Struct RidgeClassifierCV 

pub struct RidgeClassifierCV<F> {
    pub alphas: Vec<F>,
    pub fit_intercept: bool,
    pub store_cv_results: bool,
}

Ridge classifier with built-in cross-validated alpha selection.

Selects a single shared regularization strength alpha from a candidate grid by leave-one-out Generalized Cross-Validation over the binarized indicator targets, then refits a multi-output Ridge at the chosen alpha. Mirrors scikit-learn’s RidgeClassifierCV (sklearn/linear_model/_ridge.py:2676).

Type Parameters

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

Fields

alphas: Vec<F>

Candidate regularization strengths to evaluate (sklearn alphas, default (0.1, 1.0, 10.0), _ridge.py:2688).

fit_intercept: bool

Whether to fit an intercept (bias) term (sklearn fit_intercept, default True, _ridge.py:2698).

store_cv_results: bool

Whether to retain the per-sample cross-validation results (cv_results_). Mirrors sklearn store_cv_results (documented default False, _ridge.py:2547/:2727; the ctor sentinel None resolves to False, _ridge.py:2349-2350). When true, the fitted model exposes the per-sample-per-target-per-alpha squared leave-one-out errors via FittedRidgeClassifierCV::cv_results.

Implementations

impl<F: Float + FromPrimitive> RidgeClassifierCV<F>

pub fn new() -> Self

Create a new RidgeClassifierCV with default settings.

Defaults: alphas = [0.1, 1.0, 10.0] and fit_intercept = true, mirroring sklearn’s ctor defaults (sklearn/linear_model/_ridge.py:2688, :2698).

pub fn with_alphas(self, alphas: Vec<F>) -> Self

Set the candidate regularization strengths (sklearn alphas, _ridge.py:2688).

Each value must be non-negative.

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

Set whether to fit an intercept term (sklearn fit_intercept, _ridge.py:2698).

pub fn with_store_cv_results(self, store_cv_results: bool) -> Self

Set whether to retain the cross-validation results (sklearn store_cv_results, default False, _ridge.py:2547/:2727).

When true, the fitted model’s FittedRidgeClassifierCV::cv_results returns the per-sample-per-target-per-alpha squared leave-one-out errors; when false (the default) it returns None.

Trait Implementations

impl<F: Clone> Clone for RidgeClassifierCV<F>

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

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug> Debug for RidgeClassifierCV<F>

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

Formats the value using the given formatter.

impl<F: Float + FromPrimitive> Default for RidgeClassifierCV<F>

fn default() -> Self

Returns the “default value” for a type.

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

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

Fit the RidgeClassifierCV model.

Binarizes y into a {-1, +1} indicator matrix (binary → single column, multiclass → one-hot), selects a single shared alpha by leave-one-out Generalized Cross-Validation over that multi-target problem (mirroring sklearn’s _RidgeGCV path on the binarized Y, _ridge.py:2876-2881; scoring=None → -squared_errors.mean(), _ridge.py:2148-2150/:2211-2218), then refits a multi-output Ridge at the chosen alpha.

Errors

• FerroError::ShapeMismatch if x and y have different numbers of samples.
• FerroError::InvalidParameter if x contains a non-finite value (NaN/±Inf), or if alphas is empty or contains a non-positive value (alpha <= 0; the GCV path is undefined at alpha = 0).
• FerroError::InsufficientSamples if there are no samples or fewer than two distinct classes.

type Fitted = FittedRidgeClassifierCV<F>

The fitted model type returned by fit.

type Error = FerroError

The error type returned by fit.