A Fair Comparison of Graph Neural Networks for Graph Classification (ICLR 2020)

Summary

The library includes data and scripts to reproduce the experiments reported in the paper.

If you happen to use or modify this code, please remember to cite our paper:

Federico Errica, Marco Podda, Davide Bacciu, Alessio Micheli: A Fair Comparison of Graph Neural Networks for Graph Classification. Proceedings of the 8th International Conference on Learning Representations (ICLR 2020).

@inproceedings{errica_fair_2020,
    title = {A fair comparison of graph neural networks for graph classification},
    booktitle = {Proceedings of the 8th {International} {Conference} on {Learning} {Representations} ({ICLR})},
    author = {Errica, Federico and Podda, Marco and Bacciu, Davide and Micheli, Alessio},
    year = {2020}
}

--

If you are interested in an introduction to Deep Graph Networks (and a new library!), check this out:

Bacciu Davide, Errica Federico, Micheli Alessio, Podda Marco: A Gentle Introduction to Deep Learning for Graphs, Neural Networks, 2020. DOI:

10.1016/j.neunet.2020.06.006

Installation

We provide a script to install the environment. You will need the conda package manager, which can be installed from here.

To install the required packages, follow there instructions (tested on a linux terminal):

1) clone the repository

git clone https://github.com/diningphil/gnn-comparison

2)

cd

cd gnn-comparison

3) run the install script

source install.sh []

Where

cpu

cu92

cu100

cu101

cpu

gnn-comparison

bash

source

Instructions

To reproduce the experiments, first preprocess datasets as follows:

python PrepareDatasets.py DATA/CHEMICAL --dataset-name  --outer-k 10

python PrepareDatasets.py DATA/SOCIAL_1 --dataset-name  --use-one --outer-k 10

python PrepareDatasets.py DATA/SOCIAL_DEGREE --dataset-name  --use-degree --outer-k 10

Where

data_splits

Please note that dataset folders should be organized as follows:

CHEMICAL:
    NCI1
    DD
    ENZYMES
    PROTEINS
SOCIAL[_1 | _DEGREE]:
    IMDB-BINARY
    IMDB-MULTI
    REDDIT-BINARY
    REDDIT-MULTI-5K
    COLLAB

Then, you can launch experiments by typing:

cp -r DATA/[CHEMICAL|SOCIAL_1|SOCIAL_DEGREE]/ DATA

python Launch_Experiments.py --config-file  --dataset-name  --result-folder  --debug

Where

Additional Notes

You can only use CUDA with the

--debug

Troubleshooting

If you would like PyTorch not to spawn multiple threads for each process (highly recommended), append

export OMP_NUM_THREADS=1