mcmc-webdemo (GPL 2 or higher)
Markov Chain Monte Carlo (MCMC) is an exciting simulation and integration
technique which uses a randomly evolving process to explore a
specified target distribution. The original idea is due to Metropolis
et al (1953), who aimed to sample the Bolzmann distribution.
If your browser supports Java applets, you can try out and observe several MCMC algorithms in action right here.
- webdemo.tar.gz (MD5)
- This is the Java source code for the Markov Chain Monte Carlo demonstration applets
which you can find elsewhere on this site. The code is provided here AS IS,
and is not currently being actively developed.
If you just want to run the applets locally, it is easiest
if you download instead the compiled jar file, and
an appropriate Java Runtime Environment (JRE, version 1.1 or above).
Then follow the instructions in the table below.
| To run these algorithms... | you should type... |
| classical Metropolis-Hastings | jre -cp MCMC.jar ClassMHFrame |
| M-H roadmap | jre -cp MCMC.jar RoadMapFrame |
| simulated annealing | jre -cp MCMC.jar SimAFrame |
| field coupler | jre -cp MCMC.jar FCFrame |
|
