0
$\begingroup$

Is it possible to generate points pattern with Poisson and Thomas process in R using spatstat package with specified number of points? I had genereted Strauss point pattern using the Metropolis-Hastings algorithm with specified number of points (20 points) like this:

Model <- list(cif="straush",par=list(beta=10,gamma=0.5,r=0.5,hc=0.5),
          w=c(0,10,0,10))
P <- rmh(model=Model,start=list(n.start=20),
              control=list(p=1,nrep=1e5,nverb=5000))

but I can not generate Poisson or Thomas point pattern using the Metropolis-Hastings algorithm.

$\endgroup$
1
$\begingroup$

As also pointed out in off-site correspondence:

A Poisson process conditional on the number of points is just a collection of independent and identically distributed points with density proportional to the intensity, so you can use the spatstat function rpoint to simulate that process.

I don't think it is mathematical tractable to express the distribution of a Thomas process conditional on the total number of points inside the observation window.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.