Struct SelectFwe 

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

Select features controlling the family-wise error rate via the Bonferroni correction.

A feature is selected if its p-value is strictly less than alpha / n_features.

Examples

use ferrolearn_preprocess::stat_selectors::SelectFwe;
use ferrolearn_core::traits::{Fit, Transform};
use ndarray::array;

let sel = SelectFwe::<f64>::new(0.05);
let p_values = array![0.001, 0.5, 0.03, 0.9];
let fitted = sel.fit(&p_values, &()).unwrap();
// Bonferroni threshold = 0.05/4 = 0.0125; only feature 0 qualifies
assert_eq!(fitted.selected_indices(), &[0]);

Implementations

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

pub fn new(alpha: f64) -> Self

Create a new SelectFwe with the given significance level.

pub fn alpha(&self) -> f64

Return the significance level.

Trait Implementations

impl<F: Clone> Clone for SelectFwe<F>

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

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<F: Debug> Debug for SelectFwe<F>

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

Formats the value using the given formatter.

impl<F: Float + Send + Sync + 'static> Fit<ArrayBase<OwnedRepr<F>, Dim<[usize; 1]>>, ()> for SelectFwe<F>

fn fit(&self, x: &Array1<F>, _y: &()) -> Result<FittedSelectFwe<F>, FerroError>

Fit using the Bonferroni correction: p < alpha / n_features.

Errors

• FerroError::InvalidParameter if p-values are empty or alpha is not in (0, 1].

type Fitted = FittedSelectFwe<F>

The fitted model type returned by fit.

type Error = FerroError

The error type returned by fit.