papers.GSPSJ06: Fast and Exact Simulation of Large Gaussian Lattice Systems...

Description References Examples

Description

Here, the code of the paper on ‘Fast and Exact Simulation of Large Gaussian Lattice Systems in R2’ is given.

References

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
27
28
29
30
RFoptions(seed=0) ## *ANY* simulation will have the random seed 0; set
##                   RFoptions(seed=NA) to make them all random again

## Figure 1 (pretty time consuming)
stabletest <- function(alpha, theta, size=512) {
  RFoptions(trials=1, tolIm = 1e-8, tolRe=0, force = FALSE,
            useprimes=TRUE, strategy=0, skipchecks=!FALSE,
             storing=TRUE)
  model <- RMcutoff(diameter=theta, a=1, RMstable(alpha=alpha))
  RFcov(dist=0, model=model, dim=2, seed=0)
  r <- RFgetModelInfo(modelname="RMcutoff", level=3)$storage$R_theor
  x <- seq(0, r, by= r / (size - 1)) * theta
  err <- try(RFsimulate(x, x, model=RPcirculant(model), n=0))
  return(if (class(err) == "try-error") NA else r)
}

alphas <- seq(1.52, 2.0, 0.02) 
thetas <- seq(0.05, 3.5, 0.05)

m <- matrix(NA, nrow=length(thetas), ncol=length(alphas))
for (it in 1:length(thetas)) {
  theta <- thetas[it]
  for (ia in 1:length(alphas)) {
  alpha <- alphas[ia]
  cat("alpha=", alpha, "theta=", theta,"\n")
  m[it, ia] <- stabletest(alpha=alpha, theta=theta)
  if (is.na(m[it, ia])) break
  }
  if (any(is.finite(m))) image(thetas, alphas, m, col=rainbow(100))
}

RandomFields documentation built on Feb. 6, 2020, 5:13 p.m.