bojaxns 1.1.1

Bayesian Optimisation with JAXNS

Main Status:

Develop Status:

Mission: To make advanced Bayesian Optimisation easy.

What is it?

Bojaxns is:

1. a Bayesian Optimisation package for easily performing advanced non-myopic Bayesian optimisation.
2. using JAXNS under the hood to marginalise over multiple models.
3. using multi-step lookahead to plan out your next step.
4. available for academic use and non-commercial use (without permission) read the license.

Documentation

For examples, check out the documentation (still in progress).

Install

Notes:

1. Bojaxns requires >= Python 3.9.
2. It is always highly recommended to use a unique virtual environment for each project. To use miniconda, have it installed, and run

# To create a new env, if necessary
conda create -n bojaxns_py python=3.11
conda activate bojaxns_py

For end users

Install directly from PyPi,

pip install bojaxns

For development

Clone repo git clone https://www.github.com/JoshuaAlbert/bojaxns.git, and install:

cd bojaxns
pip install -r requirements.txt
pip install -r requirements-tests.txt
pip install .

Change Log

15 Jan, 2024 -- Bojaxns 1.1.0/1 released. Bumped to jaxns 2.4.3/4. 20 July, 2023 -- Bojaxns 1.0.0 released