rhomat: Calculate the correlation matrix according to the squared...

In demu: Optimal Design Emulators via Point Processes

Description

rhomat() is a helper function that constructs a correlation matrix according to the squared exponential model with parameterized by correlation parameters rho taking values in [0,1) and the exponent parameter alpha. The default of alpha=2 results in the Gaussian correlation while selecting alpha=1 corresponds to the Exponential correlation model. The design must have been already formated in distlist format using the function makedistlist().

Usage

rhomat(l.d,rho,alpha=2)

Arguments

l.d	Current design distance matrices in distlist format
rho	A vector of correlation parameters taking on values in [0,1)
alpha	Exponent parameter

Value

A list containing the constructed correlation matrix.

See Also

demu-package matern52 wendland1 wendland2 generalized.wendland

Examples

library(demu)

design=matrix(runif(10,0,1),ncol=2,nrow=5)
rho=rep(0.01,2)
l.d=makedistlist(design)
R=rhomat(l.d,rho)$R
R

demu documentation built on Jan. 13, 2020, 5:06 p.m.