We are plotting evokeds for a group of channels with mne.viz.plot_compare_evokeds, but the head shape shown by show_sensors shows up as a tiny dot, rather than encompassing the electrodes.
Here is a zoom in, where you can see a tiny ‘unfilled dot’ that is actually a tiny version of the head shape with ears/nose.
Here is the full image
I assume that the channel locations in your dataset are specified in mm, but MNE assumes it’s in meters. Hence the electrodes are placed so far away from the head, which appears tiny in comparison.
You can verify this by looking at the output of evoked.get_montage()
There is unfortunately no built-in way to fix this (but it can be fixed, it’s just not convenient).
A quick solution is to apply one of our built-in montages, e.g.
evoked.set_montage('easycap-M1')
(the choice of the montage depends on your electrode layout)
Hi Richard,
I just realized I had forgotten to reply to this.
We ended up scaling up our xyz coordinates for our montage so they are correctly interpreted as meters rather than millimeters.
This solved the issue.
Thanks!