Enum DistanceMetric 

pub enum DistanceMetric {
    Cosine,
    Euclidean,
    DotProduct,
}

Distance metric for similarity search.

Determines how vector similarity is computed. All metrics return lower values for more similar vectors.

Choosing a Metric

• Cosine: Best for text embeddings where magnitude doesn’t matter. Measures the angle between vectors. Range: [0, 2].
• Euclidean: Best for spatial data where absolute positions matter. Measures straight-line distance. Range: [0, ∞).
• DotProduct: Best for pre-normalized vectors or when magnitude should influence similarity. Range: (-∞, ∞), negated so lower = more similar.

Examples

use synadb::distance::DistanceMetric;

// Identical vectors have distance 0 for all metrics
let v = vec![1.0f32, 2.0, 3.0];
assert!(DistanceMetric::Cosine.distance(&v, &v) < 1e-6);
assert!(DistanceMetric::Euclidean.distance(&v, &v) < 1e-6);

Variants

Cosine

Cosine distance: 1 - cos(θ) where θ is the angle between vectors.

Range: [0, 2] where 0 = identical direction, 1 = orthogonal, 2 = opposite. Invariant to vector magnitude - only measures direction.

Euclidean

Euclidean (L2) distance: sqrt(Σ(a_i - b_i)²).

Range: [0, ∞) where 0 = identical vectors. Sensitive to both direction and magnitude.

DotProduct

Negative dot product: -Σ(a_i * b_i).

Range: (-∞, ∞), negated so lower values indicate higher similarity. For normalized vectors, equivalent to cosine similarity.

Implementations

impl DistanceMetric

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

Compute distance between two vectors.

Lower values indicate more similar vectors for all metrics.

Arguments

• a - First vector
• b - Second vector (must have same length as a)

Returns

The distance between the vectors according to this metric.

Panics

Debug builds will panic if vectors have different lengths. Release builds have undefined behavior for mismatched lengths.

Performance

O(n) where n is the vector dimension. Approximately 1μs for 768-dim vectors.

Examples

use synadb::distance::DistanceMetric;

let query = vec![0.1f32; 128];
let candidate = vec![0.2f32; 128];

let dist = DistanceMetric::Cosine.distance(&query, &candidate);
println!("Distance: {:.4}", dist);

Requirements: 1.4

pub fn from_u8(value: u8) -> Self

Convert from u8 representation

pub fn to_u8(&self) -> u8

Convert to u8 representation

Trait Implementations

impl Clone for DistanceMetric

fn clone(&self) -> DistanceMetric

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for DistanceMetric

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

Formats the value using the given formatter.

impl PartialEq for DistanceMetric

fn eq(&self, other: &DistanceMetric) -> 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 Copy for DistanceMetric

impl Eq for DistanceMetric

impl StructuralPartialEq for DistanceMetric