I keep getting this error when running brm function from brms package for building bayesian model. I couldnt find what causes this error
Anyone have any suggestions on how to solve this? Thanks a lot for help!
Asked
Active
Viewed 837 times
3
-
1This happens to me often, but it has dropped in frequency once I started save my models and not running multiple models before restarting R. I normally find that running a bunch of models or intensive functions on models like `loo()` will eventually lead to R Studio crashing – sjp Mar 06 '22 at 13:16
-
1I've heard from numerous people that using R from the console leads to this happening less. But I haven't found anything that stops this from happening – sjp Mar 06 '22 at 13:19
-
@sjp I did everything to solve this but failed. I deleted and re-install r and rstudio but still getting the same error – Ozgur Alptekın Mar 06 '22 at 13:39
-
It only happens to me at the very end of a long run (200K with 10K warmup), when I can see that the last chains have reached 90% sampling! I was using `adapt_delta=0.90`. I'm going to try package upgrade as mentioned here https://discourse.mc-stan.org/t/brms-causing-r-session-aborted-r-encountered-fatal-error-only-with-control-parameters-included/24361/5. – Timothy M Pollington Jul 22 '22 at 02:14