Bugfix: use subjects.index to generate node embeddings

[daokunzhang]

In the demo notebook demos/embeddings/keras-node2vec-embeddings.ipynb, demos/node-classification/attri2vec-node-classification.ipynb, demos/node-classification/keras-node2vec-node-classification.ipynb, node embeddings are generated by passing G.nodes() into the node generator. For example, in demos/embeddings/keras-node2vec-embeddings.ipynb, node embeddings are generated as

node_gen = Node2VecNodeGenerator(G, batch_size).flow(G.nodes())
node_embeddings = embedding_model.predict(node_gen, workers=4, verbose=1)

Though the order of G.nodes() is consistent with subjects, it is more desirable to use subjects.index to generate node embeddings to avoid confusing readers.

In this pull request, I replace the G.nodes() with subjects.index in the above three demo notebooks.

[review-notebook-app]

Check out this pull request on 

Review Jupyter notebook visual diffs & provide feedback on notebooks.

---

Powered by ReviewNB

[codeclimate]

Code Climate has analyzed commit e0084b1 and detected 0 issues on this pull request.

View more on Code Climate.