Dear ROOTERS / RooStats-ers / Roofitters,
I would like to know if there is already in ROOT/Roostat/Roofit a package that enable to do bayesian inference using MCMC integration such as JavaBayes, BUGS [1] or JAGS [2]; or if you think it is easy to implement it with available tools.
For example, given this simple problem, I would like to know if there is already something done:
suppose I have a process modelled by a binomial distribution, and I measure “k” successes given “n” trials, I would like to infer the probability “p” that describes the probability of success given a “user defined” prior on “p”.
This simple example can be done also analytically, but as soon as the model become more complex the computer help is somehow needed.
However if there is already something that solves this problem, then the more complex step can be easily performed.
Thanks,
Andrea
[1] mrc-bsu.cam.ac.uk/bugs/
[2] www-ice.iarc.fr/~martyn/software/jags/