Brms: view priors after customizing some of them

brms provides get_prior(formula, data), which shows the default priors. Is it possible to get a similar view of the priors after customizing some of them? I can’t find a get_prior(formula, data, priors) method.

You mean that after having sampled the model you would like to examine the priors that were used? If that is the case prior_summary(myModel).

I mean before sampling, to see the priors that will be used – the combination of the default priors and the priors I customize in the priors= argument.

Aah, the workflow you could use is;

  1. p <- get_prior(myFormula, data, likelihood)
  2. p$prior[1] <- "normal(0,1)" # set a prior manually
  3. p # check that prior has been set
  4. Now you can use p in the call to brm() as prior=p
2 Likes