Use priors and covariates to generate a prior predictive distribution

Use priors and covariates to generate a prior predictive distribution

prior_pd_binomial(
  .formula,
  .data,
  verbose = TRUE,
  .prior = c(prior_string("normal(0, 1)", class = "b"), prior_string("normal(0, 1)",
    class = "sd"), prior_string("normal(0, 1)", class = "Intercept")),
  .iter = 1000,
  .warmup = floor(.iter/2),
  .cores = 2,
  .chains = 2,
  .backend = "rstan",
  .seed = 2138
)

Arguments

.formula

model specification

.data

collapsed survey dataset, built from ccesMRPprep::build_counts

verbose

Whether to show iteration messages

.prior

prior specification that can be interpreted by brms. The default is a standard normal prior, which is tighter than the brms default but has shown to have good prior posterior draws

.iter

Number of total iterations.

.warmup

Of the iterations, how much are burn-ins. Defaults to half.

.cores

Number of cores to uses

.chains

Number of chains to pass on fit_brms

.backend

The backend argument of brms. Defaults to "rstan", can also be "cmdstanr"

.seed

seed for randomization to pass into brm