fastGraph
A fast and lightweight package designed for Graph Embedding.
Built for DeepWalk, node2vec walk and likely walk(sampled with edges of transition probability).
Built for both undirected, directed and both weighted-, binary-edge graph.
Dev Usage
python fastGraph/__main__.py --input example_graph/wiki-Vote.csv --output example_graph/wiki-Vote.model --walk-type node2vec --ngram 1 --number-paths 5 --embed-size 64 --path-length 40
Package Installation & Usage
cd fastGraph
pip install -r requirements.txt
fastGraph --input example_graph/wiki-Vote.csv --output example_graph/wiki-Vote.model --walk-type node2vec --ngram 1 --number-paths 5 --embed-size 64 --path-length 40
Pip Installation
pip install fastGraph
Dataset
There are various graph dataset available on the internet, for example here's a collection_ from Stanford. .. _collection: https://snap.stanford.edu/data/
Acknowledgement
Inspired by the great work from deepwalk.
