Posteriordb v 0.4.0

Modeling
1

Hi everyone!

Now I and @avehtari are done with the PR review and finally, a new version of posteriordb is out. There is still a need for more complex posteriors > 100 parameters or models that take longer to estimate (minutes or hours rather than seconds). The ideal posteriors are complex posteriors that have already been published where the data is open and can be included. Feel free to reach out if you have implemented such posteriors. I’m happy to include them.

News in posteriordb 0.4.0

EDIT by Aki: added link to the github repo

4 Likes
2

Hi Måns,

Perhaps this could be an option?

GitHub - torkar/BDA-PL: Replication package for manuscript on Bayesian analysis on programming languages data (not my data)
GitHub - torkar/affective_states: Data and scripts for the paper on affective states

Send me an email if you have questions :)

1 Like
3

I could also offer models from my last paper (OSF | Prediction can be safely used as a proxy for explanation in causally consistent Bayesian generalized linear models) but they might be a tad too large. I can’t loo them with my 32gb of Ram and the brms files are >1GB

4

Sounds great! We might not have enough memory for the draws, but the stan code and data should work.