UpdatePiWeighted: Update pi and u.

In NestedCategBayesImpute: Modeling, Imputing and Generating Synthetic Versions of Nested Categorical Data in the Presence of Impossible Combinations

Description

Update pi – the vector of the probabilities for the group-level latent classes – and u – the vector of the beta-distributed variables in the stick breaking representation of the group-level latent classes when the weighting/capping option is used. The weighting options allows capping the number of impossible households to sample and re-weight the multinomial counts within each latent class back to the expected truth.

Usage

UpdatePiWeighted(alpha, G_all, FF, struc_weight)

Arguments

alpha	Concentration parameter in the Dirichlet process for the group-level latent classes
G_all	A vector of the household-level latent class indexes for all households both in the original data and the sampled impossible households.
FF	Maximum number of household-level latent classes allowed.
struc_weight	A vector of weights by household sizes used in capping the number of sampled impossible households.

Details

Function for obtaining a posterior sample of pi when the weighting/capping option is used.

Value

A list containing the updated (posterior) values for pi and u based on the corresponding full conditional distributions.

Author(s)

Quanli wang, Olanrewaju Akande

NestedCategBayesImpute documentation built on May 2, 2019, 3:26 p.m.