Lattice-based statistics, often called nearest neighbor statistics, are area-based (e.g. county, blockgroup) methods to evaluate autocorrelation, identify clusters and to enhance traditional regression by including a spatial component.

performs the key lattice-based methods including calculating Moran’s I and Geary’s C, and conducting joint count analysis and spatial linear regression.

 

 

   

 

 


© 2007 ZevRoss Spatial Analysis

All Rights Reserved