SetToGraphPaper

This repository holds the code for the paper: https://arxiv.org/abs/2008.02831 mostly the same as https://github.com/hadarser/SetToGraphPaper with some small modifications to the inference and baseline models.

Data

Before running the code for the jets experiments, the data should be downloaded using the following commands:

cd SetToGraphPaper
python download_jets_data.py

This script will download all the data from Zenodo links.

Code

Prerequisites

You can use the following code to install a compatible environment (using anaconda), make sure to change the cuda toolkit version to the one that fits.

conda create -n s2g_env -c pytorch pytorch=1.5 cudatoolkit=10.2 torchvision  
conda activate s2g_env
CUDA=cu102  # for cuda 10.2
pip install torch-scatter==latest+${CUDA} -f https://pytorch-geometric.com/whl/torch-1.5.0.html
pip install torch-sparse==latest+${CUDA} -f https://pytorch-geometric.com/whl/torch-1.5.0.html
pip install torch-cluster==latest+${CUDA} -f https://pytorch-geometric.com/whl/torch-1.5.0.html
pip install torch-spline-conv==latest+${CUDA} -f https://pytorch-geometric.com/whl/torch-1.5.0.html
pip install torch-geometric
conda install -c conda-forge -c anaconda tqdm easydict rdkit uproot

Running the tests

The folder main_scripts contains scripts that run different experiments:

1. To run the main paticle-physics jets experiment with our chosen hyper-parameters, run the one of the following:

python main_scripts/main_jets.py --method=lin2  # for S2G

or

python main_scripts/main_jets.py --method=lin5  # for S2G+

or change --method=... with --baseline=siam/rnn for running a baseline.