Residual from apply_inverse() for EEG not correct?

  • MNE version: e.g. 1.9.0
  • operating system: Windows 11 Enterprise

I modified the tutorial Source localization with MNE, dSPM, sLORETA, and eLORETA — MNE 1.9.0 documentation to include EEG in addition to MEG. See code changes below.
The plot from cell [12] suggests that the residual for EEG is the same as the evoked response. For mags/grads the residual is clearly much smaller, as it should be.
I uploaded the picture with the output of cell [12].

I changed the following in Source localization with MNE, dSPM, sLORETA, and eLORETA — MNE 1.9.0 documentation to include EEG:

cell [2]: picks=(“meg”, “eeg”, “eog”),
cell [5]: evoked = epochs.average()
cell [12] : fig, axes = plt.subplots(3, 1)