Struct BinaryQuantizer 

pub struct BinaryQuantizer {
    pub dimension: usize,
    pub thresholds: Vec<f32>,
}

Binary quantizer

Maps float32 values to binary {0, 1}, achieving 32x compression. Each dimension becomes a single bit based on sign or threshold.

Fields

dimension: usize

Vector dimension

thresholds: Vec<f32>

Threshold per dimension (typically mean or median)

Implementations

impl BinaryQuantizer

pub fn train(vectors: &[Vec<f32>]) -> Result<Self>

Train quantizer using mean as threshold

pub fn train_sign_based(dimension: usize) -> Self

Train using zero threshold (sign-based binarization)

pub fn encode(&self, vector: &[f32]) -> Result<Vec<u8>>

Encode a vector to binary representation

Returns packed bits: each byte stores 8 dimensions

pub fn hamming_distance(&self, a: &[u8], b: &[u8]) -> u32

Hamming distance between binary vectors

Counts differing bits (fast XOR + popcount)

pub fn approximate_cosine(&self, a: &[u8], b: &[u8]) -> f32

Approximate cosine similarity from Hamming distance

cos(θ) ≈ 1 - 2 * (hamming_distance / dimension)

pub fn memory_usage(&self, num_vectors: usize) -> usize

Memory footprint in bytes

Trait Implementations

impl Clone for BinaryQuantizer

fn clone(&self) -> BinaryQuantizer

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for BinaryQuantizer

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for BinaryQuantizer

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for BinaryQuantizer

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.