Struct VectorJoinOperator 

pub struct VectorJoinOperator { /* private fields */ }

Vector join operator for hybrid graph + vector queries.

Takes entities from the left input and finds similar entities on the right based on vector embeddings. This enables queries like:

MATCH (u:User)-[:LIKED]->(liked:Movie)
WITH avg(liked.embedding) AS taste
VECTOR JOIN (m:Movie) ON m.embedding
WHERE cosine_similarity(m.embedding, taste) > 0.7
RETURN m.title

Output Schema

Output columns: [left columns…, right_node: Node, score: Float64]

Implementations

impl VectorJoinOperator

pub fn with_static_query( left: Box<dyn Operator>, store: Arc<LpgStore>, query_vector: Vec<f32>, right_property: impl Into<String>, k: usize, metric: DistanceMetric, ) -> Self

Creates a vector join with a static query vector.

Use this when the query vector is known upfront (e.g., a parameter).

Arguments

• left - Left input operator
• store - LPG store for vector property lookup
• query_vector - The query vector for similarity search
• right_property - Property containing right-side embeddings
• k - Number of nearest neighbors per left row
• metric - Distance metric

pub fn entity_to_entity( left: Box<dyn Operator>, store: Arc<LpgStore>, left_node_column: usize, left_property: impl Into<String>, right_property: impl Into<String>, k: usize, metric: DistanceMetric, ) -> Self

Creates a vector join for entity-to-entity similarity.

For each entity on the left, fetches its embedding and finds similar entities on the right.

Arguments

• left - Left input operator
• store - LPG store for property lookup
• left_node_column - Column index containing NodeId in left input
• left_property - Property containing left-side embeddings
• right_property - Property containing right-side embeddings
• k - Number of nearest neighbors per left entity
• metric - Distance metric

pub fn with_right_label(self, label: impl Into<String>) -> Self

Sets a label filter for right-side entities.

pub fn with_min_similarity(self, threshold: f32) -> Self

Sets minimum similarity threshold (cosine metric only).

pub fn with_max_distance(self, threshold: f32) -> Self

Sets maximum distance threshold.

pub fn with_ef(self, ef: usize) -> Self

Sets the HNSW search ef parameter.

pub fn with_chunk_capacity(self, capacity: usize) -> Self

Sets the output chunk capacity.

Trait Implementations

impl Operator for VectorJoinOperator

fn next(&mut self) -> OperatorResult

Pulls the next batch of data. Returns None when exhausted.

fn reset(&mut self)

Resets to initial state so you can iterate again.

fn name(&self) -> &'static str

Returns a name for debugging/explain output.