Struct Lts 

pub struct Lts { /* private fields */ }

A configured Least Trimmed Squares estimator.

Like crate::regression::SEstimator, it is reproducible by default (a fixed-seed rand_chacha::ChaCha8Rng sub-stream per subsample); configure with Lts::seed / Lts::fit_with_rng.

Implementations

impl Lts

pub fn new() -> Self

A max-breakdown LTS with default search settings.

pub fn coverage(self, fraction: f64) -> Self

Retain a fraction ∈ (0, 1] of the observations instead of the max-breakdown default; the coverage is h = ⌊fraction · n⌋ (clamped to [p, n]). Larger fraction trades breakdown for efficiency.

pub fn n_subsamples(self, n: usize) -> Self

Set the number of random elemental subsets (default 500).

pub fn seed(self, seed: u64) -> Self

Set the master RNG seed (default is a fixed internal constant).

pub fn control(self, control: Control) -> Self

Override the concentration-step convergence control.

pub fn fit( &self, x: &Array2<f64>, y: &Array1<f64>, ) -> Result<LtsFit, RobustError>

Fit reproducibly from the configured seed.

pub fn fit_with_rng<G: Rng>( &self, x: &Array2<f64>, y: &Array1<f64>, rng: &mut G, ) -> Result<LtsFit, RobustError>

Fit drawing the master seed from a caller-supplied generator.

Trait Implementations

impl Clone for Lts

fn clone(&self) -> Lts

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for Lts

impl Debug for Lts

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

Formats the value using the given formatter.

impl Default for Lts

fn default() -> Self

Max-breakdown coverage, 500 subsamples.