GNNs-Baseline

A Pytorch based implementation of classical GNNs.
GNN基准测试模型.
Some components of this code are adapted from GPRGNN and H2GCN.

Requirement

This project should be able to run without any modification after following packages installed.

pytorch
torch_geometric
networkx
torch-sparse

GNNs supported

H2GCN FAGCN APPNP GCN GAT ChebNet APPNP JKNet GPRGNN

Unsupervised models supported

Network Embedding as Matrix Factorization: Unifying DeepWalk, LINE, PTE, and node2vec

DeepWalk Line NetMF

Datasets supported

Download by yourself, dataset_utils.py will help you.

cora, citeseer, pubmed, computers, photo, chameleon, squirrel, film, texas, cornell, wisconsin

Run experiment with GCN & Cora:

go to folder src

python train_model.py --RPMAX 1 \
        --net GCN \
        --train_rate 0.05 \
        --val_rate 0.05 \
        --dataset cora 

Visualization of Confusion Matrix