proxsuite-nlp: a package for nonlinear optimization on manifolds

proxsuite-nlp is a C++ library, implementing a primal-dual augmented Lagrangian-type algorithm for nonlinear optimization on manifolds, as well as some modelling tools.

Installation

From Conda

From our channel

conda install -c simple-robotics proxsuite-nlp

From source

Clone this repo using

git clone [url-to-repo] --recursive

Create a build tree using CMake, build and install:

cd your/checkout/folder/
cmake -S . -B build
cmake --build build/ --config Release --target install

Dependencies

CMake (with the JRL CMake modules)
Eigen>=3.3.7
fmtlib>=9.1.0, <11
Boost>=1.71
(optional) eigenpy>=3.2.0 | conda (Python bindings)
(optional) pinocchio | conda
a C++-14 compliant compiler

Python dependencies:

numpy
matplotlib
typed-argument-parser
tqdm
meshcat-python

Notes

To build against a Conda environment, activate the environment and add export CMAKE_PREFIX_PATH=$CONDA_PREFIX before running CMake.
To build the documentation:
```
cd build/
make doc
```

Credits

The following people have been involved in the development of proxsuite-nlp and are warmly thanked for their contributions:

Wilson Jallet (LAAS-CNRS/Inria): main developer and manager of the project
Sarah El Kazdadi (Inria): linear algebra modules developer
Fabian Schramm (Inria): core developper
Joris Vaillant (Inria): core developer
Justin Carpentier (Inria): project coordinator
Nicolas Mansard (LAAS-CNRS): project coordinator

Acknowledgments

The development of proxsuite-nlp is actively supported by the Willow team @INRIA and the Gepetto team @LAAS-CNRS.