Define a Plackett–Luce ranking-mixture model

Ranking rows are assigned to latent preference groups. Each group has a separate vector of item strengths.

Usage

pl_mixture_model(
  latent_strength = .default_latent_strength(),
  ranking_clustering = gnedin_prior(gnedin_hyperparameter = 0.5),
  identifiability = pl_identifiability(),
  ability_prior = NULL,
  ranking_prior = NULL
)

Arguments

latent_strength

Gamma prior for positive component strengths.

ranking_clustering

Prior on latent ranking groups.

identifiability

PL reporting constraint.

ability_prior

Deprecated alias for latent_strength.

ranking_prior

Deprecated alias for ranking_clustering.

Value

A btsbm_model.