ArnabKar / PyGauss

Companion package of the review paper entitled 'High-dimensional Gaussian sampling: A review and a unifying approach based on a stochastic proximal point algorithm' by Maxime Vono et al.

PyGauss: High-dimensional Gaussian sampling with Python

Introduction

Efficient sampling from a high-dimensional Gaussian distribution is an old but high-stake issue. In past years, multiple methods have been proposed from different communities to tackle this difficult sampling task ranging from iterative numerical linear algebra to Markov chain Monte Carlo (MCMC) approaches. PyGauss is a Python library that puts together all exact and approximate sampling algorithms for high-dimensional Gaussian sampling. It stands for the companion package of the review paper entitled High-dimensional Gaussian sampling: A review and a unifying approach based on a stochastic proximal point algorithm, publicy available on arXiv and accepted for publication in SIAM Review.

How to cite this work

If you use the PyGauss toolbox, please consider citing it with this piece of BibTeX:

@article{Vono_SIREV_2022,
author = {Maxime Vono and Nicolas Dobigeon and Pierre Chainais},
title = {High-dimensional Gaussian sampling: A review and a unifying approach based on a stochastic proximal point algorithm},
year = {2022},
journal = {SIAM Review}
url = {https://arxiv.org/abs/2010.01510}
}

Installation

PyGauss works with Python 3.6+.

Dependencies

This project depends on the following libraries and modules:

• NumPy
• SciPy
• math

The following dependencies are optional, and unlock extra functionality if installed:

• Sphinx to modify and rebuild the documentation

Installation instructions

1. If you have a GitHub account please consider forking PyGauss and use git to clone your copy of the repository

   cd <directory_of_your_choice>
   git clone https://github.com/<username>/PyGauss.git

2. If you only use git, clone this repository

   cd <directory_of_your_choice>
   git clone https://github.com/mvono/PyGauss.git

3. Otherwise simply download the project

4. In any case, install the project with

   cd PyGauss
   pip install .

How to use it

The main PyGauss documentation is available online at http://pygauss-gaussian-sampling.readthedocs.io. You can also find examples using PyGauss in the Jupyter notebook reproducing the results of the companion paper at https://github.com/mvono/PyGauss/notebooks/.

Building the documentation

The documentation is generated locally with Sphinx and then built online by ReadTheDocs. If you wish to contribute to the documentation or just play with it locally, you can install the necessary dependencies and then:

• Generate the docs locally

  cd PyGauss/docs
  make html

• Open the local HTML version of the documentation located at PyGauss/docs/build/html/index.html

  open build/html/index.html