Lachlan Smith, Joshua, Lachie

Run

This repository can be run using

python main.py

To view all functionality we recommend running the below command

python main.py -i 215 -p -h -v --plot

Usage

We've provided command line arguments to run the bin packing problem in various different ways.

Instances

To run a different instance of the problem use the --instance argument (default 1).

python main.py --instance 215

Preprocess

To run the preprocessor on the solver use the --preprocess argument.

python main.py --instance 215 --preprocess

The preprocess argument can be made to run with more granular control, set --preprocess to 010 to only assign large items to the solver, and not incompatible or conflicting items, etc

python main.py --instance 215 --preprocess 010

Subproblem

To set the parameters of subproblem use the --subproblem argument. The subproblem has three parameters which can be set to 0 off or 1 on.

• try the heuristic guess for each subproblem
• use minimize the bin for each subproblem
• use ORTools instead of gurboi as the constraint program solver

python main.py --instance 215 --subproblem 010

If no argument is provided the subproblem is run with 110 (use heuristic, minimize bins and use gurobi constraint solver)

Verbose

To see debug output that includes preprocess assignments and real-time feasibility cuts provide the --verbose argument.

python main.py -i 215 -p --verbose

To see gurobi output instead of real-time cuts set --verbose to 2.

python main.py -i 215 -p -h --verbose 2

Heuristic

To run the problem with the heuristic provide the --heuristic argument.

python main.py --instance 0 --heuristic

Plot

To plot the solution provide the --plot argument (default grid).

python main.py --plot

To plot the heuristic solution set --plot to heuristic

python main.py --heuristic --plot heuristic.

To plot the solver solution the same as heuristic set --plot to box

python main.py --plot box.

Extract

To extract the solutions x,y placements but not plot the solution provide the --extract argument.

python main.py --extract

Benchmarks

To run the benchmark tests for all 500 instances use

python benchmark.py

To get stats of the bench test use

python stats.py