ADGCN

Pytorch Implementation for paper "Adversarial Graph Disentanglement". Note: a well-organized version will be coming soon!

2022/05/30 Update: The organized version has been released!

Introduction

Requirements

• PyTorch >= 1.1.0
• python 3.6
• networkx
• scikit-learn
• scipy
• munkres

Run from

preset hyperparameters version:

source ./pre_ADGCN.sh

or modifying the network hyperparameters and run

python main.py --param1 xxx --param2 xxx --param3 xxx ...

You can also use "meta.py" to search for the best combination of hyperparameters on each dataset:

python meta.py --datname $dataset_name

Data

We provide the citation network datasets under data/. Due to space limit, please download AMZ co-purchase dataset from https://github.com/shchur/gnn-benchmark#datasets