Global Fit Optimisation Routine

[pathfinder49]

We may want a routine for global fit optimisation of binomial (or other) distributed data.

This should give the global optimum parameter estimates and their covariances. Ideally, it should be capable of producing a plot of P(parameters|data).

Broken out from this issue

[pathfinder49]

Markov Chain Monte Carlo might me the appropreate method for this.

There are library implementations in PyMC and PyStan.
A simple introduction is here
A discussion can be found here

[hartytp]

FWIW I'm not convinced this would be a useful addition to oitg. The aim should be to cover 95% of cases well using robust simple code, and leave the tricky cases to users to fit offline. Writing a toolkit that can robustly do global optimization without significant amounts of per-case user input is a very hard problem.