Documentation
¶
Overview ¶
Package similarity provides vector distance functions for semantic search.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func CosineDistance ¶
CosineDistance computes the cosine distance between two vectors. Returns a value in [0, 2] where 0 means identical direction and 2 means opposite direction. Lower values indicate more similar vectors.
func CosineSimilarity ¶
CosineSimilarity computes the cosine similarity between two vectors. Returns a value in [-1, 1] where 1 means identical direction, 0 means orthogonal, and -1 means opposite direction. Both vectors must have the same length.
func NewEmbeddingClient ¶ added in v0.10.1
func NewEmbeddingClient(cfg *types.OptimizerConfig) (types.EmbeddingClient, error)
NewEmbeddingClient creates an EmbeddingClient from the given optimizer configuration. It returns (nil, nil) if cfg is nil or no embedding service URL is configured, meaning semantic search will be disabled.
Types ¶
type FakeEmbeddingClient ¶
type FakeEmbeddingClient struct {
// contains filtered or unexported fields
}
FakeEmbeddingClient is a deterministic embedding client for testing. It hashes input text with SHA-256 and uses the hash as a seed to generate reproducible float32 vectors. The vectors are L2-normalized to unit length.
func NewFakeEmbeddingClient ¶
func NewFakeEmbeddingClient(dimension int) *FakeEmbeddingClient
NewFakeEmbeddingClient creates a FakeEmbeddingClient that produces vectors of the given dimension.
func (*FakeEmbeddingClient) Close ¶
func (*FakeEmbeddingClient) Close() error
Close is a no-op for the fake client.
func (*FakeEmbeddingClient) Embed ¶
Embed returns a deterministic, unit-normalized vector for the given text.
func (*FakeEmbeddingClient) EmbedBatch ¶
EmbedBatch returns deterministic embeddings for each input text.
Source Files