Closed kyleam closed 5 months ago
I've updated the series. I put it on top param_estimates_batch
fix (gh-648) and made the follow changes:
updated model_summaries_onetry()
to put NULL
as the summary object when a model_summary
errors (696ea867)
dropped the warning handling from summary_log_impl()
given the model_summary.bbi_stan_model()
to error instead of warn (a443ed4b)
log message updates for the fact that model_summary.bbi_{stan,nmbayes}_model
now error
Once those changes are pushed to https://github.com/metrumresearchgroup/bbr.bayes/pull/123, I'll add a comment here pointing to the Drone run.
Once those changes are pushed to metrumresearchgroup/bbr.bayes#123, I'll add a comment here pointing to the Drone run.
That's here: https://github-drone.metrumrg.com/metrumresearchgroup/bbr.bayes/445/1/4. The failure is unrelated and probably due to rate limiting. I'll update this message with all green build once we have one, but this is ready for review.
Edit: Passing builds (on restart):
@seth127 Thanks for taking another look (and for catching the issue with putting draws in the summary_log output)
summary_log
is incompatible with nmbayes models: https://github.com/metrumresearchgroup/bbr.bayes/issues/40To make it work, we need
summary_log
logic to work with the nmbayes, in particular the object returned bymodel_summary.bbi_nmbayes_model()
. (Edit: The bbr.bayesmodel_summary
methods now error; discussion)This series handles the second item above.