Module vector 

Vector embeddings and similarity search

This module provides vector storage and approximate nearest neighbor search using IVF (Inverted File) and PQ (Product Quantization) algorithms.

Components

• types - Core types (VectorStoreConfig, VectorManifest, Fragment, etc.)
• store - Columnar vector store with fragment-based storage
• distance - Distance functions (cosine, euclidean, dot product)
• ivf - IVF index for approximate nearest neighbor search
• pq - Product quantization for vector compression
• ivf_pq - Combined IVF-PQ index for efficient approximate nearest neighbor search

Re-exports

pub use distance::batch_cosine_distance;

pub use distance::batch_dot_product_distance;

pub use distance::batch_squared_euclidean;

pub use distance::cosine_distance;

pub use distance::cosine_similarity;

pub use distance::dot_product;

pub use distance::dot_product_at;

pub use distance::euclidean_distance;

pub use distance::l2_norm;

pub use distance::normalize;

pub use distance::normalize_in_place;

pub use distance::squared_euclidean;

pub use distance::squared_euclidean_at;

pub use ivf::IvfError;

pub use ivf::IvfIndex;

pub use ivf::IvfStats;

pub use ivf::KMeansConfig;

pub use ivf::KMeansError;

pub use ivf::KMeansResult;

pub use ivf::SearchOptions;

pub use ivf_pq::deserialize_ivf_pq;

pub use ivf_pq::ivf_pq_serialized_size;

pub use ivf_pq::serialize_ivf_pq;

pub use ivf_pq::IvfPqConfig;

pub use ivf_pq::IvfPqError;

pub use ivf_pq::IvfPqIndex;

pub use ivf_pq::IvfPqSearchOptions;

pub use ivf_pq::IvfPqStats;

pub use pq::PqError;

pub use pq::PqIndex;

pub use pq::PqSearchResult;

pub use pq::PqStats;

pub use store::create_vector_store;

pub use store::vector_store_batch_insert;

pub use store::vector_store_clear;

pub use store::vector_store_delete;

pub use store::vector_store_fragment_stats;

pub use store::vector_store_get;

pub use store::vector_store_get_all_vectors;

pub use store::vector_store_get_by_id;

pub use store::vector_store_get_location;

pub use store::vector_store_get_node_id;

pub use store::vector_store_get_vector_id;

pub use store::vector_store_has;

pub use store::vector_store_insert;

pub use store::vector_store_seal_active;

pub use store::vector_store_stats;

pub use store::FragmentStats;

pub use store::VectorStoreError;

pub use store::VectorStoreStats;

pub use types::DistanceMetric;

pub use types::Fragment;

pub use types::FragmentState;

pub use types::IvfConfig;

pub use types::MultiQueryAggregation;

pub use types::PqConfig;

pub use types::RowGroup;

pub use types::VectorLocation;

pub use types::VectorManifest;

pub use types::VectorSearchResult;

pub use types::VectorStoreConfig;

Modules

compaction

Fragment compaction to remove deleted vectors

distance

Distance functions with SIMD acceleration

fragment

Fragment management

ivf

IVF (Inverted File) index

ivf_pq

IVF-PQ: Combined Inverted File Index with Product Quantization

normalize

Vector normalization utilities

pq

Product Quantization (PQ) for vector compression and fast distance computation

row_group

Row group operations

store

Columnar vector store

types

Vector embedding types