GHS_est: GHS MCMC sampler using data-augmented block (column-wise)...

Description Usage Arguments Examples

View source: R/GHS_function_test.R

Description

GHS_est returns a tuple whose first element is a p by p by nmc matrices of saved posterior samples of precision matrix, second element is the p*(p-1)/2 by nmc vector of saved samples of the local tuning parameter and the third element is the 1 by nmc vector of saved samples of the global tuning parameter

Usage

1
GHS_est(S, n, burnin, nmc)

Arguments

S

sample covariance matrix

n

sample size

burnin

number of MCMC burnins

nmc

number of saved samples

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
# This function generates positive definite matrices for testing purposes 
# with specificied eigenvalues
Posdef <- function (n,ev)
{
  Z <- matrix(ncol=n, rnorm(n^2))
  decomp <- qr(Z)
  Q <- qr.Q(decomp)
  R <- qr.R(decomp)
  d <- diag(R)
  ph <- d / abs(d)
  O <- Q %*% diag(ph)
  Z <- t(O) %*% diag(ev) %*% O
  return(Z)
}
eig1 <- rep(1,2)
eig2 <- rep(0.75,3)
#eig3 <- rep(0.25,3)
eig_val <- c(eig1,eig2)
z <- Posdef(5,eig_val)
Mu <- rep(0,5)
Sigma <- solve(z)
Y <- mvrnorm(n=5,mu=Mu,Sigma=Sigma)
S <- t(Y)%*%Y
out <- GHS_est(S,50,100,5000)
est_matrix <- apply(out[[1]],c(1,2),mean)
image(est_matrix)

Example output

Loading required package: MASS

GHS documentation built on May 1, 2019, 7:49 p.m.

Related to GHS_est in GHS...