Skip to contents

Unit-scale ELIR of a binomial power prior, interpolated in the power parameter

Source: R/power_prior_elir_cache.R
binomial_power_prior_unit_elir.Rd

Linear interpolation between the two grid nodes around power_parameter, each computed once per worker by grid_prior_unit_elir() and cached. The ELIR is close to linear in the power parameter, so the interpolation error is small against the fits' own.

Usage

binomial_power_prior_unit_elir(
  model,
  power_parameter,
  n_samples,
  step = 0.05,
  n_fits = 10L
)

Arguments

model

A binomial power prior model, whose quadrature_prior() takes the power parameter.

power_parameter

The power parameter, in \([0, 1]\).

n_samples

Draws per mixture fit.

step

Spacing of the power parameter grid.

n_fits

Number of mixture fits averaged per node.

Value

The unit-scale ELIR.