rtmvn: Truncated Multivariate Normal Distribution

In suchitm/tmvn: Generate random samples from truncated univariate and multivariate normal distributions and truncated Student-t distributions.

Description

Random vector generation for the truncated multivariate normal distribution using a Gibbs sampler.

Usage

rtmvn(n, Mean, Sigma = NULL, D, lower, upper, init, Sigma_chol = NULL)

Arguments

n	number of samples to be generated
Mean	mean vector
Sigma	covariance matrix
D	matrix of linear constraints
lower	vector of lower bounds
upper	vector of upper bounds
init	vector of initial values for the Gibbs sampler. Must satisfy the linear constraints.
Sigma_chol	the lower triangular cholesky of the covariance matrix. Only one of Sigma and Sigma_chol can be NULL.

Value

a matrix of samples with each column being an idependent sample.

References

Li, Y., & Ghosh, S. K. (2015). Efficient sampling methods for truncated multivariate normal and student-t distributions subject to linear inequality constraints. Journal of Statistical Theory and Practice, 9(4), 712-732.

Examples

Mean = rep(0,2)
rho = 0.5
Sigma = matrix(c(10,rho,rho,0.1),2,2)
D = matrix(c(1,1,1,-1),2,2)
varp = Sigma[1,1]+Sigma[2,2]+2*Sigma[1,2] # var of the sum
varm = Sigma[1,1]+Sigma[2,2]-2*Sigma[1,2] # var of the diff
sd = c(sqrt(varp),sqrt(varm))
lower = -1.5*sd; upper = 1.5*sd; init = rep(0,2)
n = 20
rtmvn(n,Mean,Sigma,D,lower,upper,init)

suchitm/tmvn documentation built on Dec. 10, 2020, 10:25 a.m.