Skip to contents

Ordered Factor Prior

Source: R/priors.R
prior_ordered.Rd

Create priors on ordered factor effects by separating a scalar total effect from its allocation across ordered-level increments.

Usage

prior_ordered(
  total,
  allocation = NULL,
  contrast = "cumulative",
  id = NULL,
  prior_weights = 1
)

Arguments

total

scalar prior for the sum of ordered increments. With "cumulative", this is the first-to-last level difference; with "cumulative_levels", it is the last level's effect relative to the zero baseline preceding the first level.

allocation

allocation prior or fixed split. NULL creates a late-bound flat Dirichlet allocation with dimension determined by the formula term. Numeric vectors specify fixed splits. Dirichlet priors specify random allocations. For terms with multiple ordered factors, use a named list.

contrast

ordered contrast. "cumulative" uses nlevels(f) - 1 increments and sets the first level effect to zero. "cumulative_levels" uses nlevels(f) increments and gives the first level a non-zero allocation share.

id

optional allocation-sharing id. Equal ids share compatible allocation weights, not the total-effect prior. For terms with multiple ordered factors, use a named character vector.

prior_weights

numeric prior model weight.

Value

An object inheriting from prior.

Details

This is RoBMA's re-export of BayesTools::prior_ordered(). Spike-and-slab or mixture behavior belongs on total, e.g., prior_ordered(prior_spike_and_slab(...)). Unsupported ordered combinations raise typed BayesTools_ordered_unavailable conditions; constructor support does not imply support for every post-fit operation.

See also

prior_factor