Software download

If you will be bringing a laptop to the course, please could you download and install as much of the following software as possible beforehand. The intention is to do the practicals in small groups on shared laptops.

Do not worry if there are any problems with the software, as there will be an opportunity for installation during the course.

Course practicals

Practicals for the course are in a zip file and can be downloaded from here (right click on the link and choose Save Target/Link as...). Unzip these to C:\, this will create a directory C:\winbugs-hecourse with the practicals in. If you unzip them to somewhere else you will need to change some of the paths in the example scripts.

WinBUGS

This program is for Markov chain Monte Carlo (MCMC) analysis. It can be downloaded from the WinBUGS website:

http://www.mrc-bsu.cam.ac.uk/bugs/winbugs/contents.shtml which contains instructions for installation.
Please remember to install the Immortality Key for unrestricted use, or else the program won't work fully!

If you have not used WinBUGS before, it would be very helpful if you ran through the tutorial provided in the online manual (packaged with the program).

R and R packages

R is a general-purpose statistical program, a free equivalent of S-Plus. The base R package can be downloaded from the UK mirror of the R web site: http://www.stats.bris.ac.uk/R/bin/windows/base/ (select "Setup program").

After installation, and while remaining online, you should go to the Packages menu, select Install Packages, and select and install R2WinBUGS, BRugs and coda. These are packages for controlling the WinBUGS graphical interface from R (R2WinBUGS), using the BUGS "engine" to do MCMC sampling entirely within R (BRugs) and for processing MCMC output from WinBUGS (coda). If you can't find BRugs, make sure CRAN (extras) is selected on Packages->Select repositories.

A paper about R2WinBUGS can be downloaded from here. Documentation for all R packages can be read from within R: select the Help menu, select HTML help and browse to Packages.

Another R package called costeff will also be used. This is not in the repository of official R packages, but can be downloaded here (right click on the link and choose Save Target/Link as...). This is installed from within R, using the menus;
Packages -> Install packages(s) from local zip files ...