nullspace-optimizer 1.1.2

Null space algorithm for nonlinear constrained optimization

Null space optimizer

nullspace_optimizer is a package implementing the null space algorithm for nonlinear constrained optimization.

Please cite the following references when using this source:

Feppon F., Allaire G. and Dapogny C. Null space gradient flows for constrained optimization with applications to shape optimization. 2020. ESAIM: COCV, 26 90 doi:10.1051/cocv/2020015 (Open Access). HAL preprint hal-01972915.

Feppon, F. Shape and topology optimization of multiphysics systems. 2019. Université Paris-Saclay. Thèse préparée à l'École polytechnique.

Installation

With pip

# Minimal version (no extra dependencies)
pip install nullspace_optimizer

# All extra dependencies including alternate QP solver, colored output and plotting features
pip install nullspace_optimizer[osqp,colored,matplotlib]

Manual installation

Add the package to the PYTHONPATH environment variable.

Running examples

A few examples of 2-d inequality constrained optimization are available in the `examples' folder. They can be run from command line with

python -m nullspace_optimizer.examples.ex0
python -m nullspace_optimizer.examples.ex1
python -m nullspace_optimizer.examples.ex2

and so on. All the examples can be run at once with

python -m nullspace_optimizer.examples.test_all

For instance, example ex1 solves the following optimization problem:

\newcommand{\<}{\leq}
\begin{aligned} \min_{(x_0,x_1)\in\mathbb{R}^2} & \quad (x_0+1)^2+(x_1+1)^2 \\
s.t. &\quad  \left\{ \begin{aligned} x_0^2+x_1^2-1 & \< 0\\
                             x_0+x_1-1 & \< 0 \\
                             -x_1-0.7 & \<0.
                             \end{aligned}\right.
\end{aligned}

Running python -m nullspace_optimizer.examples.ex1 should produce the following figure:

Requirements

Runs with python 3.6 and the following libraries:

• numpy (>=1.12.1)
• scipy (>=0.19.1)
• cvxopt (>=1.2.1)

Optional dependencies:

• osqp (>=0.6.1) (for an alternate QP solver instead of CVXOPT)
• colored (>=1.3.93) (for colored output)
• matplotlib (>=2.0.2) (for displaying figures while running examples)

Detailed documentation

A more detailed documentation is available on the DOC.md file of the git repository.

The full documentation of classes and methods is available in the python docstring of the source code.