GNN-Pytorch

Pytorch implementation of GNN methods and models. Pytorch implementation of GNN.

THe intention of this repo is understand and integrate the methods then obtaining highger results.

Node Classification

List of data sets used, OGB (https://github.com/snap-stanford/ogb) data sets can be used if conditions permit :

Dataset	Nodes	Edges	Node Attr.	Classes	Train	Valid	Test
Cora	2708	5429	1433	7	140	500	1000
Cora-Full	2708	5429	1433	7	1208	500	1000
Citeseer	3327	4732	3703	6	120	500	1000
Citeseer-Full	3327	4732	3703	6	1827	500	1000
Pubmed	19717	44338	500	3	60	500	1000
Pubmed-Full	19717	44338	500	3	18217	500	1000

Accuracy Metrics of Node Classification

Status	Method	Paper	Cora	Citeseer	Pubmed
✔️	GCN	Kipf and Welling, 2017	0.8220	0.6960	0.780
✔️	GraphSAGE	Hamilton and Ying et al., 2017	0.850	NA	0.808
✔️	GAT	Velickovic et al., 2018	0.7840	0.7060	NA
✔️	SP_GAT	Yang Ye, Shihao Ji et al., 2021	0.8180	0.7070	0.7750
✔️	FastGCN*	Chen and Ma et al., 2018	0.8240	0.7710	0.8780
✔️	GRAND	Feng and Zhang et al., 2020	0.839	0.726	0.797
✔️	GWNN	Bingbing Xu, Huawei Shen, Qi Cao et al., 2019	0.7990	0.7710	NA

* NA is because was not able to run in my system

Packages

Package	Version	Installation
python	3.8.6	conda create --name gnn python=3.8.6
numpy	1.20.0	pip install numpy==1.20.0
scipy	1.6.0	pip install scipy==1.6.0
pyyaml	5.4.1	pip install pyyaml==5.4.1
scikit-learn	0.24.1	pip install scikit-learn==0.24.1
pytorch	1.7.1	conda install pytorch==1.7.1 cudatoolkit=11.0 -c pytorch
torch-geometric	1.6.3	Installation

Reference Links

GCN : https://github.com/tkipf/pygcn

GAT : https://github.com/Diego999/pyGAT

SP_GAT : https://github.com/Diego999/pyGAT

GraphSage : https://github.com/dsgiitr/graph_nets

GWNN : https://github.com/Yanqi-Chen/GWNN

FastGCN : https://github.com/Gkunnan97/FastGCN_pytorch

GRAND : https://github.com/THUDM/GRAND

GraphLime: https://github.com/WilliamCCHuang/GraphLIME.git

License

This is open source project collected from other open sources , feel free to give feedback and raise comments