Skip to main content

Scorer

model_selection_rs::scoring

Trait Scorer 

Source 
pub trait Scorer {
    // Required methods
    fn score(&self, y_true: &Array1<f64>, y_pred: &Array1<f64>) -> f64;
    fn name(&self) -> &str;

    // Provided method
    fn greater_is_better(&self) -> bool { ... }
}
Expand description

A scoring metric over true and predicted target vectors.

Implementors compute a single scalar from aligned y_true / y_pred. greater_is_better tells callers (e.g. model selection loops) which direction is an improvement.

Required Methods§

Source

fn score(&self, y_true: &Array1<f64>, y_pred: &Array1<f64>) -> f64

Compute the score. y_true and y_pred are assumed equal length.

Source

fn name(&self) -> &str

A short human-readable name (used to label CvResults columns).

Provided Methods§

Source

fn greater_is_better(&self) -> bool

Whether a larger score is better (e.g. accuracy, R²) or smaller is better (e.g. MAE, MSE, RMSE). Defaults to true.

Dyn Compatibility§

This trait is dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementors§