Describe the bug
When I try to get the channel-adjacency-matrix for Gradiometer from Neuromag122-Data with find_ch_adjacency
, a matrix with dimensions 306x306 is given (when read from ./mne/channels/data/neighbors/neuromag122_neighb.mat).
And when I try to compute the adjacency with mne.channels.channels._compute_ch_adjacency, I get the following result:
File "C:\Users\marti\anaconda3\envs\mne\lib\site-packages\mne\channels\channels.py", line 1426, in _compute_ch_adjacency
xy = _find_topomap_coords(info, picks, sphere=HEAD_SIZE_DEFAULT)
File "C:\Users\marti\anaconda3\envs\mne\lib\site-packages\mne\channels\layout.py", line 624, in _find_topomap_coords
pos = _auto_topomap_coords(
File "C:\Users\marti\anaconda3\envs\mne\lib\site-packages\mne\channels\layout.py", line 723, in _auto_topomap_coords
raise ValueError('The following electrodes have overlapping positions,'
ValueError: The following electrodes have overlapping positions, which causes problems during visualization:
MEG 001, MEG 002, MEG 003, [...]
Steps to reproduce
import mne
data_path = "./neuromag122_test_file-raw.fif"
info = mne.io.read_info(data_path)
# Trying the load the channel-adjacency
ch_adjacency = mne.channels.find_ch_adjacency(info, 'grad')
print(ch_adjacency[0].shape)
# Trying to bypass mne.channels.read_ch_adjacency()
mne.channels.channels._compute_ch_adjacency(info, 'grad')
Expected results
A 122x122 Matrix
Actual results
A 306x306 Matrix and a Value Error
Additional information
I uploaded an anonymized raw-file from our Neuromag122 for reproduction of the issue.
Describe the bug
When I try to get the channel-adjacency-matrix for Gradiometer from Neuromag122-Data with
find_ch_adjacency, a matrix with dimensions 306x306 is given (when read from ./mne/channels/data/neighbors/neuromag122_neighb.mat).
And when I try to compute the adjacency with
mne.channels.channels._compute_ch_adjacency, I get the following result:Steps to reproduce
Expected results
A 122x122 Matrix
Actual results
A 306x306 Matrix and a Value Error
Additional information
I uploaded an anonymized raw-file from our Neuromag122 for reproduction of the issue.