Simulation eval_coords bug

[j-c-c]

This PR addresses the fact that Cov3d clustering indices maybe permuted from the Simulation volume indices or states, mentioned in #823

[codecov]

Codecov Report

Merging #873 (1681892) into develop (b60c0ee) will not change coverage.
The diff coverage is 100.00%.

❗ Current head 1681892 differs from pull request most recent head 15e5a18. Consider uploading reports for the commit 15e5a18 to get more accurate results

@@           Coverage Diff            @@
##           develop     #873   +/-   ##
========================================
  Coverage    88.68%   88.68%           
========================================
  Files          116      116           
  Lines         9439     9439           
========================================
  Hits          8371     8371           
  Misses        1068     1068           

Impacted Files	Coverage Δ
src/aspire/source/simulation.py	99.44% <100.00%> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[garrettwrong]

This problem is famous. For a statistic see adjusted rand index.

For similar loosely related algorithms, see both Edit Distance family, the Hungarian Problem family, and bipartite matching. Although I think the statistic I mention is the exact thing to use if we are trying to solve this for the general case.

Let me know if the statistic will cover the issue, thanks.

[garrettwrong]

Code lgtm, thanks.

Can you squash this down to one commit? While doing that, you can add the Co-authored-by: on the commit so I get that elusive Pair Extraordinaire badge 😇 .

[j-c-c]

Code lgtm, thanks.

Can you squash this down to one commit? While doing that, you can add the Co-authored-by: on the commit so I get that elusive Pair Extraordinaire badge 😇 .

Will do! Sorry forgot to add you to that!

[garrettwrong]

I think we can spare Joakim on reviewing this one if you want to merge it in (we talked about it yesterday). Ty!