This repository provides the implementation of benchmark functions for Bayesian optimization. The details of benchmark functions can be found in these notes.
We recommend installing it with virtualenv.
You can choose one of three installation options.
- Using PyPI repository (for user installation)
To install the released version in PyPI repository, command it.
pip install bayeso-benchmarks- Using source code (for developer installation)
To install bayeso-benchmarks from source code, command the following in the bayeso-benchmarks root.
pip install .- Using source code (for editable development mode)
To use editable development mode, command the following in the bayeso-benchmarks root.
pip install -e .If you want to install the packages required for development, you can simply add [dev].
For example, pip install .[dev] or pip install -e .[dev].
- Uninstallation
If you would like to uninstall bayeso-benchmarks, command it.
pip uninstall bayeso-benchmarksA simple example on Branin function is shown below.
from bayeso_benchmarks import Branin
obj_fun = Branin()
bounds = obj_fun.get_bounds()
X = obj_fun.sample_uniform(100)
Y = obj_fun.output(X)
Y_noise = obj_fun.output_gaussian_noise(X)@misc{KimJ2023software,
author={Kim, Jungtaek},
title={{BayesO Benchmarks}: Benchmark Functions for {Bayesian} Optimization},
doi={10.5281/zenodo.7577330},
url={https://github.com/jungtaekkim/bayeso-benchmarks},
howpublished={\url{https://doi.org/10.5281/zenodo.7577330}},
year={2023}
}