This repository contains the code for the Graph+AI World session From Dataframes to Graph: Data Science with pyTigerGraph by Parker Erickson.
Unfortunately, this code will only work on UNIX systems, and they must have Java installed for the Python GSQL interface. This is something we are trying to remedy for future versions of pyTigerGraph.
Download the data here. The directory structure should be:
graphAIWorldDataframeToGraph
|-data
| |- artistData.tsv
| |_ etc.
|
|-createSchemaAndLoad.ipynb
|-genreClassification.ipynb
|_ ...
The notebooks assume that you are using a tgcloud instance of TigerGraph. You can create a free instance, and choose "Blank" for the starter kit.
This repository uses Pandas, pyTigerGraph, Pytorch, and DGL (might be missing a few there). Make sure you have these installed before trying to use the notebooks.
Building out this site with many different papers and videos to get you started with graph machine learning.