Struct OrthogonalMatchingPursuit 

pub struct OrthogonalMatchingPursuit {
    pub max_k: usize,
}

Orthogonal Matching Pursuit (OMP) for sparse signal recovery.

Greedily selects the most correlated column of the measurement matrix at each step and performs a least-squares fit on the selected support.

Fields

max_k: usize

Maximum sparsity (number of non-zero coefficients to recover).

Implementations

impl OrthogonalMatchingPursuit

pub fn new(max_k: usize) -> Self

Create an OMP solver with sparsity bound max_k.

pub fn solve(&self, a: &[Vec<f64>], b: &[f64]) -> Vec<f64>

Recover a sparse signal from measurements.

• a — measurement matrix (m × n, row-major)
• b — measurement vector (length m)

Returns the recovered coefficient vector of length n.

pub fn support(&self, a: &[Vec<f64>], b: &[f64]) -> Vec<usize>

Return the support (indices of non-zero components) for sparsity level k.

Runs OMP for exactly k steps and returns the selected column indices.