Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
415218 | Computational Statistics & Data Analysis | 2009 | 11 Pages |
Abstract
In this paper we propose a new class of spatial concentration measures to verify hypotheses on the spatial concentration of empirical phenomena. Our proposed measures incorporate the ideas of both a-spatial concentration and spatial autocorrelation. We discuss many issues related to the approximate sampling theory and we suggest a Monte Carlo test for the presence of significant spatial concentration. A simulation experiment studying the relationship of the new measure to different spatial patterns and to different levels of a-spatial variability is also reported. A discussion related to irregular space is included.
Related Topics
Physical Sciences and Engineering
Computer Science
Computational Theory and Mathematics
Authors
Giuseppe Arbia, Gianfranco Piras,