Sequential model-based optimization toolbox.
People: The scikit-optimize contributors
|Travis Status| |CircleCI Status| |binder| |gitter| |Zenodo DOI|
Scikit-Optimize, or ``skopt``, is a simple and efficient library to
minimize (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
.. figure:: https://github.com/scikit-optimize/scikit-optimize/blob/master/media/bo-objective.png
:alt: Approximated objective
Approximated objective function after 50 iterations of ``gp_minimize``.
Plot made using ``skopt.plots.plot_objective``.
- Static documentation - `Static
- Example notebooks - can be found in the
`examples directory <https://github.com/scikit-optimize/scikit-optimize/tree/master/examples>`_.
- Issue tracker -
- Releases - https://pypi.python.org/pypi/scikit-optimize
The latest released version of scikit-optimize is v0.5.1, which you can install
pip install scikit-optimize
In addition there is a `conda-forge <https://conda-forge.org/>`_ package
conda install -c conda-forge scikit-optimize
Using conda-forge is probably the easiest way to install scikit-optimize on
Find the minimum of the noisy function ``f(x)`` over the range
``-2 < x < 2`` with ``skopt``:
.. code:: python
import numpy as np
from skopt import gp_minimize
return (np.sin(5 * x) * (1 - np.tanh(x ** 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``
.. code:: python
from skopt import Optimizer
opt = Optimizer([(-2.0, 2.0)])
for i in range(20):
suggested = opt.ask()
y = f(suggested)
print(\'iteration:\', i, suggested, y)
Read our `introduction to bayesian
and the other
The library is still experimental and under heavy development. Checkout
for the plans for the next release or look at some `easy
to get started contributing.
The development version can be installed through:
git clone https://github.com/scikit-optimize/scikit-optimize.git
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 <https://docs.pytest.org/en/latest/mark.html>`__. If a tests runs longer than 1 second, it is marked as slow, else as fast.
All contributors are welcome!
Making a Release
The release procedure is almost completely automated. By tagging a new release
travis will build all required packages and push them to PyPI. To make a release
create a new issue and work through the following checklist:
* update the version tag in ``setup.py``
* update the version tag in ``__init__.py``
* update the version tag mentioned in the README
* check if the dependencies in ``setup.py`` are valid or need unpinning
* check that the ``CHANGELOG.md`` is up to date
* did the last build of master succeed?
* create a `new release <https://github.com/scikit-optimize/scikit-optimize/releases>`__
* ping `conda-forge <https://github.com/conda-forge/scikit-optimize-feedstock>`__
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 in
``setup.py`` and ``__init__.py``. Mark a release candidate as a \"pre-release\"
on GitHub when you tag it.
Feel free to `get in touch <mailto:email@example.com>`_ if you need commercial
support or would like to sponsor development. Resources go towards paying
for additional work by seasoned engineers and researchers.
Made possible by
The scikit-optimize project was made possible with the support of
.. image:: https://avatars1.githubusercontent.com/u/18165687?v=4&s=128
:alt: Wild Tree Tech
.. image:: https://i.imgur.com/lgxboT5.jpg
:alt: NYU Center for Data Science
.. image:: https://i.imgur.com/V1VSIvj.jpg
.. image:: https://i.imgur.com/3enQ6S8.jpg
:alt: 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.
.. |Travis Status| image:: https://travis-ci.org/scikit-optimize/scikit-optimize.svg?branch=master
.. |CircleCI Status| image:: https://circleci.com/gh/scikit-optimize/scikit-optimize/tree/master.svg?style=shield&circle-token=:circle-token
.. |Logo| image:: https://avatars2.githubusercontent.com/u/18578550?v=4&s=80
.. |binder| image:: https://mybinder.org/badge.svg
.. |gitter| image:: https://badges.gitter.im/scikit-optimize/scikit-optimize.svg
.. |Zenodo DOI| image:: https://zenodo.org/badge/54340642.svg
You can download the latest distribution from PyPI here: http://pypi.python.org/pypi/scikit-optimize
You can install scikit-optimize for yourself from the terminal by running:
pip install --user scikit-optimize
If you want to install it for all users on your machine, do:
pip install scikit-optimize
On Linux, do sudo pip install scikit-optimize
If you don't yet have the pip tool, you can get it following
This package was discovered in PyPI.