MRC Biostatistics UnitThe following examples are in no particular order – please see BUGS resources on the web for a lot more examples provided by others.
1.4 only means that the example will not run in WinBUGS 1.3.
| Example name and description | Text file (either plain text or for decoding) | .odc File |
|---|---|---|
| Hips: integrated evidence synthesis and cost-effectiveness analysis. These programs accompany the paper: Spiegelhalter DJ and Best NB (2002) “Bayesian approaches to multiple sources of evidence and uncertainty in complex cost-effectiveness modelling” [PDF], |
||
| FunShapes: arbitrary constraints on functions of parameters Attractive illustration of the `step trick’. Use the correlations tool to draw the pictures. |
||
| Hepatitis – measurement error on covariates The example used in Chapter 2 of Gilks, Richardson and Spiegelhalter (1996) |
||
| Rats-drop – informative dropout The old rats example, illustrating the impact of an informative dropout assumption. |
||
| Camel – multivariate normal with missing data Tanner and Wong’s example of structured missing data which gives a bimodal posterior for the correlation. 1.4 only |
||
| Pines – Bayes factors using the Carlin and Chib approach The old example from the Classic BUGS Examples Vol 2, page 47. |
||
| Jama – use of the interp-lin function in radiocarbon calibration. Example provided by Andrew Millard. 1.4 only |
||
| StVeit – radiocarbon calibration with stratification. Example provided by Andrew Millard. 1.4 only |
||
| Eye-tracking – Dirichlet process prior for mixture of Poissons Adapted from Congdon (2001), Ex 6.27, to allow learning of baseline distribution.. 1.4 only |
||
| Pigweights – Non-wishart prior on MVN precision matrix Histogram smoothing using a multinomial-logistic model, with autoregressive random effects. Illustrates use of structured precision matrix (and also struictured covariance matrix,m but this is very slow due to need for matrix inversions). Adapted from Congdon (2001), Ex 5.9. 1.4 only |