Struct SparseProjection 

pub struct SparseProjection { /* private fields */ }

Sparse random projection matrix for dimensionality expansion

Uses a sparse random matrix to project low-dimensional inputs into high-dimensional space while maintaining computational efficiency.

Properties

• Sparse connectivity (typically 10-20% connections)
• Gaussian-distributed weights
• Deterministic (seeded) for reproducibility

Performance

• Time complexity: O(input_dim × output_dim × sparsity)
• Space complexity: O(input_dim × output_dim)

Implementations

impl SparseProjection

pub fn new( input_dim: usize, output_dim: usize, sparsity: f32, seed: u64, ) -> Result<Self>

Create a new sparse random projection

Arguments

• input_dim - Input vector dimension
• output_dim - Output vector dimension (should be >> input_dim)
• sparsity - Connection probability (typically 0.1-0.2)
• seed - Random seed for reproducibility

Example

use ruvector_nervous_system::SparseProjection;

let projection = SparseProjection::new(128, 10000, 0.15, 42);

pub fn project(&self, input: &[f32]) -> Result<Vec<f32>>

Project input vector to high-dimensional space

Arguments

• input - Input vector of size input_dim

Returns

Output vector of size output_dim

Example

use ruvector_nervous_system::SparseProjection;

let projection = SparseProjection::new(128, 10000, 0.15, 42).unwrap();
let input = vec![1.0; 128];
let output = projection.project(&input).unwrap();
assert_eq!(output.len(), 10000);

pub fn input_dim(&self) -> usize

Get input dimension

pub fn output_dim(&self) -> usize

Get output dimension

pub fn sparsity(&self) -> f32

Get sparsity level

Trait Implementations

impl Clone for SparseProjection

fn clone(&self) -> SparseProjection

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for SparseProjection

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

Formats the value using the given formatter.