Closed caiw closed 8 months ago
@caiw - can you give more info of what this issue refers to? In the above plot, I can see that at least 28 sensors are attributed to the centre line (i.e. are outlined dots) - which I think is too many, is this what this issue refers to?
I'm not 100% sure to be honest :) I was removing the #todo
as part of #184.
The current code looks like this:
_left_right_sensors: tuple[AxisAssignment, AxisAssignment] = (
AxisAssignment(axis_name="left",
axis_channels=[
sensor
for sensor, (x, y) in get_meg_sensor_xy().items()
if x <= 0
] + eeg_sensors()),
AxisAssignment(axis_name="right",
axis_channels=[
sensor
for sensor, (x, y) in get_meg_sensor_xy().items()
if x >= 0
] + eeg_sensors()),
)
So it looks as if the EEG sensors are all getting assigned to both hemis, which is how we are defining medial sensors at the moment:
# Symmetric difference
both_chans = top_chans & bottom_chans
top_chans -= both_chans
bottom_chans -= both_chans
So I assume the comment relates the fact that we aren't actually using the location of EEG sensors to assign them to left and right.
I'll amend the issue description accordingly.
@neukym Please add a file which maps from the sensor names in EEG1015.lay to the sensor names we use.
Done. Reassigning to @caiw.
Should be ready now
Currently EEG sensors are assigned to both hemispheres: we aren't using any positional information to assign them.
Note to the person who looks at this issue next
I have a memory that there was a reason we couldn't easily access the positional info from the EEG sensors, or that we decided not to spend the time properly parsing it, so this may not be a super straightforward fix.
Removed
#todo
Removed comment
# TODO: these EEGs aren't split appropriately
in_left_right_sensors = ...
inplot.py
as part of #173/#184