BSGS.Sample {BSGS}R Documentation

Sample version of group-wise Gibbs sampler for sparse group selection.

Description

Generate the posterior samples by an approximation sampling method to perform Bayesian sparse group selection to identify the important groups of variables and variables within those.

Usage

BSGS.Sample(Y, X, Group.Index, r.value, eta.value, beta.value, tau2.value, 
rho.value, theta.value, sigma2.value, nu, lambda, Num.of.Iter.Inside.CompWise, 
Num.Of.Iteration, MCSE.Sigma2.Given)

Arguments

Y

vector of observations of length n.

X

design matrix of dimension n \times p.

Group.Index

Specify the group label to which the variable belongs.

r.value

Initial values of indicator variables for individual variables.

eta.value

Initial values of indicator variables for the group variables.

beta.value

Initial values of regression coefficients, β.

tau2.value

Variance in the prior distribution for regression coefficients.

rho.value

Prior inclusion probability for a variable.

theta.value

Prior inclusion probability for a group.

sigma2.value

Initial value of σ^2.

nu

The hyperparameter in the prior distribution of σ^2.

lambda

The hyperparameter in the prior distribution of σ^2.

Num.of.Iter.Inside.CompWise

Specify the number of iterations within component wise Gibbs sampler for variable selection within a group.

Num.Of.Iteration

Specify the number of iterations for sparse group variable selection.

MCSE.Sigma2.Given

Prespecified value which is used to stop simulating samples. When the MCSE of estimate of σ^2 less than the given value, the simulation is terminated.

Value

A list is returned with posterior random samples of regression coefficients, β, binary variables for group selection, η, binary variables for variable selection, γ, variance, σ^2 and the number of iterations performed and the elapsed time in second required for the run.

Examples



## Not run: 
output = BSGS.Sample(Y, X, Group.Index, r.value, eta.value, beta.value, tau2.value, 
	rho.value, theta.value, sigma2.value, nu, lambda, 
	Num.of.Iter.Inside.CompWise, Num.Of.Iteration, MCSE.Sigma2.Given)

## End(Not run)


[Package BSGS version 2.0 Index]