
13
Gibbs Models
This chapter introduces Gibbs (including Markov) point process models, which are flexible and
natural models for point patterns with dependence between points. We explain the basic concepts,
and show how to build a Gibbs model ap propriate to the data, to fit the model to the data, to interpret
the results, and to identify w eaknesses in the analysis.
13.1 Introduction
Dependence between the points in a spatial point process is often loosely called interaction. But
the word ‘interaction’ suggests a very specific k ind of behaviour, in whic h individual points ‘act’
on each other, causing them to stay away from each other (‘inhibition’