cyipopt-wheels 1.7.0.dev0

A Cython wrapper to the IPOPT optimization package

Ipopt (Interior Point OPTimizer, pronounced eye-pea-opt) is a software package for large-scale nonlinear optimization. Ipopt is available from the COIN-OR initiative, under the Eclipse Public License (EPL).

cyipopt is a Python wrapper around Ipopt. It enables using Ipopt from the comfort of the Python programming language.

Status

Anaconda
PyPI
Read the Docs

History

This repository was forked in 2016 from https://bitbucket.org/amitibo/cyipopt and is now considered the primary repository. The fork includes a SciPy-style interface and ability to handle exceptions in the callback functions.

As of version 1.1.0 (2021-09-07), the distribution is released under the name “cyipopt” on PyPi (https://pypi.org/project/cyipopt). Before version 1.1.0, it was released under the name “ipopt” (https://pypi.org/project/ipopt).

Installation

We recommend using conda to install cyipopt on Linux, Mac, and Windows:

conda install -c conda-forge cyipopt

Other installation options are present in the documentation.

Building manylinux wheels

manylinux wheels can be built for a tagged version (GIT_TAG below) of cyipopt via docker by running (while in the root of this repo):

docker run -v $(pwd):/wheels --rm --platform=linux/amd64 quay.io/pypa/manylinux_2_28_x86_64 bash /wheels/build_manylinux_wheels.sh GIT_TAG

for linux/amd64 and:

docker run -v $(pwd):/wheels --rm --platform=linux/aarch64 quay.io/pypa/manylinux_2_28_aarch64 bash /wheels/build_manylinux_wheels.sh GIT_TAG

for linux/aarch64 platforms. Built wheels appear at the folder the command was executed from.

Warning

Docker supports emulating non-native platforms to e.g. produce ARM binaries from an AMD64 host. However this can be quite slow (~1h for our case).

Using in Pyodide (browser/WASM)

cyipopt can run in the browser via Pyodide, with Ipopt compiled to WebAssembly. Pre-built wheels are available from GitHub releases.

Install with micropip in a Pyodide environment:

import micropip
await micropip.install(
    "https://github.com/mechmotum/cyipopt/releases/download/VERSION/cyipopt-VERSION-cp312-cp312-pyodide_2024_0_wasm32.whl"
)

Then use it normally:

import numpy as np
from cyipopt import minimize_ipopt
from scipy.optimize import rosen, rosen_der

x0 = np.array([1.3, 0.7, 0.8, 1.9, 1.2])
res = minimize_ipopt(rosen, x0, jac=rosen_der)
print(res.x)

Full example: see the grecov calculator for a working web app that uses cyipopt in Pyodide to solve nonlinear optimization problems entirely in the browser.

License

cyipopt is open-source code released under the EPL license, see the LICENSE file.

Contributing

For bug reports, feature requests, comments, patches use the GitHub issue tracker and/or pull request system.

Contributors (made with contrib.rocks):