Prior Configuration for HBSAE Models — hbm_priors • hbsaems

Bundles the shrinkage-prior arguments of hbm into a single named list. Same opt-in pattern as hbm_control.

Usage

hbm_priors(
  prior_type = c("default", "horseshoe", "r2d2"),
  prior = NULL,
  hs_df = 1,
  hs_df_global = 1,
  hs_df_slab = 4,
  hs_scale_global = NULL,
  hs_scale_slab = 2,
  hs_par_ratio = NULL,
  hs_autoscale = TRUE,
  r2d2_mean_R2 = 0.5,
  r2d2_prec_R2 = 2,
  r2d2_cons_D2 = NULL,
  r2d2_autoscale = TRUE
)

Arguments

prior_type: Character. One of "default", "horseshoe", "r2d2".
prior: Optional brmsprior for explicit priors that override the registry default.
hs_df, hs_df_global, hs_df_slab, hs_scale_global, hs_scale_slab, hs_par_ratio, hs_autoscale: Horseshoe-prior hyperparameters; see hbm.
r2d2_mean_R2, r2d2_prec_R2, r2d2_cons_D2, r2d2_autoscale: R2D2-prior hyperparameters; see hbm.

Value

A named list of arguments for hbm.

See also

hbm_control, hbm_nonlinear, hbm

Examples

p_hs   <- hbm_priors(prior_type = "horseshoe", hs_df = 1, hs_df_slab = 4)
p_r2d2 <- hbm_priors(prior_type = "r2d2",      r2d2_mean_R2 = 0.5)