Enum DistanceFunction 

pub enum DistanceFunction {
    L2,
    Cosine,
    NegativeDotProduct,
}

Distance function for vector similarity

Variants

L2

L2 / Euclidean distance

Cosine

Cosine distance (1 - cosine similarity)

NegativeDotProduct

Negative inner product (for maximum inner product search)

Implementations

impl DistanceFunction

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

Compute distance between two vectors

Uses SIMD-accelerated implementations when simd feature is enabled. Falls back to optimized scalar implementations otherwise.

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

Compute distance for internal comparisons (optimized, may not be actual distance)

For L2: returns squared distance (skips sqrt) since relative ordering is preserved For Cosine/NegativeDotProduct: same as distance()

This is ~10-15% faster for L2 in HNSW search_layer.

pub fn comparison_to_actual(&self, d: f32) -> f32

Convert comparison distance back to actual distance

For L2: applies sqrt For others: identity

Trait Implementations

impl Clone for DistanceFunction

fn clone(&self) -> DistanceFunction

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for DistanceFunction

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for DistanceFunction

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl From<DistanceFunction> for Metric

fn from(d: DistanceFunction) -> Self

Converts to this type from the input type.

impl From<Metric> for DistanceFunction

fn from(m: Metric) -> Self

Converts to this type from the input type.

impl PartialEq for DistanceFunction

fn eq(&self, other: &DistanceFunction) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for DistanceFunction

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

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Copy for DistanceFunction

impl Eq for DistanceFunction

impl StructuralPartialEq for DistanceFunction