Tests for isotropy in spatial point patterns – A comparison of statistical indices
Journal article, 2022

Isotropy of a point process, defined as invariance of the distribution under rotation, is often assumed in spatial statistics. Formal tests for the hypothesis of isotropy can be created by comparing directional summary statistics in different directions. In this paper, the statistical powers of tests based on a variety of summary statistics and several choices of deviance measures are compared in a simulation study. Four models for anisotropic point processes are considered covering both regular and clustered cases. We discuss the robustness of the results to changes of the tuning parameters, and highlight the strengths and limitations of the methods.

Test statistics

Power analysis

Monte-Carlo test

Spectral analysis

Fry plot

Second-order characteristics

Wavelets

Spatial summary statistics

Author

Tuomas Rajala

Natural Resources Institute Finland (Luke)

Claudia Redenbach

Technische Universität Kaiserslautern

Aila Särkkä

Chalmers, Mathematical Sciences, Applied Mathematics and Statistics

Martina Sormani

Fraunhofer-Institut fur Techno- und Wirtschaftsmathematk

Spatial Statistics

2211-6753 (ISSN)

Vol. 52 100716

Extracting information from complicated point patterns

Swedish Research Council (VR) (2018-03986), 2019-01-01 -- 2022-12-31.

Roots

Basic sciences

Subject Categories

Probability Theory and Statistics

Areas of Advance

Materials Science

DOI

10.1016/j.spasta.2022.100716

More information

Latest update

1/16/2024