PaPILO - Parallel Presolve for Integer and Linear Optimization. PaPILO, a C++14-based software package, provides parallel presolve routines for (mixed integer) linear programming problems. The routines are implemented using templates which allows switching to higher precision or rational arithmetic using the boost multiprecision package. Additionally to the distribution here under the LGPLv3, PaPILO is also distributed as part of the SCIP Optimization Suite which is available under https://scipopt.org/. PaPILO can be used as a header-based library and also provides an executable. Using the executable it is possible to presolve and postsolve MILP instances based on files. Additionally, PaPILO can be linked to SCIP, SoPlex, and HiGHS (https://github.com/ERGO-Code/HiGHS) solvers and act as a frontend. In this setting PaPILO passes the presolved problem to those solvers and applies the postsolve step to the optimal solution. When PaPILO is compiled as part of the SCIP Optimization Suite linking of SoPlex and SCIP solvers is performed automatically.
References in zbMATH (referenced in 1 article )
Showing result 1 of 1.
- Eifler, Leon; Gleixner, Ambros: A computational status update for exact rational mixed integer programming (2021)