Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Roll our own summary #6

Closed
rlouf opened this issue Jan 3, 2023 · 3 comments
Closed

Roll our own summary #6

rlouf opened this issue Jan 3, 2023 · 3 comments

Comments

@rlouf
Copy link
Member

rlouf commented Jan 3, 2023

Numpyro has a nice numpyro.diagnostics.print_summary function to print diagnostics for the chains. However:

  1. It feels wasteful to import numpyro just for this functionality;
  2. Blackjax implements the same diagnostics
  3. Blackjax may have more diagnostics in the future, see Add (Kernelized) Stein discrepancy diagnostic blackjax#384 and Add nested Rhat diagnostic blackjax#278

I think we should thus implement our own function that prints a summary with different diagnostics; this will certainly be useful when comparing the results from different algorithms.

@junpenglao
Copy link
Member

I think we should use https://python.arviz.org/en/stable/ for this.

Note that we have rhat and ess currently for the reason that we might want to explore using these for tuning. So I will apply the same logic when consider adding new diagnostics)

@albcab
Copy link
Member

albcab commented Apr 13, 2023

Agree with @junpenglao that we don't need to be a diagnostics library, just use arviz.

But maybe we can use some of these diagnostics on testing?

@rlouf
Copy link
Member Author

rlouf commented Apr 16, 2023

I agree with both of you, closing for now.

@rlouf rlouf closed this as not planned Won't fix, can't repro, duplicate, stale Apr 16, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants