---

Computing/software for this course

---

R
To install R and to find documentation and packages etc. go to the CRAN website . Here are some useful R links: general R links .
To download any R packages, you simply need to use "Install packages" under the "Packages" menu in the R gui or from the R command line type: install.packages(package name within quotes)
Note: You can download and install R for Windows, Mac and Linux/Unix.

---

R for spatial:
The geoR package in R is useful for a variety of basic spatial functions. More on this can be found at Ribeiro's webpage geoR documentation
The geoRglm package.
The maps package in R is useful for mapping/plotting functions:
Important R spatial libraries: spatial spatstat (for point process data).

---

WINBUGS:
WINBUGS software which automates the process of running/fitting Bayesian models via Markov chain Monte Carlo. The current version is easily downloaded here.
GeoBUGS is now part of the latest version of WINBUGS (1.4.1 onwards): It is a useful package for spatial modelers. For more information try the GeoBUGS page .
Note: WINBUGS only works on Windows (hence the name) although OpenBUGS works on Linux as well.

---

WINBUGS and R:
If the thought of dealing with R and WINBUGS separately bothers you, you may try to combine the two with the R package called BRUGS which can be downloaded from here: BRUGS . This accesses OpenBUGS which is an open source version of WINBUGS. Note: BRUGS only works for Windows although OpenBUGS works on Linux as well.
I expect the software to be stable (since the authors are also the WINBUGS authors) but I have no experience with BRUGS or OpenBUGS myself. Please share any positives/negatives with me and the class if you use it.
Also see Prof.Carlin's page: BRugs

---

spBayes:
This is a brand new R package developed at the University of Minnesota (the authors include two of the authors of our course textbook): SpBayes
This promises to do much more efficient and flexible fitting of Bayesian spatial models. It currently fits geostatistics models but future releases will include models for areal data as well.

---