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

Include tail ESS if possible #468

Open
bwiernik opened this issue Oct 4, 2021 · 6 comments
Open

Include tail ESS if possible #468

bwiernik opened this issue Oct 4, 2021 · 6 comments
Labels
Enhancement 💥 Implemented features can be improved or revised

Comments

@bwiernik
Copy link
Contributor

bwiernik commented Oct 4, 2021

Currently, we only report the bulk ESS. The tail ESS is useful for diagnosising reliability of the interval quantiles and detecting poor mixing. It's available by default from brms objects and can be computed for all Stan models using posterior::ess_tail() (vs. posterior::ess_bulk()).

@bwiernik
Copy link
Contributor Author

bwiernik commented Oct 4, 2021

To use rstanarm objects currently, we can use posterior functions by extracting the $stanfit slot from the object.

@mattansb
Copy link
Member

Looking at the posterior code, looks like its:

I.05 <- x <= quantile(x, 0.05)
I.95 <- x <= quantile(x, 0.95)
min(ess(I.05), ess(I.95))

@bwiernik
Copy link
Contributor Author

The Stan packages seem to all be moving to relying on posterior for these sorts of computations, so a soft dependency for Stan models could be okay.

@bwiernik
Copy link
Contributor Author

Would be really nice if BayesFactor and INLA would also update to work with the posterior S3 generics...

@mattansb
Copy link
Member

Would be really nice if BayesFactor and INLA would also update to work with the posterior S3 generics...

Unlikely... That's what we're here for!

I still cannot wrap my head around the rvar data :/

@bwiernik
Copy link
Contributor Author

My understanding is that rvar are helpful, but we can also just work with numeric R vectors

@strengejacke strengejacke added the Enhancement 💥 Implemented features can be improved or revised label Oct 27, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Enhancement 💥 Implemented features can be improved or revised
Projects
None yet
Development

No branches or pull requests

3 participants