I see three popular python drivers: neo4j, neo2py, graphdatascience. Each one seem to do similar things.
1) What is a preferred long term direction?
2) I am a data scientist. I want to run many gds algorithms. Is there a guidance on what is one library I should use? There is no clear documentation on this topic. There is no comparison of their capability? Can use just one of these libraries or do I have to use more than one to make it work?
>>I see three popular python drivers: neo4j, neo2py, graphdatascience.
neo2py? or are you suggesting py2neo
As to which to use if nothing more the Neo4j Python driver is maintained by Neo4j and as such has regular support/improvements etc. py2neo is a community driver and as such may or may not have same commttment
If you want to use GDS, go with the Graph Data Science client that wraps the official Python driver.
For general application development, I'd advise to stick with the official neo4j driver. This one is here to stay.
Py2neo has been maintained by Nigel Small, a long-time Neo4j employee. It offers high-level features, but Nigel does not work at Neo anymore so what the future holds for Py2neo is quite unclear.