Find answers from the community

Updated last year

Postgres

At a glance

The community member is working with PGVectorStore and has data already uploaded to PostgreSQL. They want to know if they can create indexes on existing tables, as the documentation mainly covers uploading documents and creating indexes. The community members have not found a solution for this use case yet. Another community member responds that using an existing PostgreSQL database is something they haven't figured out how to do yet, and suggests that the community member submit a PR if they have ideas on how to do this. The community member also asks for more detailed documentation on the methods and objects available for VectorStoreIndex, as they've only found information on the VectorStoreIndex.from_documents method. A final community member suggests reading the source code for more information.

Useful resources
Hey, so I have been working on PGVectorStore : https://docs.llamaindex.ai/en/stable/examples/vector_stores/postgres.html. For my use case I already have data uploaded on PostGres, I wanted to know if we could create index on existing tables - because in the docs it is mainly for uploading a doc and creating indexes. I have searched quite a bit and did not find any relevant matches. would appreciate any help/ insights on this.
L
V
5 comments
Using an existing postgres db is something we haven't quite figured out how to do yet
Ideally llama-index creates it πŸ˜…

If you have ideas on how to do this, would love to see a PR
Thank you for your response, I also wanted to know if you have any detailed explanation on the methods/ objects that are available for VectorStoreIndex
because i have only come across this: https://docs.llamaindex.ai/en/latest/api_reference/indices/vector_store.html
I want to see if there is any documentation that explains the other methods as well, because in most example i have seen "VectorStoreIndex.from_documents"
tbh I would just read the source code πŸ˜…
Add a reply
Sign up and join the conversation on Discord