post_mix_dist: Generate the posterior distribution when the error...

In dcgerard/stramash: STRAMASH: Specifying The Right Amount of Mixture-likelihood for Adaptive SHrinkage

Description

Generate the posterior distribution when the error distribution is a mixture.

Usage

post_mix_dist(g, betahat, errordist)

Arguments

g	The prior distribution. Either of class "normalmix" or "unimix".
betahat	The observations, a vector of numerics.
errordist	A list of mixture distributions.

Value

A list of arrays containing the parameters for the posterior. The first dimension is the observations, the second is the mixture components for the prior, and the third is the mixture components of the error.

weights The mixing proportions

means The location parameter if the posterior is either a mixture of normals or truncated normals.

variances The scale parameter if the posterior is either a mixture of normals or truncated normals.

lower The lower bound of support if the posterior is either a mixture of truncated normals of uniforms.

upper The upper bound of support if the posterior is either a mixture of truncated normals of uniforms.

Author(s)

David Gerard

dcgerard/stramash documentation built on May 15, 2019, 1:24 a.m.