If this software was used for academic purposes, please cite our paper with the below information:
@article{turner2023context, title={{A Context-Aware Cutting Plane Selection Algorithm for Mixed-Integer Programming}}, author={Turner, Mark and Berthold, Timo and Besan{\c{c}}on, Mathieu}, journal={arXiv preprint arXiv:2307.07322}, year={2023} }
Requirements: Python 3.9 / Debian 11 (Also tested with 3.11). We use SLURM https://slurm.schedmd.com/overview.html as a job manager. All calls go through a central function however, and in theory SLURM could be replaced by python's default multiprocessing package.
Run the bash script init_venv. If you don't use bash, configure the shebang (first line of the script) to be your shell interpreter.
./init_venv
After installing the virtual environment, make sure to always activate it with the script printed beneath. This is so your python path is appended and files at different directory levels can import from each other.
source ./set_venv
Now go and install SCIP from (https://www.scipopt.org/index.php#download / https://github.com/scipopt/scip)
For Ubuntu / Debian, the .sh installer is the easiest choice if you don't want to configure it yourself).
The latest release version of SCIP is required for this project, or the development branch mt/ensemble_cutsel.
You can test if SCIP is installed by locating /bin/scip and calling it through the command line. SCIP should hopefully open.
One then needs to install PySCIPOpt https://github.com/scipopt/PySCIPOpt. I would recommend following the INSTALL.md guide. Make sure to have set your environment variable pointing to SCIP! You can test if this has been properly installed by running one of the tests, or by trying to import Model. This research was done using PySCIPOpt 4.3.0.
We use Nohup https://en.wikipedia.org/wiki/Nohup to run all of our jobs and to capture output of the main function calls. It also allows jobs to be started through a SHH connection. Using this is not necessary however, so feel free to use whatever software you prefer. An example call to redirect output to nohup/nohup.out and to run the process in the background would be
nohup python dir/example.py > nohup/nohup.out &
For our experiments we used MIPLIB 2017, StrIPLib, and SNDlib-MIPS. A combination of these instances were used for training, where we evaluated our final model on the MIPLIB 2017 benchmark set.
To select instances in a similar to us, please download all the appropriate instances from the respective websites (or feel free to use instances that you have on hand). Place all instances in a single directory. We reduce the amount of instances for training by running:
python Slurm/generate_instance_subset.py instance_dir solution_dir instance_subset_dir results_dir outfiles_dir
After the instances have been pre-filtered, we can now get their feature embeddings to perform a maximally diverse selection. To do this run
python scripts/get_instance_embeddings.py instance_dir feature_dir
Then to actually select a maximally diverse subset run
python scripts/get_diverse_instance_subset.py instance_dir selected_instance_dir feature_dir plot_stuff
here plot_stuff is an argument about whether you'd like the end result visualised. You can feel free to manually
change between PCA or 2-SNE as the embedding options for visualisation. (You have to manually gzip all .sol files at
this point)
Now we can train a SMAC model. To do this run
nohup python Slurm/main.py bool_1 seed instance_dir solution_dir results_dir outfiles_dir > nohup/test.out &
bool_1 here is whether ot not you want to first generate default runs to compare against (This is an option in
case you already have generated your default runs and don't want to constantly reproduce identical results).
This main function trains a SMAC model, with output being redirected to nohup/test.out. In that file you will find
the training progression along with the best parameter choice found at the end of training as well as how it
performed. For more information, one can view the SMAC run history file that was also created.
To test a given configuration on some test set, you can run the following:
nohup python Slurm/test_configuration.py config_file bool instance_dir solution_dir results_dir outfiles_dir > nohup/test.out &