Comparing Three Spatial Cluster Tests from Rare to Common Spatial Events

In the past few years, several new tests for spatial clustering have been proposed. With ever increasing capability of GIS and wider availability of spatial statistic functions, spatial analysts are likely to face challenge of properly using these tests. Seemingly gaps also exist between the development of new tests and follow up evaluations against various assumptions. In this research note, I compare three cluster tests along a range of distribution from rare to common spatial events. The results not only revealed sensitive data feature that each test is designed to detect, but also clarified the interpretation based on the nature of the test.