loglik.approx: A function to calculate the approximated log-likelihood

In yuanfang90/LNMVGA: Cluster microbiome / compositional data using a mixture of logistic-normal multinomial models

A function to calculate the approximated log-likelihood

Description

This function calculate approximated log-likelihood using the elbo.

Usage

loglik.approx(W, m, V, Vmat, mu, Sig, red_sig, pi_g, G, it)

Arguments

W	observation data: the count data, n*(K+1) matrix where K is the number of taxa.
m	variational parameter m, vector of n.
V	variational parameter V, list of n matrices, each element is a (K+1)*(K+1) matrix.
Vmat	internal parameter, a matrix of n rows as n being number of observations; each row is the diagonal element of the variational parameter V.
mu	mean parameter for the latent Gaussian variable, list of g vectors of (K+1).
Sig	variance parameter for the latent Gaussian variable, list of g (K+1)*(K+1) matrices.
red_sig	internal parameter, a list of G K*K dimensional matrices from the variance parameter.
pi_g	proportion vector of each component.
G	number of component.
it	number of iteration, keep track of iterations for condition checking
xi	variational parameter xi, vector of n.

Examples

loglik.approx()

yuanfang90/LNMVGA documentation built on Jan. 29, 2024, 8:24 a.m.