Struct LinearModel 

pub struct LinearModel { /* private fields */ }

A fitted linear model: an intercept plus one slope per predictor column.

Fields are private so the fitted parameters stay an invariant pair (the intercept and slopes always come from the same fit); read them through LinearModel::intercept and LinearModel::coefficients, and evaluate the model with LinearModel::predict / LinearModel::r_squared.

Implementations

impl LinearModel

pub const fn intercept(&self) -> f64

Returns the fitted intercept term (β₀).

pub fn coefficients(&self) -> &[f64]

Returns the fitted slope coefficients, one per predictor column in input column order.

pub fn predict(&self, row: &[f64]) -> f64

Predicts the target for a single observation row.

Computes β₀ + Σ_j β_j x_j. Coordinates of row beyond the fitted slope count are ignored, and missing coordinates contribute zero, so a row whose length differs from the training width still yields a finite prediction (callers are responsible for passing rows of the fitted dimension).

Arguments

• row — the predictor values for one observation, in fitted column order.

Returns

The model’s scalar prediction ŷ for row.

pub fn r_squared(&self, x: &[Vec<f64>], y: &[f64]) -> f64

Computes the coefficient of determination R² of the model on x/y.

R² = 1 − SS_res / SS_tot, where SS_res = Σ(yᵢ − ŷᵢ)² and SS_tot = Σ(yᵢ − ȳ)². Matches scikit-learn’s score: when SS_tot is zero (a constant target) the result is 1.0 if the residuals also vanish, else 0.0. R² can be negative when the model fits worse than the mean.

Arguments

• x — design matrix to score against, one inner slice per observation.
• y — observed targets, one per row of x.

Returns

The R² score in (−∞, 1.0]. An empty y yields 0.0.

Trait Implementations

impl Clone for LinearModel

fn clone(&self) -> LinearModel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LinearModel

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

Formats the value using the given formatter.

impl PartialEq for LinearModel

fn eq(&self, other: &LinearModel) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for LinearModel