Learn R Programming

BNPMIXcluster (version 1.1.0)

sampling_Omega_ij: Simulation of "\(\Omega_{i,j}\)" in the mixdpcluster model for bayesian clustering.

Description

Generates simulation of the posterior distribution of the \((i,j)\) element of the \(\Omega\) matrix in the mixdpcluster model for bayesian clustering. The simulation is done via Metropolis-Hastings method.

Usage

sampling_Omega_ij(n = 1, Omega.ini, i, j, delta = 4, Z, mu_Z, Lambda,
  sampling_prob, n.burn = 0, n.thin = 0, max.time = Inf, verbose = F,
  USING_CPP = USING_CPP)

Arguments

n

number of simulations to be generated

Omega.ini

matrix \(\Omega\) with an initialization value for \(\Omega_{i,j}\).

i

indicates the row for \(\Omega_{i,j}\)

j

indicates the column for \(\Omega_{i,j}\)

delta

defines the maximum jump on each iteration of the MCMC as \(1/delta\) of the feasible interval for \(\Omega_{i,j}\)

Z

bla

mu_Z

bla

Lambda

bla

sampling_prob

bla

n.burn

number of iterations in the simulation considered in the burn-in period.

n.thin

number of iterations discarded between two simulated values (for thinning of the MCMC chain).

max.time

maximum allowed time for the simulation process. The function returns Error if exceeded.

verbose

if T, the function reports extra information on progress.

Value

A list with two elements:

$omega_ij.chain

A numeric vector with the simulated values from the posterior distribution of \(\Omega_{i,j} \).

$accept.indic

A logical vector indicating whether or not omega_ij.prop was accepted.

References

Carmona C., Nieto-Barajas L., Canale A. (2017). Model based approach for household clustering with mixed scale variables.