Join our forum for any questions related to the solver!
The OSQP (Operator Splitting Quadratic Program) solver is a numerical optimization package for solving convex quadratic programs in the form
where x is the optimization variable and P ∈ S+n a positive semidefinite matrix.
Code available on GitHub.
Citing OSQP
If you are using OSQP for your work, we encourage you to
Cite the related papers <citing>
- Put a star on GitHub
We are looking forward to hearing your success stories with OSQP! Please share them with us.
Features
- Efficient
It uses a custom ADMM-based first-order method requiring only a single matrix factorization in the setup phase. All the other operations are extremely cheap. It also implements custom sparse linear algebra routines exploiting structures in the problem data.
- Robust
The algorithm is absolutely division free after the setup and it requires no assumptions on problem data (the problem only needs to be convex). It just works!
- Detects primal / dual infeasible problems
When the problem is primal or dual infeasible, OSQP detects it. It is the first available QP solver based on first-order methods able to do so.
- Embeddable
It has an easy interface to generate customized embeddable C code with no memory manager required.
- Library-free
It requires no external library to run. Only the setup phase requires the AMD and SparseLDL from Timothy A. Davis that are already included in the sources.
- Efficiently warm started
It can be easily warm-started and the matrix factorization can be cached to solve parametrized problems extremely efficiently.
- Interfaces
It can be interfaced to C, C++, Python and Matlab.
License
OSQP is distributed under the Apache 2.0 License
Credits
The following people have been involved in the development of OSQP:
- Bartolomeo Stellato (University of Oxford): main development
- Goran Banjac (University of Oxford): main development
- Nicholas Moehle (Stanford University): methods, maths, and code generation
- Paul Goulart (University of Oxford): methods, maths, and Matlab interface
- Alberto Bemporad (IMT Lucca): methods and maths
- Stephen Boyd (Stanford University): methods and maths
Bug reports and support
Please report any issues via the Github issue tracker. All types of issues are welcome including bug reports, documentation typos, feature requests and so on.
Numerical benchmarks
Numerical benchmarks against other solvers are available here.
solver/index installation/index interfaces/index parsers/index codegen/index examples/index contributing/index citing/index