Struct KernelDensityEstimator 

pub struct KernelDensityEstimator {
    pub data: Vec<f64>,
    pub bandwidth: f64,
}

Kernel density estimator using a Gaussian kernel.

For a dataset x_1, …, x_n, the KDE at point x is: f̂(x) = (1/(n·h)) Σ_i K((x - x_i)/h) where K is the standard Gaussian kernel.

Fields

data: Vec<f64>

Training data points.

bandwidth: f64

Bandwidth h (Silverman’s rule of thumb by default).

Implementations

impl KernelDensityEstimator

pub fn new(data: Vec<f64>) -> Self

Creates a KDE with Silverman’s rule-of-thumb bandwidth: h = 1.06 · σ̂ · n^{-1/5}.

pub fn with_bandwidth(data: Vec<f64>, bandwidth: f64) -> Self

Creates a KDE with an explicit bandwidth.

pub fn density(&self, x: f64) -> f64

Evaluates the kernel density estimate at point x.

pub fn density_grid(&self, lo: f64, hi: f64, m: usize) -> Vec<(f64, f64)>

Evaluates the KDE over a grid of m equally spaced points in [lo, hi].