openMCMC

openMCMC is a package for constructing Bayesian models from distributional components, and then doing parameter estimation using Markov Chain Monte Carlo (MCMC) methods. The package supports a number of standard distributions used in Bayesian modelling (e.g. Normal, gamma, uniform), and a number of simple functional forms for the parameters of these distributions. For a model constructed in the toolbox, a number of different MCMC algorithms are available, including simple random walk Metropolis-Hastings, manifold MALA, exact samplers for conjugate distribution choices, and reversible-jump MCMC for parameters with an unknown dimensionality.

Installing openMCMC as a package

Suppose you want to use this openMCMC package in a different project. You can install it from PyPi through pip pip install openmcmc. Or you could clone the repository and install it from the source code. After activating the environment you want to install openMCMC in, open a terminal, move to the main openMCMC folder where pyproject.toml is located and run pip install ., optionally you can pass the -e flag is for editable mode. All the main options, info and settings for the package are found in the pyproject.toml file which sits in this repo as well.

Examples

For some examples on how to use this package please check out these Examples

Contribution

This project welcomes contributions and suggestions. If you have a suggestion that would make this better you can simply open an issue with a relevant title. Don't forget to give the project a star! Thanks again!

Contributor setup with uv

This repository now uses uv for contributor environment management.

1. Install uv by following the official instructions.
2. From the repository root (where pyproject.toml is located), create and sync the environment with contributor tools:

uv sync --group contributor --group dev

3. Run commands in the project environment using uv:

uv run pytest .
uv run black . --check
uv run isort . --check
uv run pydocstyle .
uv run pylint src/

For more details on contributing to this repository, see the Contributing guide.

Licensing

Distributed under the Apache License Version 2.0. See the license file for more information.