Struct SparseEmbeddingSpace 

pub struct SparseEmbeddingSpace { /* private fields */ }

Metadata about a sparse embedding space.

A sparse embedding space defines the properties of sparse embeddings stored within it:

• The maximum dimension (vocabulary size for SPLADE, etc.)
• The distance metric used for similarity search

Unlike dense embedding spaces, sparse spaces don’t require a fixed dimension since sparse vectors only store non-zero elements.

Example

use manifoldb_vector::types::{EmbeddingName, SparseEmbeddingSpace};
use manifoldb_vector::distance::sparse::SparseDistanceMetric;

let name = EmbeddingName::new("splade_embedding").unwrap();
let space = SparseEmbeddingSpace::new(name, 30522, SparseDistanceMetric::DotProduct);

assert_eq!(space.max_dimension(), 30522); // BERT vocab size
assert_eq!(space.distance_metric(), SparseDistanceMetric::DotProduct);

Implementations

impl SparseEmbeddingSpace

pub const fn new( name: EmbeddingName, max_dimension: u32, distance_metric: SparseDistanceMetric, ) -> Self

Create a new sparse embedding space.

pub fn name(&self) -> &EmbeddingName

Get the name of the embedding space.

pub fn max_dimension(&self) -> u32

Get the maximum dimension (vocabulary size) for this space.

pub fn distance_metric(&self) -> SparseDistanceMetric

Get the distance metric used for similarity search.

pub fn to_bytes(&self) -> Result<Vec<u8>, VectorError>

Encode the sparse embedding space to bytes.

Format:

• 1 byte: version (2 for sparse)
• 2 bytes: name length (big-endian u16)
• N bytes: name (UTF-8)
• 4 bytes: max dimension (big-endian u32)
• 1 byte: distance metric

Errors

Returns an error if the name length exceeds u16::MAX.

pub fn from_bytes(bytes: &[u8]) -> Result<Self, VectorError>

Decode a sparse embedding space from bytes.

Errors

Returns an error if the bytes are invalid or truncated.

Trait Implementations

impl Clone for SparseEmbeddingSpace

fn clone(&self) -> SparseEmbeddingSpace

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for SparseEmbeddingSpace

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

Formats the value using the given formatter.

impl PartialEq for SparseEmbeddingSpace

fn eq(&self, other: &SparseEmbeddingSpace) -> 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 Eq for SparseEmbeddingSpace

impl StructuralPartialEq for SparseEmbeddingSpace