notest1.cmd and notest2.cmd run the same changepoint problem with
different starting points. As pointed out in the WinBUGS manual.
This parameterization leads to very poor mixing, hence these examples
are not included in the test suite. You can run them with:

jags notest1.cmd
jags notest2.cmd

Thin from within R you can then analyze the results with the coda package

coda
x <- mcmc.list(read.coda("stag1"), read.coda("stag2"))
plot(x)

test1.cmd and test2.cmd run the changepoint problem with a better
parameterization
