compPmlvarsel: Computes the log PML for SUGS VarSel in the case of Gaussian...

In ococrook/sugsvarsel: SUGS algorithm of Wang and Dunson (2011), SUGS VarSel algorithm Crook et al.

Description

Computes the log PML for SUGS VarSel in the case of Gaussian mixtures

Usage

compPmlvarsel(X, K, N, D, n, phi, betaHat, m, nu, lambda, S, mu_0, nu_0,
  lambda_0, S_0, intfeature, w)

Arguments

X	The data matrix with observations as rows
K	The number of currently occupied clusters
N	The number of observations
D	The number of variables
n	The vector indicating the number of observations in each cluster
phi	Current value of posterior dirichlet weights.
betaHat	A grid of hyperparameters for the dirichlet concentration parameter, the default is c(0.01, 0.1, 1, 5, 10, 15, 30, 50, 100).
m	The current posterior mean
nu	The current posterior degrees of freedom
lambda	The current posterior mean variance
S	The current posterior scale vector
mu_0	The mean hyperparameter, default is the column means of the data matrix.
nu_0	The degrees of freedom hyperparameter, the default value is D, where D is the number of variables.
lambda_0	The variance of the Guassian mean prior, the dafault value is 0.01.
S_0	The scale hyperparamter, the deault value is a fifth of the column variance of the data matrix.
intfeature	A binary vector indicating whether features are irrelevant (0) or relevant (1).
w	A numerical vector of length 2 giving the prior probability of a variable being irrelevant or relevant. The first slot is irrelevant the second relevant.

Value

The log PML

ococrook/sugsvarsel documentation built on May 27, 2019, 12:12 p.m.