Cluster test adjacency matrix

External Email - Use Caution

Dear MNE Community!

In my research I intend to perform the 1-sample cluster permutation test on
the time course data extracted with PCA according to Desikan-Killiany
cortical atlas.
In order to perform the cluster test correctly I should provide the
information on spatial connectivity of the 68 extracted Desikan-Killiany
cortical labels (i.e., the adjacency matrix).
However, while there are well defined functions such as
mne.find_ch_connectivity or mne.spatial_src_connectivity, I had no luck
figuring out any method to calculate the adjacency matrix for labels.
Do you have any ideas on how I should treat the adjacency matrix problem in
the cluster test if I use parcellised cortical labels?

Great thank you in advance!


External Email - Use Caution

We don't currently have a function for this but we could add one fairly
easily I think based on the FreeSurfer triangulations. Could you open a
GitHub issue about it?


External Email - Use Caution

Hi Anna and Eric,

Correct me if there is already an issue opened in github since I haven't
found one to follow. I would like to use a similar test in the future.
Shall I make a feature request?
Let me know. Thanks!