Modelling Spatial Point Patterns in R

This paper describes practical techniques for fitting stochastic models to spatial point pattern data using the statistical language R. The techniques are demonstrated with a detailed analysis of two real datasets. We have implemented the techniques as a package spatstat in the R language. Both spatstat and R are freely available from the R website [19]. Sections 2 and 3 introduce the spatstat package. Theory of point process models is covered in Section 4, while Section 5 describes how to fit models in spatstat, and Section 6 explains how to interpret the fitted models obtained from the package. Models involving external covariates are discussed in Section 7, and models for multitype point patterns in Section 8. Estimation of irregular parameters is discussed in Section 9. Section 10 discusses formal inference for models. Examples are analysed in Sections 11–12.