sample.theta: Sample theta and vk parameters

In drvalle1/EcoCluster: Bayesian Clustering using Truncated Stick-Breaking priors

Description

Sample vk parameters from their full conditional distributions and convert these vk parameters into theta parameters

Usage

sample.theta(ngroup.loc, gamma.v, burnin, gibbs.step, theta, psi, z)

Arguments

ngroup.loc	maximum number of groups for locations (KL)
gamma.v	truncated stick-breaking prior parameter for the location groups. This value should be between 0 and 1, and small values enforce more parsimonius results (i.e., fewer groups)
burnin	number of MCMC samples that are going to be thrown out as part of the burn-in phrase
gibbs.step	current iteration of the gibbs sampler
theta	vector of size KL containing the current estimate of theta (i.e., probability of each location group)
psi	KL x KS matrix containing the current estimate of psi
z	this is a vector of length L, storing the current membership of each location

Value

this function returns a list containing 4 items (theta, vk, psi, and z)

drvalle1/EcoCluster documentation built on Jan. 25, 2022, 12:46 a.m.