HMC Lab // Python#

These pages provide the documentation of the Python side of HMC Lab. Here you’ll find notebooks detailing how different concepts in Bayesian inference using Markov chain Monte Carlo work. In particular, extra attention is given on how to use the Hamiltonian Monte Carlo algorithm on (your) geophysical inverse problems.

HMC Lab is tested on Python 3.8+. In theory, it should work well on any system that has access to Conda.

Quickstart#

To download the repo, create a Conda environment, and install all dependencies, run the following:

> $ git clone https://github.com/larsgeb/hmclab.git
> $ cd hmclab
> $ conda env create -f environment.yml
> $ conda activate hmclab
> $ pip install -e .

The resulting Conda environment should be able to run all notebooks found in hmclab/notebooks. See the installation page for more detailed instructions.

Andrea Zunino, Andreas Fichtner, Lars Gebraad