Issues in leveraging created vector index for local implementation

I have created the movie plots vector index (refer below screengrab) as mentioned in the article [Neo4j & LLM Fundamentals | Neo4j & LLMs | Free Neo4j Courses from GraphAcademy]

But when I am trying to leverage the same in the local code implementation of the Build a Neo4j-backed Chatbot using Python, I keep getting the ValueError: The specified vector index name does not exist. Make sure to check if you spelled it correctly. Please suggest the way to resolve the same.

Hi,

The error is because (for whatever reason) Python cannot find the vector index in the database the program is connected to.

Without seeing your code, database and environment setup is difficult to provide advice.

I would check:

  • that the vector index name is the same in your code as in the database (it is case sensitive)
  • that the program is connecting to the same database that you have created the index in

Martin

Hi Martin,

Thanks for the reply.

I had earlier set it up on the sandbox system to which I do not have access currently. I intend to set it up on the free instance made available by Neo4j but unfortunately, I am not able to replicate the steps there (having issues with loading exact movies dataset and so on). Can anyone from the team connect with me for 10 mins to look into the same? If any other document can be made available that I can act as a ready reckoner (would be able to run the steps as it is) would be really helpful. Sort of seeking handholding as I am pretty new to this.

GraphAcademy uses sandbox's "prebuilt" datasets, in the case of these courses, the "recommendations" dataset.

The same datasets are not available on Aura, although you could potentially restore them from a dump file which are available on github - recommendations/data at main · neo4j-graph-examples/recommendations · GitHub

You can also access sandbox directly (rather than through graphacademy) at https://sandbox.neo4j.com/