spatstat.model (version 3.4-2)

simulate.ppm: Simulate a Fitted Gibbs Point Process Model

Description

Generates simulated realisations from a fitted Gibbs or Poisson point process model.

Usage

# S3 method for ppm
simulate(object, nsim=1, ...,
                         singlerun = FALSE,
                         start = NULL,
                         control = default.rmhcontrol(object, w=w),
                         w = window, 
                         window = NULL, 
                         project=TRUE, new.coef=NULL, newdata=NULL, 
                         verbose=FALSE, progress=(nsim > 1),
                         drop=FALSE)

Value

A list of length nsim containing simulated point patterns (objects of class "ppp"). It also belongs to the class "solist", so that it can be plotted, and the class "timed", so that the total computation time is recorded.

Arguments

object

Fitted point process model. An object of class "ppm".

nsim

Number of simulated realisations.

singlerun

Logical. Whether to generate the simulated realisations from a single long run of the Metropolis-Hastings algorithm (singlerun=TRUE) or from separate, independent runs of the algorithm (singlerun=FALSE, the default). Ignored when the model is a Poisson process.

start

Data determining the initial state of the Metropolis-Hastings algorithm. See rmhstart for description of these arguments. Defaults to list(n.start=npoints(data.ppm(object))) meaning that the initial state of the algorithm has the same number of points as the original dataset.

control

Data controlling the running of the Metropolis-Hastings algorithm. See rmhcontrol for description of these arguments.

w, window

Optional. The window in which the model should be simulated. An object of class "owin".

...

Further arguments passed to rmhcontrol, or to rmh.default, or to covariate functions in the model.

project

Logical flag indicating what to do if the fitted model is invalid (in the sense that the values of the fitted coefficients do not specify a valid point process). If project=TRUE the closest valid model will be simulated; if project=FALSE an error will occur.

verbose

Logical flag indicating whether to print progress reports from rmh.ppm during the simulation of each point pattern.

progress

Logical flag indicating whether to print progress reports for the sequence of simulations.

new.coef

New values for the canonical parameters of the model. A numeric vector of the same length as coef(object).

newdata

List of pixel images, functions or numeric values to replace the data argument of the original object.

drop

Logical. If nsim=1 and drop=TRUE, the result will be a point pattern, rather than a list containing a point pattern.

Simulating a model on a different window

It is possible to simulate the model object on a window that is different from the original window (the window containing the data to which the model was fitted). To do this, specify the argument w as the window in which the simulated patterns should be generated, as in simulate(object, w=wsim). If the model is stationary, then this is sufficient. However if the model depends on spatial covariates, then the values of these covariates inside the new simulation window will be required. If the original covariates are functions, or pixel images that are defined on a region that includes the new simulation window, then no further action is needed, and simulate(object, w=wsim) should work. Otherwise, the values of the required covariates inside the new simulation window should be supplied in the argument newdata.

Author

Adrian Baddeley and Rolf Turner

Details

This function is a method for the generic function simulate for the class "ppm" of fitted point process models.

Simulations are performed by rmh.ppm.

If singlerun=FALSE (the default), the simulated patterns are the results of independent runs of the Metropolis-Hastings algorithm. If singlerun=TRUE, a single long run of the algorithm is performed, and the state of the simulation is saved every nsave iterations to yield the simulated patterns.

In the case of a single run, the behaviour is controlled by the parameters nsave,nburn,nrep. These are described in rmhcontrol. They may be passed in the ... arguments or included in control. It is sufficient to specify two of the three parameters nsave,nburn,nrep.

See Also

ppm, simulate.kppm, simulate

Examples

  op <- spatstat.options(rmh.nrep=10)
  fit <- ppm(japanesepines ~1, Strauss(0.1))
  simulate(fit, 2)
  simulate(fit, 2, singlerun=TRUE, nsave=1e4, nburn=1e4)
  spatstat.options(op)