update_SigmaINV_xCC: Gibbs sampling for Sigma^{-1} for xCC models

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

Description

Gibbs sampling for Σ^{-1} for xCC models

Usage

update_SigmaINV_xCC(x_data, z, y, Lambda, mu, K, alpha_sigma, beta_sigma)

Arguments

x_data	n\times p matrix containing the observed data
z	Allocation vector
y	n\times q matrix containing the latent factors
Lambda	K\times p\times q array with factor loadings
mu	K\times p array containing the marginal means
K	Number of components
alpha_sigma	Prior parameter alpha
beta_sigma	Prior parameter beta

Value

p\times p matrix with the common variance of errors per component Σ^{-1} = σ I_p.

Author(s)

Panagiotis Papastamoulis

Examples

library('fabMix')
# simulate some data
n = 8                # sample size
p = 5                # number of variables
q = 2                # number of factors
K = 2                # true number of clusters
sINV_diag = 1/((1:p))    # diagonal of inverse variance of errors
set.seed(100)
syntheticDataset <- simData(sameLambda=TRUE,K.true = K, n = n, q = q, p = p, 
                        sINV_values = sINV_diag)

# use the real values as input and update SigmaINV
update_SigmaINV_xCC(x_data = syntheticDataset$data, 
	z = syntheticDataset$class, 
	y = syntheticDataset$factors, 
	Lambda = syntheticDataset$factorLoadings, 
	mu = syntheticDataset$means, 
	K = K, 
	alpha_sigma = 0.5, beta_sigma = 0.5)

fabMix documentation built on Feb. 20, 2020, 1:09 a.m.