This is the implementation of paper:

Label Contrastive Coding based Graph Neural Network for Graph Classification

Requirements

The code is implemented in Python 3.7. Package used for development are just below.

networkx       
numpy              
scipy              
torch == 1.4.0
torch_geometric == 1.6.0

###Instructions for running the code

For LCGNN with different encoders, the training scripts are in separate files (e.g., ./for_gin).

1, Enter the for_gin file

cd ./for_gin

2, Run the code

python3 train_powerfulgnn_oneenc.py

for the momentum weight $\alpha = 0$ condition; or run the code

python3 train_powerfulgnn_twoenc.py

for other conditions.

###Note:

1, The default setting includes using the GPU. 2, To change model configurations, (e.g., set the epoch numbers of training as NNUMBER), add config --epochs NUMBER. 3, For the size limitation of Github, you can get the dataset from https://www.dropbox.com/sh/kc7xf42kz4lqx9a/AAC9wKim768TBNocN1JNPudFa?dl=0