✨ Include labels with atlases

[shnizzedy]

Related problem

how do you know what regions the rows and columns represent in the connectivity tsvs?

They’re fully dependent on the atlases in the config. The Schaefer atlases come from https://github.com/ThomasYeoLab/CBIG/tree/v0.22.6-Edit_KRR_LITE_to_only_allow_for_corr_kernel/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal/Parcellations/MNI, so I think they’d correspond to the labels in these CSVs: https://github.com/ThomasYeoLab/CBIG/tree/v0.22.6-Edit_KRR_LITE_to_only_allow_for_corr_kernel/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal/Parcellations/MNI/Centroid_coordinates

[…]

I thought that the sidecars for the tsvs would list the row/column info for the tsvs, but it looks like they have the processing info for the bold scan

As far as I know, C-PAC never even loads the labels, it just does the correlating. I don’t know why we wouldn’t want to add that functionality though.
I think both CC atlases are from here, but I don’t see any labels for them

The people downloading RBC will definitely want to know what the regions in the connectivity matrices are

― pennbbl#rbc-collab/p1718043001881159 🧵

C-PAC, through v1.8.7 at least, just uses the integer keys for the labels in the connectivity matrices it generates.

Proposed feature

If we package labels with the atlases we package, we can include clear labels in these matrices like the examples in nilearn's documentation

### Acceptance criteria
- [ ] Labels are included with atlases (where relevant)
- [ ] C-PAC handles included labels
- [ ] C-PAC handles atlases without labels