine_evoked(all_evoked, weights)[source]

Merge evoked data by weighted addition or subtraction.

Each Evoked in all_evoked should have the same channels and the same time instants. Subtraction can be performed by passing weights=[1, -1].


Other than cases like simple subtraction mentioned above (where all weights are -1 or 1), if you provide numeric weights instead of using 'equal' or 'nave', the resulting Evoked object’s .nave attribute (which is used to scale noise covariance when applying the inverse operator) may not be suitable for inverse imaging.

all_evokedlist of Evoked

The evoked datasets.

weightslist of float | ‘equal’ | ‘nave’

The weights to apply to the data of each evoked instance, or a string describing the weighting strategy to apply: 'nave' computes sum-to-one weights proportional to each object’s nave attribute; 'equal' weights each Evoked by 1 / len(all_evoked).


The new evoked data.


New in version 0.9.0.