Dan Schult

This looks quite like the community detection functions in that it tries to find a partition that optimizes some objective function. Your objective function counts the number of within-group edges....

I think all of those methods are approximate or iterative in nature. It's too hard a problem to solve exactly for any reasonably sized problem of interest.

It seems that you need "triage" permissions to add a label. I have done that here. Maybe we should rename that test to indicate a `core developer` needs to add...

Yes, upon looking at this again, I still think this problem is community detection problem -- or rather that this is where it fits best in NetworkX. Those methods are...

That sounds good -- and the name using "community" might fit better. You should probably include a comment in the doc_string that mentions the "coalition" name for anyone coming from...

Hi Sorry for the delay. I have tried to think through which is better and on such short timescale I'm not sure I know which is better. What have you...

I was most of the way to fixing the issues with dispatchable when I ran into configuration problems and haven’t been able to get back to it. I’m aiming for...

I've merged the PR with `main` and reapplied your PR to change the name of dispatch to dispatchable. It looks like it is working -- hope for some more green...

Yes! Could you push those changes up now... That will likely reduce the wait for review of that later -- and not add much to the review time for this...

The geospatial examples are failing -- definitely not related to this PR. And other PRs started having this trouble about the same time. Something changed in the CI's testing environment....