search

paul-buerkner / brms

brms R package for Bayesian generalized multivariate non-linear multilevel models using Stan
https://paul-buerkner.github.io/brms/
GNU General Public License v2.0
1.28k stars 181 forks source link

file_refit = "on_change" after changing priors #1148

Closed clementbbp closed 3 years ago

clementbbp commented 3 years ago

Hi,

I'm a relatively new user to brms and bayesian statistics, so please excuse if my request is already implemented in the package somewhere, or is unfeasible.

Would it make sense to implement a feature, perhaps similar to file_refit = "on_change", which updates the model not only after changing the formula, but also after calibrating the priors? In my short experience, you may find the priors too wide or constraining on the posterior and want to calibrate them, but you don't have the benefit of being able to save the model immediately using file = " ", as this would simply load the old model. I think it would be a nice addition to the workflow, but I might be wrong since I can also imagine that you want to think very hard about choosing your priors before finally "saving" the model.

Any answer would be greatly appreciated!

//Clem

paul-buerkner commented 3 years ago

changes to the prior should already be recognized by file_refit = "on_change". if you don't think that's the case please provide a minimal reproducible example.

clementbbp @.***> schrieb am Sa., 24. Apr. 2021, 10:00:

Hi,

I'm a relatively new user to brms and bayesian statistics, so please excuse if my request is already implemented in the package somewhere, or is unfeasible.

Would it make sense to implement a feature, perhaps similar to file_refit = "on_change", which updates the model not only after changing the formula, but also after calibrating the priors? In my short experience, you may find the priors too wide or constraining on the posterior and want to calibrate them, but you don't have the benefit of being able to save the model immediately using file = " ", as this would simply load the old model. I think it would be a nice addition to the workflow, but I might be wrong since I can also imagine that you want to think very hard about choosing your priors before finally "saving" the model.

Any answer would be greatly appreciated!

//Clem

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/paul-buerkner/brms/issues/1148, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADCW2AAMMCRXBT7LZZVTSILTKJ3DBANCNFSM43P4LULA .