/bayesian-free-energies

Repository to host our developmental bayesian free energy methods

Primary LanguageJupyter Notebook

Methods for the Bayesian estimation of free energies

UNDER DEVELOPEMENT

Repository to host tools for Bayesian estimate of free energies.

Bayesian adaptive mixture sampling (BAMS)

An online method to calculate the ratios of the normalizing constants in mixtures of distributions. The method is inspired by thinking of error estimates for free energies calculated with Self Adjusted Mixture Sampling. The result is a method that is similar in principle to the method by Bartel and Karplus.

Installation

This package is still in development, so to install type

python setup.py install

Dependencies

  • numpy
  • scipy
  • emcee
  • autograd (dependency to be removed soon)

Directories

bams/ Contains a tools for Bayesian adjusted mixture sampling

references/ Papers and references of prior Bayesian free energy work