optlis

Code for the project Logistique d'Intervention suite aux catastrophes industrielles developed at the LITIS lab, Université Le Havre Normandie.

Requirements

To use this package you will need gcc, python >= 3.8 and CPLEX previously installed in your system.

Installation

Once you are inside the project folder, start by building the C code:

make

And export the lib directory as an environment variable using the absolute path:

export OPTLIS_LIB=path/to/optlis/lib

It is highly recommended to install this package inside a virtual environment.

Once you have downloaded or cloned the repository, install it by using (make sure you are inside a virtual environment if you opted to use it):

pip install -e .

To verify that the installation worked, test it by using:

pytest

You will also have to manually install CPLEX python interface if you want to run the integer models.

Usage

Generate the instance benchmark:

invoke export-benchmark path/to/inst-dir/

Solve the instance benchmark (all parameters pre configured):

invoke --help bulk-solve # Display help

invoke bulk-solve [solver] path/to/inst-dir/ -o path/to/sol-dir -l path/to/log-dir # static problem
invoke bulk-solve [solver] -d path/to/inst-dir/ -o path/to/sol-dir -l path/to/log-dir # dynamic problem

Replace [solver] by cplex or ils and define a [time limit] for cplex (in seconds) or a max number of evaluation calls for ils.