Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ysig/GraKeL
A scikit-learn compatible library for graph kernels
https://github.com/ysig/GraKeL
bioinformatics chemoinformatics graph-classification graph-kernels graph-mining graph-similarity graph-similarity-algorithms scikit-learn
Last synced: about 1 month ago
JSON representation
A scikit-learn compatible library for graph kernels
- Host: GitHub
- URL: https://github.com/ysig/GraKeL
- Owner: ysig
- License: other
- Created: 2017-10-31T13:05:27.000Z (about 7 years ago)
- Default Branch: master
- Last Pushed: 2024-10-24T13:11:10.000Z (2 months ago)
- Last Synced: 2024-10-25T15:14:30.851Z (2 months ago)
- Topics: bioinformatics, chemoinformatics, graph-classification, graph-kernels, graph-mining, graph-similarity, graph-similarity-algorithms, scikit-learn
- Language: Python
- Homepage: https://ysig.github.io/GraKeL/
- Size: 26.4 MB
- Stars: 594
- Watchers: 10
- Forks: 97
- Open Issues: 9
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
--------------------------------------------------------------------------------
> [!NOTE]
> **GraKeL needs your support!**
> Our goal with GraKeL was to have a stable library for graph kernels but as tools change it requires maintenance:
> Lots of libraries change requiring constant pull requests to keep up with newer python/numpy versions,
> or for example modernizing our ci-cycle was not finalized and the owner currently lacks knowledge and time to go over changes.
> If there is a community of researchers that want to be part of grakel's support please reach out,
> to `ioannis siglidis enpc fr` and we would be happy to gradually move you into maintenance.
> Thank you! 💕
[](https://pypi.org/pypi/grakel/)
[](https://codecov.io/gh/ysig/GraKeL)
[](https://circleci.com/gh/ysig/GraKeL)
**[Documentation](https://ysig.github.io/GraKeL/)** | **[Paper](http://jmlr.org/papers/volume21/18-370/18-370.pdf)**
*GraKeL* is a library that provides implementations of several well-established graph kernels. The library unifies these kernels into a common framework. Furthermore, it provides implementations of some frameworks that work on top of graph kernels. Specifically, GraKeL contains 16 kernels and 2 frameworks. The library is compatible with the [scikit-learn](http://scikit-learn.org/) pipeline allowing easy and fast integration inside machine learning algorithms.
--------------------------------------------------------------------------------
In detail, the following kernels and frameworks are currently implemented:
* **[Vertex histogram kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.VertexHistogram.html)**
* **[Edge histogram kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.EdgeHistogram.html)**
* **[Shortest path kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.ShortestPath.html)** from Borgwardt and Kriegel: [Shortest-path kernels on graphs](https://www.dbs.ifi.lmu.de/~borgward/papers/BorKri05.pdf) (ICDM 2005)
* **[Graphlet kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.GraphletSampling.html)** from Shervashidze *et al.*: [Efficient graphlet kernels for large graph comparison](http://proceedings.mlr.press/v5/shervashidze09a/shervashidze09a.pdf) (AISTATS 2009)
* **[Random walk kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.RandomWalk.html)** from Vishwanathan *et al.*: [Graph Kernels](http://www.jmlr.org/papers/volume11/vishwanathan10a/vishwanathan10a.pdf) (JMLR 11(Apr))
* **[Neighborhood hash graph kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.NeighborhoodHash.html)** from Hido and Kashima: [A Linear-time Graph Kernel](https://ieeexplore.ieee.org/abstract/document/5360243) (ICDM 2009)
* **[Weisfeiler-Lehman framework](https://ysig.github.io/GraKeL/latest/generated/grakel.WeisfeilerLehman.html)** from Shervashidze *et al.*: [Weisfeiler-Lehman Graph Kernels](http://www.jmlr.org/papers/volume12/shervashidze11a/shervashidze11a.pdf) (JMLR 12(Sep))
* **[Neighborhood subgraph pairwise distance kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.NeighborhoodSubgraphPairwiseDistance.html)** from Costa and De Grave: [Fast Neighborhood Subgraph Pairwise Distance Kernel](https://pdfs.semanticscholar.org/7a10/f6a406b664d1159e7c4fefbdd6ac275aee53.pdf) (ICML 2010)
* **[Lovasz-theta kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.LovaszTheta.html)** from Johansson *et al.*: [Global graph kernels using geometric embeddings](http://proceedings.mlr.press/v32/johansson14.pdf) (ICML 2014)
* **[SVM-theta kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.SvmTheta.html)** from Johansson *et al.*: [Global graph kernels using geometric embeddings](http://proceedings.mlr.press/v32/johansson14.pdf) (ICML 2014)
* **[Ordered decompositional DAG kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.OddSth.html)** from Da San Martino *et al.*: [A Tree-Based Kernel for Graphs](https://pdfs.semanticscholar.org/69ee/18dd7a214d4d656b5b95742212f050dabeac.pdf) (SDM 2012)
* **[GraphHopper kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.GraphHopper.html)** from Feragen *et al.*: [Scalable kernels for graphs with continuous attributes](https://papers.nips.cc/paper/5155-scalable-kernels-for-graphs-with-continuous-attributes.pdf) (NIPS 2013)
* **[Propagation kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.Propagation.html)** from Neumann *et al.*: [Propagation kernels: efficient graph kernels from propagated information](https://link.springer.com/content/pdf/10.1007/s10994-015-5517-9.pdf) (Machine Learning 102(2))
* **[Pyramid match kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.PyramidMatch.html)** from Nikolentzos *et al.*: [Matching Node Embeddings for Graph Similarity](https://www.aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14494/14426) (AAAI 2017)
* **[Subgraph matching kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.SubgraphMatching.html)** from Kriege and Mutzel: [Subgraph Matching Kernels for Attributed Graphs](https://arxiv.org/ftp/arxiv/papers/1206/1206.6483.pdf) (ICML 2012)
* **[Multiscale Laplacian kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.MultiscaleLaplacian.html)** from Kondor and Pan: [The Multiscale Laplacian Graph Kernel](https://papers.nips.cc/paper/6135-the-multiscale-laplacian-graph-kernel.pdf) (NIPS 2016)
* **[Core framework](https://ysig.github.io/GraKeL/latest/generated/grakel.CoreFramework.html)** from Nikolentzos *et al.*: [A Degeneracy Framework for Graph Similarity](https://www.ijcai.org/proceedings/2018/0360.pdf) (IJCAI 2018)
* **[Weisfeiler-Lehman optimal assignment kernel](https://ysig.github.io/GraKeL/latest/generated/grakel.WeisfeilerLehmanOptimalAssignment.html)** from Kriege *et al.*: [On Valid Optimal Assignment Kernels and Applications to Graph Classification](http://papers.nips.cc/paper/6166-on-valid-optimal-assignment-kernels-and-applications-to-graph-classification.pdf) (NIPS 2016)
--------------------------------------------------------------------------------
To learn how to install and use GraKeL, and to find out more about the implemented kernels and frameworks, please read our [documentation](https://ysig.github.io/GraKeL/). To learn about the functionality of the library and about example applications, check out our [examples](https://github.com/ysig/GraKeL/tree/master/examples) in the `examples/` directory and our [tutorials](https://github.com/ysig/GraKeL/tree/master/tutorials) in the `tutorials/` directory.
In case you find a bug, please open an [issue](https://github.com/ysig/GraKeL/issues). To propose a new kernel, you can open a [feature request](https://github.com/ysig/GraKeL/issues).
## Installation
The GraKeL library requires the following packages to be installed:
* Python (>=2.7, >=3.5)
* NumPy (>=1.8.2)
* SciPy (>=0.13.3)
* Cython (>=0.27.3)
* cvxopt (>=1.2.0) [optional]
* future (>=0.16.0) (for python 2.7)
To install the package, run:
```sh
$ pip install grakel
```
## Running tests
To test the package, execute:
```sh
$ pytest
```
## Running examples
```
$ cd examples
$ python shortest_path.py
```
## Cite
If you use GraKeL in a scientific publication, please cite our paper (http://jmlr.org/papers/volume21/18-370/18-370.pdf):
```bibtex
@article{JMLR:v21:18-370,
author = {Giannis Siglidis and Giannis Nikolentzos and Stratis Limnios and Christos Giatsidis and Konstantinos Skianis and Michalis Vazirgiannis},
title = {GraKeL: A Graph Kernel Library in Python},
journal = {Journal of Machine Learning Research},
year = {2020},
volume = {21},
number = {54},
pages = {1-5}
}
```
## License
GraKeL is distributed under the __BSD 3-clause__ license. The library makes use of the C++ source code of [BLISS](http://www.tcs.hut.fi/Software/bliss) (a tool for computing automorphism groups and canonical labelings of graphs) which is __LGPL__ licensed. Futhermore, the [cvxopt](https://cvxopt.org/) package (a software package for convex optimization) which is an optional dependency of GraKeL is __GPL__ licensed.
## Acknowledgements
We would like to thank [@eddiebergman](https://github.com/eddiebergman) for modernizing our CI and extending our python support.