s-news
[Top] [All Lists]

Re: cluster simulation

To: gebove@geog.buffalo.edu
Subject: Re: cluster simulation
From: Rolf Turner <rolf@math.unb.ca>
Date: Mon, 30 Aug 2004 10:11:51 -0300 (ADT)
Cc: s-news@wubios.wustl.edu
Gerald E Bove Jr wrote:

> I only have experience running simple simulations in splus, would
> like to run a simulation which would represent both a random spatial
> pattern (~100pts.), and a spatial pattern which is clustered around
> specified points in the study area, is there a way to do this?

Yes and no.  No in Splus, as such (as far as I know).  Yes if you get
and fire up R, and install the spatstat package from CRAN.  In
spatstat there are a number of functions which simulate clustered
spatial point processes: rMatClust, rNeymanScott, rThomas, and rmh.
I would guess that you'd find rThomas satisfactory.

Note that (a) R is free (GNU public license) and (b) its syntax is
***almost*** the same as that of Splus --- so there's not too huge an
amount for you to learn in order to use it.

                                        cheers,

                                                Rolf Turner
                                                rolf@math.unb.ca

<Prev in Thread] Current Thread [Next in Thread>