feat(graphql): Add support for ef and distance_threshold in generated GraphQL queries for similarity search #9562
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
github-actions
bot
added
area/testing
matthewmcneely
commented
Jan 9, 2026
|
|
||
| if metric == schema.SimilarSearchMetricDotProduct { | ||
| distanceFormula = "math(( 1.0 - (($search_vector) dot v2)) /2.0)" | ||
| distanceFormula = "math(1.0 - (($search_vector) dot v2))" |
Contributor
Author
There was a problem hiding this comment.
@raphael-istari For some reason, the calculated distances for cosine and dp were being divided by 2. Not sure why, maybe to normalize them to [0,1] instead of their natural [0,2] (which is how they're actually stored in the index). The distance_threshold param exposes this oddness. So this (a breaking change to GraphQL queries) computes them as they're stored--and the distance_threshold parameter actually works as expected.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
DO NOT MERGE
This PR adds support for the new effort and distance threshold in generated GraphQL queries for types that have embedding vectors.
Note. There's a breaking change in the resulting computed distance in that for cosine and dotproduct indexes, the computed distance is no longer divided by 2. The results now correctly represent the distances stored in the index.
Checklist
Conventional Commits syntax, leading
with
fix:,feat:,chore:,ci:, etc.docs repo staged and linked here. This process can
be simplified by going to the public docs site and clicking the
"Edit this page" button at the bottom of page(s) relevant to your changes. Ensure that you
indicate in the PR that this is an unreleased feature so that it does not get merged into
the main docs prematurely.