lccp-bnp

The repository contains the branch-and-price algorithm and instance-wise results for the paper "Branch and Price for the Length-Constrained Cycle Partition Problem", you can find the preprint here.

Installation

The easiest way to get started is to create a new conda environment using the env.yml file.

conda env create -f env.yml

Then activate the newly created environment

conda activate lccp

To compile the pricer module

maturin develop --release -m rs_pricing/Cargo.toml

Run

Example for running the branch and price algorithm on an instance

python main.py data/t84_eil51.npcc

All improvements techniques are enabled by default, to run with specific techniques, run the following to display all options

python main.py --help

Instance-wise results

Refer to results/README.md for the instance-wise results.