Skip to contents

The prior-predictive probability of every pair of target responder counts under each component, as the matrices egidi_binomial_conflict_pvalue() reads: one row per control count and one column per treatment count. On the lattice, the table is t(B_c) K B_t, with K the component in target control and treatment rates and B the binomial probabilities of each count at each rate.

Usage

binomial_rmp_predictive_tables(
  components,
  source_counts,
  n_control,
  n_treatment
)

Arguments

components

Output of binomial_rmp_components().

source_counts

The source counts the components were built from, which identify them in the cache.

n_control, n_treatment

Target arm sizes.

Value

A list with informative and weak tables.