Struct Rfecv 

pub struct Rfecv {
    pub min_features_to_select: usize,
    pub step: usize,
    pub cv_folds: usize,
    /* private fields */
}

Recursive Feature Elimination with Cross-Validated selection of the optimal number of features.

Mirrors sklearn.feature_selection.RFECV. For each candidate n_features_to_select in min..=n_features, runs Rfe on each k-fold split, scores on the held-out fold, averages — picks the size with the highest mean CV score, then runs RFE on the full data to that size.

Fields

min_features_to_select: usize

step: usize

cv_folds: usize

Implementations

impl Rfecv

pub fn new<I, S>( min_features_to_select: usize, importance_fn: I, score: S, ) -> Self

where I: Fn(&Array2<f64>, &Array1<f64>) -> Result<Array1<f64>> + Send + Sync + 'static, S: Fn(&Array2<f64>, &Array1<f64>) -> Result<f64> + Send + Sync + 'static,

pub fn with_cv_folds(self, k: usize) -> Self

pub fn with_step(self, step: usize) -> Self

pub fn fit(&self, x: &Array2<f64>, y: &Array1<f64>) -> Result<FittedRfecv>