SLEQP: An open‐source package for nonlinear programming

Abstract

We present SLEQP, an open source C‐package consisting of an active‐set method capable of solving large‐scale nonlinear programming (NLP) problems based on successive linear programming and equality constrained quadratic programming techniques. The method includes a feasibility restoration phase and second‐order corrections (SOCs), achieving robust practical performance. It has also been adapted for the special cases of unconstrained optimization, box‐constrained problems, and nonlinear least squares problems and contains interfaces to both Python and MATLAB. To demonstrate the performance of the package, we perform a computational study based on the well‐known CUTest suite of NLP problems.