This repository contains some scripts to create LP-files for randomly generated multi-objective integer programming problems. Note the peculiar format of these files, the objective (as defined by the LP file format) is 0, the multiple objectives are all given as constraints. The last "x" constraints will have right-hand-side values of 1,2,3,...,x, where x is the number of objectives for your problem.
Currently, the only known (to the author) solver which handles such files is moip_aira.
If you want to support some other option, feel free to contact author.
The author would also very much like to be a part of any working group developing an open standard for multi-objective problem file formats.
Usage: ./generate_AP.py
where lpfile is the name of the file you wish to create, num-tasks is the number of tasks (equivalently number of workers) in the problem and num-objectives is the number of objectives you wish to minimise.
Example: ./generate_AP.py 5AP-50.lp 50 5
This will create an assignment problem with 50 tasks, 50 workers, and 5 objective functions.
Usage: ./generate_KP.py
where lpfile is the name of the file you wish to create, num-vars is the number of variables in the knapsack problem, and num-objectives is the number of objectives you wish to maximise.
Example: ./generate_KP.py KD4D-100.lp 100 4
This will create a knapsack problem with 100 variables and 4 objective functions.
Usage: ./generate_TSP.py
where lpfile is the name of the file you wish to create, num-cities is the number of cities in the problem, and num-objectives is the number of objectives you wish to maximise.
Example: ./generate_TSP.py TSP2_100.lp 100 2
This will generate a travelling salesman problem with 100 cities and 2 objective functions.