# complete.log.likelihood_q0: Complete log-likelihood function for xUx models and q=0 In fabMix: Overfitting Bayesian Mixtures of Factor Analyzers with Parsimonious Covariance and Unknown Number of Components

 complete.log.likelihood_q0 R Documentation

## Complete log-likelihood function for xUx models and q=0

### Description

Complete log-likelihood function for models with different error variance per component (xUx) and diagonal covariance structure per component (q=0.

### Usage

complete.log.likelihood_q0(x_data, w, mu, SigmaINV, z)


### Arguments

 x_data n\times p matrix containing the data w a vector of length K containing the mixture weights mu K\times p matrix containing the marginal means per component SigmaINV K\times p\times p precision matrix (inverse covariance) per component z A vector of length n containing the allocations of the n datapoints to the K mixture components

### Value

complete log-likelihood value

### Author(s)

Panagiotis Papastamoulis

### Examples

	library('fabMix')
data(waveDataset1500)
x_data <- scale(waveDataset1500[ 1:20, -1]) # data
z <-  waveDataset1500[ 1:20, 1]	# class
p <- dim(x_data)[2]
q <- 2
K <- length(table(z))		# 3 classes
# give some arbitrary values to the parameters:
set.seed(1)
w <- rep(1/K, K)
mu <- array( -0.1 + 0.2*runif(K * p), dim = c(K,p) )
SigmaINV <- array( 1, dim = c(K,p,p))
# compute the complete.log.likelihood ( -inf )
complete.log.likelihood_q0(x_data = x_data, w = w, mu = mu,
SigmaINV = SigmaINV, z = z)


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