BayesO Benchmarks: Benchmark Functions for Bayesian Optimization

This repository provides the implementation of benchmark functions for Bayesian optimization. The details of benchmark functions can be found in these notes.

• https://bayeso.org

Installation

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-benchmarks

Simple Example

A 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)

Citation

@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}
}

License

MIT License