Struct PolynomialFit 

pub struct PolynomialFit {
    pub coefficients: [f64; 3],
    pub confidence: f64,
}

The rest of the animation layer: tweens chained and curved, the scroll simulations and their constants. A polynomial fitted to a set of weighted points, and how well it fits.

Only what the velocity estimate needs: degree ≤ 2, so three coefficients.

Fields

coefficients: [f64; 3]

coefficients[i] multiplies xⁱ.

confidence: f64

The fraction of the data’s variance the fit explains (an r-squared), in 0..=1. Higher is better.

Implementations

impl PolynomialFit

pub fn solve(x: &[f64], y: &[f64], degree: usize) -> Option<PolynomialFit>

Fits a polynomial of degree (≤ 2) to (x, y) by least squares, all points weighted equally. None when there is not enough data, or when the points are degenerate (all at the same instant, say).

The method is a QR decomposition by Gram–Schmidt, then back-substitution — the standard approach, and numerically far better behaved than forming and inverting the normal equations. It runs in f64 even though frus is f32 throughout: orthogonalising powers of the abscissae loses precision quickly, and this is the number every fling depends on.

Trait Implementations

impl Clone for PolynomialFit

fn clone(&self) -> PolynomialFit

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for PolynomialFit

impl Debug for PolynomialFit

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

Formats the value using the given formatter.

impl PartialEq for PolynomialFit

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

Equality operator ==.

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

Inequality operator !=.

impl StructuralPartialEq for PolynomialFit