spatstat.random

Random Generation and Simulation for the spatstat family

The original spatstat package has been split into several sub-packages (see spatstat/spatstat)

This package spatstat.random is one of the sub-packages. It contains the functions for random generation of data and simulation of models.

You are viewing the GitHub repository which holds the latest development version of spatstat.random. For the latest public release on CRAN, click the green badge above.

Contents:

• Overview of package
• Detailed contents of package
• Installing the package

---

<a name="overview"></a> Overview

spatstat.random supports

• generating random spatial patterns of points according to many simple rules (complete spatial randomness, binomial process, random grid, systematic random, stratified random, simple sequential inhibition, cell process),

• randomised alteration of patterns (thinning, random shift, jittering),

• generating simulated realisations of spatial point processes (Poisson processes, Matern inhibition models, Matern cluster processes, Neyman-Scott cluster processes, log-Gaussian Cox processes, product shot noise cluster processes, Gibbs point processes)

• generating simulated realisations of Gibbs point processes (Metropolis-Hastings birth-death-shift algorithm; perfect simulation/ dominated coupling from the past; alternating Gibbs sampler)

• generating random spatial patterns of line segments

• generating random tessellations

• generating random images (random noise, random mosaics).

Exceptions:

• generation of determinantal point processes is provided in spatstat.model

• generation of quasi-random patterns is provided in spatstat.geom

---

<a name="detailed"></a> Detailed contents

Generating random patterns

• binomial random patterns (runifpoint, rpoint, rmpoint, runifdisc)
• completely random patterns (rpoispp, rmpoispp)
• systematic random patterns (rstrat, rsyst)

Point process models defined by simple constructions

• simple sequential inhibition (rSSI)
• Matern inhibition models (rMaternI, rMaternII)
• cell process (rcell)

Randomly changing an existing point pattern

• random shift (rshift)
• random thinning (rthin)
• random (re)labelling (rlabel)
• block resampling (quadratresample)

Clustered point processes

• log-Gaussian Cox process (rLGCP)
• Neyman-Scott cluster processes (rThomas, rMatClust, rCauchy, rVarGamma)
• general Neyman-Scott cluster process (rNeymanScott)
• general Poisson cluster process (rPoissonCluster)
• Gauss-Poisson process (rGaussPoisson)

Gibbs point processes

• perfect simulation algorithms for specific Gibbs models (rHardcore, rStrauss, rStraussHard, rDiggleGratton, rDGS, rPenttinen,
• Metropolis-Hastings simulation algorithm for Gibbs models (rmh)
• alternating Gibbs sampler for multitype Gibbs processes (rags, ragsMultiHard)
• alternating Gibbs sampler for area-interaction process (ragsAreaInter)

random points along lines

• random points along specified line segments (runifpointOnLines, rpoisppOnLines)

random pixel images and random sets

• random pixel noise (rnoise)
• random mosaic (rMosaicField, rMosaicSet)

random line segment patterns

• Poisson line process (rpoisline)

random tessellations

• tessellation using Poisson line process (rpoislinetess)

three-dimensional point patterns

• uniform random points in 3D (runifpoint3)
• Poisson point process in 3D (rpoispp3)

multi-dimensional point patterns

• uniform random points in space or space-time (runifpointx)
• Poisson point process in space or space-time (rpoisppx)

probability distributions

• theoretical distribution of nearest neighbour distance (rknn)
• mixed Poisson distribution (dmixpois)

---

<a name="installing"></a> Installing the package

This repository contains the development version of spatstat.random. The easiest way to install the development version is to start R and type

repo <- c('https://spatstat.r-universe.dev', 'https://cloud.r-project.org')
install.packages("spatstat.random", dependencies=TRUE, repos=repo)

To install the latest public release of spatstat.random, type

install.packages("spatstat.random")