vOptGeneric: part of vOptSolver for non-structured problems

vOptSolver is a solver of multiobjective linear optimization problems (MOCO, MOIP, MOMILP, MOLP). This repository concerns vOptGeneric, the part of vOptSolver devoted to multiobjective non-structured problems (currently available: 2-IP, p-ILP). With vOptGeneric, the problem is expressed using JuMP algebraic language extended to multiple objectives. vOptGeneric runs on macOS, linux-ubuntu, and windows (local use), also on JuliaBox (distant use).

We suppose you are familiar with vOptSolver; if not, read first this presentation.

Instructions

For a local use, a working version of:

• Julia must be ready; instructions for the installation are available here
• your favorite MILP solver must be ready (GLPK is suggested); instructions for the installation are available here

Run Julia

On linux or in the cloud (JuliaBox):

• open a console on your computer or in the cloud
• when the prompt is ready, type in the console julia

On macOS:

• locate the application julia and
• click on the icon, the julia console comes to the screen

Installation Instructions

Before your first local or distant use,

1. run Julia and when the terminal is ready with the prompt julia on screen,
2. add as follow the mandatory packages to your Julia distribution:

julia> using Pkg
julia> Pkg.add("vOptGeneric")
julia> Pkg.add("GLPK") ; Pkg.add("GLPKMathProgInterface")

That's all folk; at this point, vOptGeneric is properly installed.

Usage Instructions

When vOptGeneric is properly installed,

1. run Julia and when the terminal is ready with the prompt julia on screen,
2. invoke vOptGeneric and the MILP solver to activate in typing in the console:

julia> using vOptGeneric
julia> using GLPK ; using GLPKMathProgInterface

vOptGeneric is ready. See examples for further informations and have fun with the solver!

Problems available

Problem	Description	Output	Method	Parameter (if required)	Name
2-ILP	bi-objective Integer Linear Program	Y_N	:epsilon	step = realValue	ϵ-constraint
2-ILP	bi-objective Integer Linear Program	Y_N	:chalmet or :Chalmet	step = realValue	Chalmet
2-ILP	bi-objective Integer Linear Program	Y_{SN}	:dicho or :dichotomy	(none)	Aneja & Nair
p-ILP	multi-objective Integer Linear Program	Y_{lex}	:lex or :lexico	(none)	Lexicographic

Examples

The folder examples provides (1) source code of problems ready to be solved and (2) selected datafiles into different formats.

Limitations

No special limitation; the solving strength of vOptGeneric is currently provided by the MILP solver (GLPK, Clp/Cbc, CPLEX, GUROBI, etc.) invoked.