benedekrozemberczki / karateclub

Karate Club: An API Oriented Open-source Python Framework for Unsupervised Learning on Graphs (CIKM 2020)
https://karateclub.readthedocs.io
GNU General Public License v3.0
2.17k stars 247 forks source link

How do I use my own data set and train. #73

Closed fiego closed 3 years ago

fiego commented 3 years ago

I have a directed graph of data, and I need to use graph embedding or node embedding to get the node embedding representation vector.

After reading the tutorial, I didn't find out how to use my own data.

1 Can you explain in code how to use my own data set? 2 Also, what should I do with my data set before it can be loaded? 3 And how do we get the embedded vector for this node after we train it?

benedekrozemberczki commented 3 years ago

The paper and the tutorial explain the input data requirements in great detail. If you need a node embedding you need a graph with numeric indices stored as a networkx graph. A graph embedding requires a list of networkx graphs. The documentation is explicit about these things. Could you star the repo and hit follow?