Installation

This is a guide to install VRPSolverEasy package.

Note

See below how to install on Mac computers with Apple Silicon ARM processors.

Installation of free version

Free version uses COIN-OR CLP linear programming solver, embedded in the solver.

Requirements

You must have installed version of Python >=3.6

If you have an old version of package setuptools, it’s recommanded to upgrade version. You can run this command line on terminal:

python -m pip install --upgrade setuptools

Installation using pip

You can run this command line:

python -m pip install VRPSolverEasy

Installation using git

  • Download the package and extract it into a local directory

  • Move to this local directory and enter:

    python -m pip install .

Installation for Apple Silicon ARM processors

VRPSolverEasy comes with x86_64 BaPCod and VRPSolver libraries only. Therefore, Python for x86_64 architecture should be used with VRPSolverEasy. One way to do it is the following:

  • install x86_64 version of Homebrew (see for example here for instructions),

  • install python package via x86_64 Homebrew:

    arch -x86_64 /usr/local/homebrew/bin/brew install python

  • use /usr/local/homebrew/bin/python3 executable (or make an alias for it) to install VRPSolverEasy and to run Python code which uses VRPSolverEasy:

    /usr/local/homebrew/bin/python3 -m pip install VRPSolverEasy
/usr/local/homebrew/bin/python3 VRPSolverEasy/demos/CVRP.py -i VRPSolverEasy/demos/data/CVRP/A-n32-k5.vrp

Copy binaries

Once the package is installed you will need to request the Bapcod distribution here: https://bapcod.math.u-bordeaux.fr/ Once you have downloaded the distribution. You just have to go to the VRPSolverEasy folder and copy the system folder corresponding to your computer and copy it into the lib folder of the VRPSolverEasy python package. For example if your computer is a Mac you will copy and replace the Darwin folder, you will then have VRPSolverEasy/lib/Darwin.

Installation of academic version

Academic version of VRPSolverEasy uses commercial CPLEX MIP solver, which can be obtained for free for academic purposes. This version has improved performance and provides a built-in (slow) MIP-based heuristic, which is useful for finding feasible solutions in the absence of an external heuristic solver.

Warning

Academic version is also free but requires an e-mail address from an academic institution to download BaPCod source code. If CPLEX is not installed in the default directory, we invite you to add the path to the dynamic libraries in the environment variables.

To install this version, you must :

  1. Clone VRPSolverEasy in your local machine

  2. Install CPLEX

  3. Download Bapcod source code on its web page.

  4. Install Bapcod using installation instructions in the Bapcod user guide .

  5. Compile bapcod-shared library using this command :

cmake --build build --config Release --target bapcod-shared

This will produce the following shared library file:

"Bapcod-path/build/Bapcod/libbapcod-shared.so" // Linux
"Bapcod-path/build/Bapcod/libbapcod-shared.dylib" // MacOs
"Bapcod-path/build/Bapcod/Release/bapcod-shared.dll" // Windows

  • Replace the old library inside the VRPSolverEasy folder lib/your-system by the newly produced one

    • libbapcod-shared.so (Linux)

    • libbapcod-shared.dylib (MacOs)

    • bapcod-shared.dll (Windows)

  • re-install VRPSolverEasy using the command

    python -m pip install .

  • Make sure to indicate solver_name='CPLEX' when specifying VRPSolverEasy parameters.

  • If you use Windows system, you have to indicate the path to CPLEX by specifying cplex_path='<path>' in parameters.

  • If you want to use build-in heuristic, indicate also heuristic_used=True in parameters.