Struct FittedSparsePCA 

pub struct FittedSparsePCA<F> { /* private fields */ }

A fitted Sparse PCA model holding the learned components.

Created by calling Fit::fit on a SparsePCA. Implements Transform<Array2<F>> to project new data onto the sparse components.

Implementations

impl<F: Float + Send + Sync + 'static> FittedSparsePCA<F>

pub fn components(&self) -> &Array2<F>

Sparse components, shape (n_components, n_features).

pub fn mean(&self) -> &Array1<F>

Per-feature mean learned during fitting.

pub fn n_iter(&self) -> usize

Number of outer iterations performed.

Trait Implementations

impl<F: Clone> Clone for FittedSparsePCA<F>

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

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug> Debug for FittedSparsePCA<F>

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

Formats the value using the given formatter.

impl<F: Float + Send + Sync + 'static> Transform<ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>> for FittedSparsePCA<F>

fn transform(&self, x: &Array2<F>) -> Result<Array2<F>, FerroError>

Least-squares projection of the data onto the sparse components.

Because the sparse components_ are not orthonormal, a plain linear projection is incorrect (see scikit-learn _BaseSparsePCA.transform, _sparse_pca.py:93-123). Instead this computes the ridge-regularised least-squares fit U = ridge_regression(components_.T, X_centered.T, ridge_alpha), i.e. U = (X - mean) · Cᵀ · (C·Cᵀ + ridge_alpha·I)⁻¹ with C = components_ and ridge_alpha = 0.01 (sklearn default).

Errors

• Returns FerroError::ShapeMismatch if the number of columns does not match the number of features seen during fitting.
• Returns FerroError::NumericalInstability if the ridge system C·Cᵀ + ridge_alpha·I is singular (should not occur for ridge_alpha > 0).

type Output = ArrayBase<OwnedRepr<F>, Dim<[usize; 2]>>

The transformed output type.

type Error = FerroError

The error type returned by transform.