Point Pattern Analysis

A set of point events such as disease, accidents, crime, emergency calls usually exhibit a spatial point pattern of random, clustered, or uniform nature. It is important to rely on simple and rapid methods to detect whether such patterns exhibit clustering and where clustering occurs. In the case of crime mapping clusters, when knowledge is gained on the type of pattern and its intensity, safety measures can be undertaken to better protect an area. This article overviews accessible techniques for point pattern analysis, which are usually supported by a geographical information system, statistical software such as R, or implementation via programming languages. The issue of scale is of paramount importance and it is important to keep in mind that it will influence the result: a spatial phenomenon may show significant clustering at fine scale but not at a coarser resolution. A sensitivity analysis that evaluates the effect of scale on the resulting point pattern is desirable.