New Zealand Statistical Association 2024 Conference

Rolf Turner

University of Auckland

Pseudo analysis of variance of K-functions

This is joint work with Peter Diggle (Lancaster University)

Data consisting of independent point patterns may be collected according to some classification structure or experimental design. In such circumstances it may be of interest to formally test whether the patterns vary in nature according to the classification structure. Such variation could be expressed in term of the K-functions (or other summary functions) of the patterns. Hahn (2012) and Diggle et al. (2000) investigated these ideas for one-way classifications. In this talk we extend their ideas to two-way (cross) classifications. A major goal is to test for an effect from one classification factor, allowing for the possible impact of a second factor. This work was fundamentally motivated by the study of point patterns of stomata in plant leaves, Peijian Shi et al. (2021).