Enum CovKernel 

#[non_exhaustive]
pub enum CovKernel {
    Gaussian {
        length_scale: f64,
        variance: f64,
    },
    Exponential {
        length_scale: f64,
        variance: f64,
    },
    Matern {
        length_scale: f64,
        variance: f64,
        nu: f64,
    },
    Brownian {
        variance: f64,
    },
    Periodic {
        length_scale: f64,
        variance: f64,
        period: f64,
    },
    Linear {
        variance: f64,
        offset: f64,
    },
    Polynomial {
        variance: f64,
        offset: f64,
        degree: u32,
    },
    WhiteNoise {
        variance: f64,
    },
    Sum(Box<CovKernel>, Box<CovKernel>),
    Product(Box<CovKernel>, Box<CovKernel>),
}

Covariance kernel specification.

Each variant encodes a family of covariance functions k(s, t). Kernels can be composed with Sum and Product to build richer covariance structures.

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

Gaussian

Squared-exponential (RBF) kernel: variance * exp(-0.5 * ((s-t)/length_scale)^2).

Fields

length_scale: f64

variance: f64

Exponential

Exponential (Ornstein-Uhlenbeck) kernel: variance * exp(-|s-t| / length_scale).

Fields

length_scale: f64

variance: f64

Matern

Matern kernel with smoothness parameter nu.

Closed-form expressions are used for nu = 0.5 (exponential), nu = 1.5, and nu = 2.5. For other values of nu the general formula with a gamma-function approximation is used.

Fields

length_scale: f64

variance: f64

nu: f64

Brownian

Brownian motion (Wiener process) kernel: variance * min(s, t).

Fields

variance: f64

Periodic

Periodic kernel: variance * exp(-2 * sin^2(pi * |s-t| / period) / length_scale^2).

Fields

length_scale: f64

variance: f64

period: f64

Linear

Linear kernel: variance * (s - offset) * (t - offset).

Fields

variance: f64

offset: f64

Polynomial

Polynomial kernel: (variance * s * t + offset)^degree.

Fields

variance: f64

offset: f64

degree: u32

WhiteNoise

White noise kernel: variance * delta(s, t).

Fields

variance: f64

Sum(Box<CovKernel>, Box<CovKernel>)

Sum of two kernels: k1(s,t) + k2(s,t).

Product(Box<CovKernel>, Box<CovKernel>)

Product of two kernels: k1(s,t) * k2(s,t).

Implementations

impl CovKernel

pub fn eval(&self, s: f64, t: f64) -> f64

Evaluate the covariance function k(s, t).

Trait Implementations

impl Clone for CovKernel

fn clone(&self) -> CovKernel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for CovKernel

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

Formats the value using the given formatter.

impl PartialEq for CovKernel

fn eq(&self, other: &CovKernel) -> 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 CovKernel