Aquery ivfflat fix

[forbug]

Description

As described in the issue, vector store index querying was not successful
when the PGVectorStore was initialized with the ivfflat_probes argument.
This problem is potentially due to the behavior of psycopg2/3 relying on server-side binding,
which does not support SET statements. However, this is just
a hypothesis and may be related to a different issue.

The proposed solution works successfully and formats the variable directly
in the SET string.

Fixes # 17877

New Package?

Did I fill in the tool.llamahub section in the pyproject.toml and provide a detailed README.md for my new integration or package?

• Yes
• No

Version Bump?

Did I bump the version in the pyproject.toml file of the package I am updating? (Except for the llama-index-core package)

• Yes
• No

Type of Change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)

How Has This Been Tested?

Your pull-request will likely not be merged unless it is covered by some form of impactful unit testing.

• I added new unit tests to cover this change

Suggested Checklist:

• I have performed a self-review of my own code
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I ran make format; make lint to appease the lint gods

[forbug]

NOTE: This integration should be updated to use psycopg, which combines the functionality of psycopg2 and asyncio. I believe it would be backwards compatible, but would need to be tested. I do not have the time at the moment to update every instance of psycopg2/asyncio within the repo to update the dependency.

[forbug]

@logan-markewich Do you know of anyone that may be able to look into this? Our solution relies on many query engines/retrievers that all use the asynchronous querying (aquery) except for our RetrieverQueryEngine, because of this problem.

We have tried:

• not setting ivfflat.probes at all, but the default produces less-deterministic results than setting it to 20
• setting ivfflat.probes at the global level for a database, but this reduces the flexibility of having multiple settings for different retrievers and our production environment cannot have OWNER privileges for the postgres user credentials and, thus, cannot set global settings

Without being able to use asynchronous querying for this one step of our workflows, our workflows are severely slowed down. Any help would be greatly appreciated!