A class of local and global K functions and their exact statistical methods

Abstract

In 1987 Getis and Franklin introduced a technique, based on secondorder methods, for quantifying clustering at various scales in mapped point patterns. Subsequently, this technique has become known as localK function analysis. In this paper we develop the local and global forms of a class of K functions and cross K functions formulated on a bounded plane that includes the technique of Getis and Franklin. Exact statistical methods are formulated or discussed and computational methods are shown for the functions.