Subtracts the common average reference from the EEG data (EEGPrep version).
This is useful for having a consistent referencing scheme across recordings
(cf. [Offner1950]).
Generally, common average re-referencing is data -= mean(data, axis=0), but
both EEGLAB/eegprep and to a greater extent MNE have additional bookkeeping around
re-referencing, in the latter case due to its focus on source localization. This
will have little effect on most machine-learning use cases; nevertheless, this
operation is included here to allow users to mirror the behavior of the end-to-end
EEGPrep pipeline by means of individual operations (for example when migrating
from one to the other form) without introducing perhaps unexpected side effects
on the MNE data structure.
The operation performed is:
\[X'_{c,t} = X_{c,t} - \frac{1}{C}\sum_{c=1}^{C} X_{c,t}\]
where \(C\) is the number of channels, \(c\) indexes the channel, and
\(t\) indexes time.
References
[Offner1950]
Offner, F. F. (1950). The EEG as potential mapping: the value of the
average monopolar reference. Electroencephalography and Clinical Neurophysiology,
2(2), 213-214.
Methods
-
apply_eeg(eeg, raw)[source]
Apply the preprocessor to an EEGLAB EEG structure.
- Return type:
dict[str, Any]
