SpatialPack: Computing the Association Between Two Spatial Processes

Abstract

An R package SpatialPack that implements routines to compute point estimators and perform hypothesis testing of the spatial association between two stochastic sequences is introduced. These methods address the spatial association between two processes that have been observed over the same spatial locations. We briefly review the methodologies for which the routines are developed. The core routines have been implemented in C and linked to R to ensure a reasonable computational speed. Three examples are presented to illustrate the use of the package with both simulated and real data. The particular case of computing the association between two time series is also considered. Besides elementary plots and outputs we also provide a plot to visualize the spatial correlation in all directions using a new graphical tool called codispersion map. The potential extensions of SpatialPack are also discussed.