jin 0.1.0

Approximate Nearest Neighbor Search: HNSW, DiskANN, IVF-PQ, ScaNN, quantization
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54

//! Unified Approximate Nearest Neighbor (ANN) search algorithms.
//!
//! Pure Rust implementations of state-of-the-art ANN algorithms:
//! - **HNSW**: Hierarchical Navigable Small World (graph-based) - see `dense::hnsw`
//! - **AnisotropicVQ-kmeans**: Anisotropic Vector Quantization with k-means Partitioning
//!   (vendor name: SCANN/ScaNN) - see `dense::scann`
//! - **IVF-PQ**: Inverted File Index with Product Quantization - see `dense::ivf_pq`
//! - **DiskANN**: Disk-based ANN for very large datasets - see `dense::diskann`
//!
//! All algorithms are optimized with SIMD acceleration and minimal dependencies.
//!
//! # Index Factory
//!
//! Use `factory::index_factory()` to create indexes from string descriptions:
//!
//! ```rust,ignore
//! use jin::ann::{index_factory, ANNIndex};
//!
//! // Create HNSW index (requires "hnsw" feature)
//! let mut index = index_factory(128, "HNSW32")?;
//! let v0 = vec![0.1; 128];
//! index.add_slice(0, &v0)?;
//! index.build()?;
//! ```

// Autotune for automatic parameter optimization
#[cfg(any(
    feature = "hnsw",
    feature = "nsw",
    feature = "ivf_pq",
    feature = "scann",
    feature = "dense"
))]
pub mod autotune;

#[cfg(any(
    feature = "hnsw",
    feature = "nsw",
    feature = "ivf_pq",
    feature = "scann",
    feature = "dense"
))]
pub mod factory;
pub mod traits;

#[cfg(any(
    feature = "hnsw",
    feature = "nsw",
    feature = "ivf_pq",
    feature = "scann",
    feature = "dense"
))]
pub use factory::{index_factory, AnyANNIndex};
pub use traits::{ANNIndex, ANNStats};