Spatial Statistics Software
The following lists some software available for fitting spatial models of various sorts, with an emphasis on spatial smoothing.
Smoothing of location-specific data
A number of packages in R allow one to smooth point-level data. The mgcv package fits additive models with the ability to model spatial surfaces as nonparametric smooth two-dimensional surfaces.
The SemiPar package fits spatial surfaces as well as additive models using a mixed models framework, with capabilities similar to mgcv. Both mgcv and SemiPar can easily fit generalized models with spatial structure, in particular spatial logistic and spatial Poisson regression.
The geoR package implements classical and Bayesian implementations of kriging, while the geoRglm package fits generalized spatial models.
The fields package implements thin plate spline fitting and kriging.
Some of these packages are also available for S-plus. Head-Bang PC Software “Head-banging” is a weighted two-dimensional median-based smoothing algorithm available from NCI.
Markov random field models
Markov random field models are designed for data aggregated into regions; they smooth data based on the neighborhood structure of the regions. The GeoBUGS software package has been developed by a team at the Department of Epidemiology and Public Health of Imperial College at St Mary’s Hospital, London. It is an add-on to WinBUGS that fits spatial models and produces a range of maps as output. Bayesian inference is used to spatially smooth the standardized incidence/mortality ratios using Markov Chain Monte Carlo (MCMC) methods. GeoBUGS implements models for data that are collected within discrete regions with smoothing done based on Markov random field models for the neighborhood structure of the regions relative to each other. GeoBUGS also has the ability to smooth point-level data via a Bayesian implementation of kriging, but this capability is less well-developed and R may be a better choice for such data.
R packages for spatial data
The R project page has a number of other packages useful for spatial modelling of various sorts. A rough categorization of the areas for which the libraries might be useful is:
GIS software interface: spdep, GRASS, shapefiles
point process analysis: spatstat, splancs
smoothing: mgcv, fields, geoR, geoRglm, gstat, spBayes
areal data analysis: spdep
general spatial statistics: fields, VR (the spatial component), gstat, sgeostat
Last modified 9/8/2011.