Struct BinaryVectorStore 

pub struct BinaryVectorStore { /* private fields */ }

Stores f32 vectors binarized to 1 bit per dimension.

Each dimension is thresholded at the per-vector mean: bit = 1 if v[i] >= mean(v) else 0.

Bits are packed into u64 words (64 dims per word), so a 384-dim vector occupies ceil(384/64) = 6 words = 48 bytes, vs 1536 bytes for f32.

Similarity is measured with Hamming distance (popcount of XOR).

Implementations

impl BinaryVectorStore

pub fn new(dim: usize) -> Self

Create an empty store for vectors of dim dimensions.

pub fn push(&mut self, vector: &[f32]) -> usize

Binarize vector (threshold at mean) and store it.

Returns the assigned ID.

pub fn hamming_distance(&self, a_id: usize, b_id: usize) -> u32

Compute the Hamming distance between two stored vectors.

Hamming distance = number of bit positions where the two vectors differ. Returns u32::MAX if either id is out of range.

pub fn bytes_per_vector(&self) -> f64

Approximate memory per stored vector (bytes).

dim_words * 8 bytes per vector, e.g. dim=384 → 6 × 8 = 48 bytes.

pub fn len(&self) -> usize

Number of stored vectors.

pub fn is_empty(&self) -> bool

Returns true if no vectors have been stored.

pub fn dim(&self) -> usize

Vector dimension.