Posterior of the binomial robust mixture prior with a given weight
Source:R/binomial_rmp.R
binomial_rmp_posterior.RdThe posterior binomial_npp_posterior() gives for the kernels
of binomial_rmp_kernels(), computed without forming them: the mixture
kernel is linear in the weight, so the posterior mass at each risk
difference is weight times that of the informative component plus
1 - weight times that of the weak one, and the posterior weight of the
informative component is the former's share of the total. The component
masses are read off the components' kernels, which do not depend on the
weight, in a single weighted sum each; the lattice points visited are
those of binomial_npp_posterior().
Usage
binomial_rmp_posterior(
components,
weight,
n_control,
n_successes_control,
n_treatment,
n_successes_treatment
)Arguments
- components
Output of
binomial_rmp_components().- weight
Prior weight of the informative component.
- n_control, n_successes_control
Target control arm.
- n_treatment, n_successes_treatment
Target treatment arm.
Value
The list binomial_npp_posterior() returns for
binomial_rmp_kernels(components, weight).