CorMat_mcmc_summary: Compute quantiles for the correlation matrix.

In fabMix: Overfitting Bayesian Mixtures of Factor Analyzers with Parsimonious Covariance and Unknown Number of Components

Compute quantiles for the correlation matrix.

Description

Compute quantiles for the correlation matrix per cluster based on the MCMC output stored in a fabMix.object.

Usage

CorMat_mcmc_summary(x, quantile_probs)

Arguments

x	An object of class fabMix.object.
quantile_probs	Vector of probabilities for computing the corresponding quantiles.

Value

quantiles	A list containing the quantiles for the correlation matrix per component. Each element is a p\times p \times K array, where p and K denote the dimension of the multivariate data and number of alive components for the selected model, respectively.
p_matrix	A p\times p\times K array, where for each k = 1,\ldots,K the p\times p matrix p_matrix[,,k] contains the posterior probability 1-2|0.5-P(\rho_{ij} > 0)| for i=1,\ldots,p; j=1,\ldots,p.

Author(s)

Panagiotis Papastamoulis

fabMix documentation built on May 29, 2024, 2:53 a.m.