Struct OrthogonalMatchingPursuit 

pub struct OrthogonalMatchingPursuit<F> {
    pub n_nonzero_coefs: Option<usize>,
    pub tol: Option<F>,
    pub fit_intercept: bool,
}

Orthogonal Matching Pursuit.

A greedy sparse approximation algorithm that selects features one at a time. At each iteration it picks the feature most correlated with the residual, adds it to the support, solves OLS on the support set, and re-computes the residual.

Termination is controlled by either n_nonzero_coefs (maximum support size) or tol (residual norm threshold), whichever is reached first.

Type Parameters

• F: The floating-point type (f32 or f64).

Fields

n_nonzero_coefs: Option<usize>

Maximum number of non-zero coefficients. Defaults to None (use all features or stop at tol).

tol: Option<F>

Residual norm tolerance. If the squared residual norm drops below this threshold the algorithm terminates. Defaults to None.

fit_intercept: bool

Whether to fit an intercept (bias) term.

Implementations

impl<F: Float> OrthogonalMatchingPursuit<F>

pub fn new() -> Self

Create a new OrthogonalMatchingPursuit with default settings.

Defaults: n_nonzero_coefs = None, tol = None, fit_intercept = true.

pub fn with_n_nonzero_coefs(self, n: usize) -> Self

Set the maximum number of non-zero coefficients.

pub fn with_tol(self, tol: F) -> Self

Set the residual norm tolerance.

pub fn with_fit_intercept(self, fit_intercept: bool) -> Self

Set whether to fit an intercept term.

Trait Implementations

impl<F: Clone> Clone for OrthogonalMatchingPursuit<F>

fn clone(&self) -> OrthogonalMatchingPursuit<F>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug> Debug for OrthogonalMatchingPursuit<F>

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

Formats the value using the given formatter.

impl<F: Float> Default for OrthogonalMatchingPursuit<F>

fn default() -> Self

Returns the “default value” for a type.

impl<F: Float + Send + Sync + ScalarOperand + FromPrimitive + 'static> Fit<ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>, ArrayBase<OwnedRepr<F>, Dim<[usize; 1]>>> for OrthogonalMatchingPursuit<F>

fn fit(&self, x: &Array2<F>, y: &Array1<F>) -> Result<FittedOMP<F>, FerroError>

Fit the OMP model.

Greedily selects features by correlation with the residual and solves OLS on the growing support set.

Errors

• FerroError::ShapeMismatch — sample count mismatch.
• FerroError::InsufficientSamples — zero samples.
• FerroError::InvalidParameter — n_nonzero_coefs exceeds features, or neither n_nonzero_coefs nor tol is set.

type Fitted = FittedOMP<F>

The fitted model type returned by fit.

type Error = FerroError

The error type returned by fit.

impl<F> PipelineEstimator<F> for OrthogonalMatchingPursuit<F>

where F: Float + FromPrimitive + ScalarOperand + Send + Sync + 'static,

fn fit_pipeline( &self, x: &Array2<F>, y: &Array1<F>, ) -> Result<Box<dyn FittedPipelineEstimator<F>>, FerroError>

Fit this estimator on the given data.