README

This is an implementation of the Metropolis Hastings algorithm.
This is used for Bayesian sampling from a distribution that's typically multidimensional and can't be numerically integrated.
It utilizes Markov chains, which are ordered lists of stochastic (random) variables.
The Markov chain wanders around, only remembering the state of the previous iteration.
When the number of samples approaches infinity, the Markov chain will converge to the posterior distribution.

Usage:
Modify the posterior and proposal distribution functions in mh.py to suit your statistical model.

references:
"Pattern Recognition and Machine Learning" by Christopher Bishop
"Information Theory, Inference, and Learning Algorithms" by David Mackay
"Machine Learning: An Algorithmic Perspective" by Stephen Marsland