Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/edisonleeeee/graphgallery
GraphGallery is a gallery for benchmarking Graph Neural Networks, From InplusLab.
https://github.com/edisonleeeee/graphgallery
adversarial-machine-learning dgl geometric-deep-learning graph graph-neural-networks graphgallery models pyg pytorch
Last synced: 1 day ago
JSON representation
GraphGallery is a gallery for benchmarking Graph Neural Networks, From InplusLab.
- Host: GitHub
- URL: https://github.com/edisonleeeee/graphgallery
- Owner: EdisonLeeeee
- License: mit
- Created: 2020-03-10T13:57:10.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2023-08-14T09:37:11.000Z (over 1 year ago)
- Last Synced: 2024-11-12T13:38:31.651Z (2 days ago)
- Topics: adversarial-machine-learning, dgl, geometric-deep-learning, graph, graph-neural-networks, graphgallery, models, pyg, pytorch
- Language: Python
- Homepage:
- Size: 9.56 MB
- Stars: 459
- Watchers: 13
- Forks: 64
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
PyTorch is all you need!
# GraphGallery
GraphGallery is a gallery for benchmarking Graph Neural Networks (GNNs) based on pure [PyTorch](https://github.com/pytorch/pytorch) backend. Alteratively, [Pytorch Geometric (PyG)](https://github.com/pyg-team/pytorch_geometric) and [Deep Graph Library (DGL)](https://www.dgl.ai/) backend are also available in GraphGallery to facilitate your implementations.
# 💨 NEWS
+ November 20, 2021: We now no longer support `TensorFlow` backend.
+ November 20, 2021: The module `graphgallery.attack` is deprecated, users may refer to [GraphWar](https://github.com/EdisonLeeeee/GraphWar) for more information.
# 🚀 Installation
Please make sure you have installed [PyTorch](https://pytorch.org/). Also, [Pytorch Geometric (PyG)](https://github.com/pyg-team/pytorch_geometric) and [Deep Graph Library (DGL)](https://www.dgl.ai/) are alternative choices.
Install from source:
```bash
# Recommended
git clone https://github.com/EdisonLeeeee/GraphGallery.git && cd GraphGallery
pip install -e . --verbose
```
where `-e` means "editable" mode so you don't have to reinstall every time you make changes.
**NOTE**: GraphGallery is a frequently updated package and DO NOT install GraphGallery with `pip`, we're currently working on releasing a binary distribution on `PyPI`, stay tuned!
# 🤖 Implementations
In detail, the following methods are currently implemented:
## Node Classification
| Method | Author | Paper | PyTorch | PyG | DGL |
| ------------------------- | ----------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------ | ------------------ | ------------------ |
| **ChebyNet** | *Michaël Defferrard et al.* | [Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering (NeurIPS'16)](https://arxiv.org/abs/1606.09375) | :heavy_check_mark: | | |
| **GCN** | *Thomas N. Kipf et al.* | [Semi-Supervised Classification with Graph Convolutional Networks (ICLR'17)](https://arxiv.org/abs/1609.02907) | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| **GraphSAGE** | *William L. Hamilton et al.* | [Inductive Representation Learning on Large Graphs (NeurIPS'17)](https://arxiv.org/abs/1706.02216) | :heavy_check_mark: | :heavy_check_mark: | |
| **FastGCN** | *Jie Chen et al.* | [FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling (ICLR'18)](https://arxiv.org/abs/1801.10247) | :heavy_check_mark: | | |
| **GAT** | *Petar Veličković et al.* | [Graph Attention Networks (ICLR'18)](https://arxiv.org/abs/1710.10903) | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| **SGC** | *Felix Wu et al.* | [Simplifying Graph Convolutional Networks (ICLR'19)](https://arxiv.org/abs/1902.07153) | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| **GWNN** | *Bingbing Xu et al.* | [Graph Wavelet Neural Network (ICLR'19)](https://arxiv.org/abs/1904.07785) | :heavy_check_mark: | | |
| **ClusterGCN** | *Wei-Lin Chiang et al.* | [Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks (KDD'19)](https://arxiv.org/abs/1905.07953) | :heavy_check_mark: | | |
| **DAGNN** | *Meng Liu et al.* | [Towards Deeper Graph Neural Networks (KDD'20)](https://arxiv.org/abs/2007.09296) | :heavy_check_mark: | | :heavy_check_mark: |
| **GDC** | *Johannes Klicpera et al.* | [Diffusion Improves Graph Learning (NeurIPS'19)](https://www.in.tum.de/daml/gdc/) | :heavy_check_mark: | | |
| **TAGCN** | *Jian Du et al.* | [Topology Adaptive Graph Convolutional Networks (arxiv'17)](https://arxiv.org/abs/1710.10370) | :heavy_check_mark: | | |
| **APPNP, PPNP** | *Johannes Klicpera et al.* | [Predict then Propagate: Graph Neural Networks meet Personalized PageRank (ICLR'19)](https://arxiv.org/abs/1810.05997) | :heavy_check_mark: | | :heavy_check_mark: |
| **PDN** | *Benedek Rozemberczki et al.* | [Pathfinder Discovery Networks for Neural Message Passing (ICLR'21)](https://arxiv.org/abs/2010.12878) | | :heavy_check_mark: | |
| **SSGC** | *Zhu et al.* | [Simple Spectral Graph Convolution (ICLR'21)](https://openreview.net/forum?id=CYO5T-YjWZV) | :heavy_check_mark: | | |
| **AGNN** | *Kiran K. Thekumparampil al.* | [Attention-based Graph Neural Network for semi-supervised learning (ICLR'18 openreview)](https://arxiv.org/abs/1803.03735) | :heavy_check_mark: | | |
| **ARMA** | *Bianchi et al.* | [Graph Neural Networks with convolutional ARMA filters (Arxiv'19)](https://arxiv.org/abs/1901.01343) | | | |
| **GraphML*P*** | *Yang Hu et al.* | [Graph-MLP: Node Classification without Message Passing in Graph (Arxiv'21)](https://arxiv.org/abs/2106.04051) | :heavy_check_mark: | | |
| **LGC, EGC, hLGC** | *Luca Pasa et al.* | [Simple Graph Convolutional Networks (Arxiv'21)](https://arxiv.org/abs/2106.05809) | | | :heavy_check_mark: |
| **GRAND** | *Wenzheng Feng et al.* | [Graph Random Neural Network for Semi-Supervised Learning on Graphs (NeurIPS'20)](https://arxiv.org/abs/2005.11079) | | | :heavy_check_mark: |
| **AlaGCN, AlaGAT** | *Yiqing Xie et al.* | [When Do GNNs Work: Understanding and Improving Neighborhood Aggregation (IJCAI'20)](https://www.ijcai.org/Proceedings/2020/0181.pdf) | | | :heavy_check_mark: |
| **JKNet** | *Keyulu Xu et al.* | [Representation Learning on Graphs with Jumping Knowledge Networks (ICML'18)](https://arxiv.org/abs/1806.03536) | | | :heavy_check_mark: |
| **MixHop** | *Sami Abu-El-Haija et al.* | [MixHop: Higher-Order Graph Convolutional Architecturesvia Sparsified Neighborhood Mixing (ICML'19)](https://arxiv.org/abs/1905.00067) | | | :heavy_check_mark: |
| **DropEdge** | *Yu Rong et al.* | [DropEdge: Towards Deep Graph Convolutional Networks on Node Classification (ICML'20)](https://arxiv.org/abs/1907.10903) | | :heavy_check_mark: | |
| **Node2Grids** | *Dalong Yang et al.* | [Node2Grids: A Cost-Efficient Uncoupled Training Framework for Large-Scale Graph Learning (CIKM'21)](https://arxiv.org/abs/2003.09638) | :heavy_check_mark: | | |
| **RobustGCN** | *Dingyuan Zhu et al.* | [Robust Graph Convolutional Networks Against Adversarial Attacks (KDD'19)](https://dl.acm.org/doi/10.1145/3292500.3330851) | :heavy_check_mark: | | :heavy_check_mark: |
| **SBVAT, OBVAT** | *Zhijie Deng et al.* | [Batch Virtual Adversarial Training for Graph Convolutional Networks (ICML'19)](https://arxiv.org/abs/1902.09192) | :heavy_check_mark: | | |
| **SimPGCN** | *Wei Jin et al.* | [Node Similarity Preserving Graph Convolutional Networks (WSDM'21)](https://arxiv.org/abs/2011.09643) | :heavy_check_mark: | | |
| **GraphVAT** | *Fuli Feng et al.* | [Graph Adversarial Training: Dynamically Regularizing Based on Graph Structure (TKDE'19)](https://arxiv.org/abs/1902.08226) | :heavy_check_mark: | | |
| **LATGCN** | *Hongwei Jin et al.* | [Latent Adversarial Training of Graph Convolution Networks (ICML@LRGSD'19)](https://graphreason.github.io/papers/35.pdf) | :heavy_check_mark: | | |
| **DGAT** | *Weibo Hu et al.* | [Robust graph convolutional networks with directional graph adversarial training (Applied Intelligence'19)](https://link.springer.com/article/10.1007/s10489-021-02272-y) | :heavy_check_mark: | | |
| **MedianGCN, TrimmedGCN** | *Liang Chen et al.* | [Understanding Structural Vulnerability in Graph Convolutional Networks](https://arxiv.org/abs/2108.06280) | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
#### Graph Purification
The graph purification methods are universal for all models, just specify:
```python
graph_transform="purification_method"
```
so, here we only give the examples of `GCN` with purification methods, other models should work.
| Method | Author | Paper |
| --------------- | ----------------------- | ------------------------------------------------------------------------------------------------------------------------------------------- |
| **GCN-Jaccard** | *Huijun Wu et al.* | [Adversarial Examples on Graph Data: Deep Insights into Attack and Defense (IJCAI'19)](https://arxiv.org/abs/1903.01610) |
| **GCN-SVD** | *Negin Entezari et al.* | [All You Need Is Low (Rank): Defending Against Adversarial Attacks on Graphs (WSDM'20)](https://dl.acm.org/doi/abs/10.1145/3336191.3371789) |
## LinkPrediction
| Method | Author | Paper | PyTorch | PyG | DGL |
| ------------- | ----------------------- | ------------------------------------------------------------------------------- | ------------------ | ------------------ | --- |
| **GAE, VGAE** | *Thomas N. Kipf et al.* | [Variational Graph Auto-Encoders (NeuIPS'16)](https://arxiv.org/abs/1611.07308) | :heavy_check_mark: | :heavy_check_mark: | |
## Node Embedding
The following methods are framework-agnostic.
| Method | Author | Paper |
| ------------- | --------------------------------- | --------------------------------------------------------------------------------------------------------------- |
| **Deepwalk** | *Bryan Perozzi et al.* | [DeepWalk: Online Learning of Social Representations (KDD'14)](https://arxiv.org/abs/1403.6652) |
| **Node2vec** | *Aditya Grover and Jure Leskovec* | [node2vec: Scalable Feature Learning for Networks (KDD'16)](https://arxiv.org/abs/1607.00653) |
| **Node2vec+** | *Renming Liu et al.* | [Accurately Modeling Biased Random Walks on Weighted Graphs Using Node2vec+](https://arxiv.org/abs/2109.08031) |
| **BANE** | *Hong Yang et al.* | [Binarized attributed network embedding (ICDM'18)](https://ieeexplore.ieee.org/document/8626170) |
# ⚡ Quick Start
## Datasets
+ [Planetoid](https://github.com/kimiyoung/planetoid): a collection of widely used benchmark datasets in graph learning tasks, including 'cora', 'citeseerr', 'pubmed' and 'nell' datasets.
+ NPZDataset: a collection of graph datasets stored with numpy `.npz` format.
you can simply run `dataset.available_datasets()` to see the available datasets, e.g.,:
```python
from graphgallery.datasets import Planetoid
print(Planetoid.available_datasets())
```
more details please refer to [GraphData](https://github.com/EdisonLeeeee/GraphData).
## Example of GCN (Node Classification Task)
It takes just a few lines of code.
```python
import torch
import graphgallery
from graphgallery.datasets import Planetoid
from graphgallery.gallery import callbacks
data = Planetoid('cora', root="~/GraphData/datasets/", verbose=True)
graph = data.graph
splits = data.split_nodes()
device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
graphgallery.set_backend("torch")
from graphgallery.gallery.nodeclas import GCN
trainer = GCN(device=device, seed=123).setup_graph(graph, feat_transform="normalize_feat").build()
cb = callbacks.ModelCheckpoint('model.pth', monitor='val_accuracy')
trainer.fit(splits.train_nodes, splits.val_nodes, verbose=1, callbacks=[cb])
results = trainer.evaluate(splits.test_nodes)
print(f'Test loss {results.loss:.5}, Test accuracy {results.accuracy:.2%}')
```
## Example of GAE (Link Prediction Task)
```python
import torch
import graphgallery
from graphgallery.gallery import callbacks
from graphgallery.datasets import Planetoid
data = Planetoid('cora', root="~/GraphData/datasets/", verbose=True)
graph = data.graph
splits = data.split_edges(random_state=15)
device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
graphgallery.set_backend("torch")
from graphgallery.gallery.linkpred import GAE
trainer = GAE(device=device, seed=123).setup_graph(graph).build()
cb = callbacks.ModelCheckpoint('model.pth', monitor='val_ap')
trainer.fit(splits.train_pos_edge_index,
val_data=(splits.val_pos_edge_index, splits.val_neg_edge_index),
verbose=1, callbacks=[cb])
results = trainer.evaluate((splits.test_pos_edge_index, splits.test_neg_edge_index))
print(results)
```
If you have any troubles, you can simply run `trainer.help()` for more information.
## Other Backends
```python
>>> import graphgallery
# Default: PyTorch backend
>>> graphgallery.backend()
PyTorch 1.9.0+cu111 Backend
# Switch to PyTorch Geometric backend
>>> graphgallery.set_backend("pyg")
# Switch to DGL PyTorch backend
>>> graphgallery.set_backend("dgl")
# Switch to PyTorch backend
>>> graphgallery.set_backend("th") # "torch", "pytorch"
```
But your codes don't even need to change.
# ❓ How to add your datasets
This is motivated by [gnn-benchmark](https://github.com/shchur/gnn-benchmark/)
```python
from graphgallery.data import Graph
# Load the adjacency matrix A, attribute (feature) matrix X and labels vector y
# A - scipy.sparse.csr_matrix of shape [num_nodes, num_nodes]
# X - scipy.sparse.csr_matrix or numpy.ndarray of shape [num_nodes, num_feats]
# y - numpy.ndarray of shape [num_nodes]
mydataset = Graph(adj_matrix=A, attr_matrix=X, label=y)
# save dataset
mydataset.to_npz('path/to/mydataset.npz')
# load dataset
mydataset = Graph.from_npz('path/to/mydataset.npz')
```
# ⭐ Road Map
- [x] Add PyTorch trainers support
- [x] Add other frameworks (PyG and DGL) support
- [x] set tensorflow as optional dependency when using graphgallery
- [ ] Add more GNN trainers
- [ ] Support for more tasks, e.g., `graph Classification` and `link prediction`
- [x] Support for more types of graphs, e.g., Heterogeneous graph
- [ ] Add Docstrings and Documentation (Building)
- [ ] Comprehensive tutorials
# ❓ FAQ
Please fell free to contact me if you have any troubles.
# 😘 Acknowledgement
This project is motivated by [Pytorch Geometric](https://github.com/rusty1s/pytorch_geometric), [Stellargraph](https://github.com/stellargraph/stellargraph) and [DGL](https://www.dgl.ai/), etc., and the original implementations of the authors, thanks for their excellent works!
# Cite
Please cite our [paper](https://www.computer.org/csdl/proceedings-article/icse-companion/2021/121900a013/1sET5DXNWJG) (and the respective papers of the methods used) if you use this code in your own work:
```bibtex
@inproceedings{li2021graphgallery,
author = {Jintang Li and Kun Xu and Liang Chen and Zibin Zheng and Xiao Liu},
booktitle = {2021 IEEE/ACM 43rd International Conference on Software Engineering: Companion Proceedings (ICSE-Companion)},
title = {GraphGallery: A Platform for Fast Benchmarking and Easy Development of Graph Neural Networks Based Intelligent Software},
year = {2021},
pages = {13-16},
publisher = {IEEE Computer Society},
address = {Los Alamitos, CA, USA},
}
```