/scikit-optimize

Sequential model-based optimization with a `scipy.optimize` interface

Primary LanguagePythonBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

Logo

pypi conda CI Status binder codecov Zenodo DOI

Scikit-Optimize

Scikit-Optimize, or skopt, is a simple and efficient library for optimizing (very) expensive and noisy black-box functions. It implements several methods for sequential model-based optimization. skopt aims to be accessible and easy to use in many contexts.

The library is built on top of NumPy, SciPy, and Scikit-Learn.

We do not perform gradient-based optimization. For gradient-based optimization algorithms look at scipy.optimize here.

Approximated objective

Approximated objective function after 50 iterations of gp_minimize. Plot made using skopt.plots.plot_objective.

Maintaining the codebase

This repo is a copy of the original repositoy at https://github.com/scikit-optimize/scikit-optimize/. As the original repo is now in read-only mode, i decided to continue the development on it on my own. I still have credentials for pypi, so I will publish new releases at https://pypi.org/project/scikit-optimize/. I did my best to include all open PR since 2021 in the new release of scikit-optimize 0.10.

https://scikit-optimize.github.io/ has been moved to http://scikit-optimize.readthedocs.io/.

Important links

Install

scikit-optimize requires

  • Python >= 3.8
  • NumPy (>= 1.20.3)
  • SciPy (>= 0.19.1)
  • joblib (>= 0.11)
  • scikit-learn >= 1.0.0
  • matplotlib >= 2.0.0

You can install the latest release with:

pip install scikit-optimize

This installs the essentials. To install plotting functionality, you can instead do:

pip install 'scikit-optimize[plots]'

This will additionally install Matplotlib.

If you're using Anaconda platform, there is a conda-forge package of scikit-optimize:

conda install -c conda-forge scikit-optimize

Using conda-forge is probably the easiest way to install scikit-optimize on Windows.

Getting started

Find the minimum of the noisy function f(x) over the range -2 < x < 2 with skopt:

import numpy as np
from skopt import gp_minimize

def f(x):
    return (np.sin(5 * x[0]) * (1 - np.tanh(x[0] ** 2)) +
            np.random.randn() * 0.1)

res = gp_minimize(f, [(-2.0, 2.0)])

For more control over the optimization loop you can use the skopt.Optimizer class:

from skopt import Optimizer

opt = Optimizer([(-2.0, 2.0)])

for i in range(20):
    suggested = opt.ask()
    y = f(suggested)
    opt.tell(suggested, y)
    print('iteration:', i, suggested, y)

Read our introduction to bayesian optimization and the other examples.

Development

The library is still experimental and under development. Checkout the next milestone for the plans for the next release or look at some easy issues to get started contributing.

The development version can be installed through:

git clone https://github.com/holgern/scikit-optimize.git
cd scikit-optimize
pip install -e .

Run all tests by executing pytest in the top level directory.

To only run the subset of tests with short run time, you can use pytest -m 'fast_test' (pytest -m 'slow_test' is also possible). To exclude all slow running tests try pytest -m 'not slow_test'.

This is implemented using pytest attributes. If a tests runs longer than 1 second, it is marked as slow, else as fast.

All contributors are welcome!

Pre-commit-config

Installation

pip install pre-commit

Using homebrew

brew install pre-commit

pre-commit --version
pre-commit 2.10.0

Install the git hook scripts

pre-commit install

Run against all the files

pre-commit run --all-files
pre-commit run --show-diff-on-failure --color=always --all-files

Update package rev in pre-commit yaml

pre-commit autoupdate
pre-commit run --show-diff-on-failure --color=always --all-files

Making a Release

The release procedure is almost completely automated. By tagging a new release, CI will build all required packages and push them to PyPI. To make a release, create a new issue and work through the following checklist:

Before making a release, we usually create a release candidate. If the next release is v0.X, then the release candidate should be tagged v0.Xrc1. Mark the release candidate as a "pre-release" on GitHub when you tag it.

Made possible by

The scikit-optimize project was made possible with the support of

Wild Tree Tech NYU Center for Data Science NSF Northrop Grumman

If your employer allows you to work on scikit-optimize during the day and would like recognition, feel free to add them to the "Made possible by" list.