Four MIP primal heuristics (FeasibilityJump, FPR, LocalMIP, Scylla) implemented and benchmarked inside HiGHS
Optees is an open-source desktop toolkit for optimization: LP (SciPy/HiGHS), MILP (OR-Tools CP-SAT), and 0/1 knapsack. Clean Architecture, TDD, dataset adapters (LPnetlib, MIPLIB, Burkardt). Minimal APIs, extensible and testable.
Benchmarks ojAlgo's pure Java LP, QP and MIP solvers against HiGHS, OR-Tools, Clarabel, CPLEX, Gurobi, MOSEK and others on Netlib, Maros-Meszaros and MIPLIB.
Compare objective values for three formulations on the same RCPSP instance.
Script that download instances from MIPLIB base on a csv
Python framework for experimenting with transformations of Mixed Integer Programming (MIP) problems. Designed for research and benchmarking, it includes tools for reordering, canonical form generation, and block decomposition. Compatible with external solvers such as Gurobi, and supports evaluation on MIPLIB benchmark instances.
This repository provides utilities for converting `.mps` and `.lp` models into IBM ILOG CPLEX Optimization Studio's OPL format, which consists of a `.mod` model file and a `.dat` data file.
1,100+ LP, QP and MIP test models (MIPLIB, QPLIB, Netlib, Maros-Meszaros, Meszaros) as a Maven artifact, for testing and benchmarking optimization solvers.
To associate your repository with the miplib topic, visit your repo's landing page and select "manage topics."