Struct ClosedFormLeaf 

pub struct ClosedFormLeaf { /* private fields */ }

Available on crate feature alloc only.

Leaf model that computes the optimal weight in closed form: weight = -grad_sum / (hess_sum + lambda).

This is the standard leaf value used in gradient boosted trees.

Implementations

impl ClosedFormLeaf

pub fn new() -> Self

Create a new zeroed closed-form leaf.

Trait Implementations

impl Default for ClosedFormLeaf

fn default() -> Self

Returns the “default value” for a type.

impl LeafModel for ClosedFormLeaf

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

Produce a prediction given input features.

fn update( &mut self, _features: &[f64], gradient: f64, hessian: f64, lambda: f64, )

Update model parameters given a gradient, hessian, and regularization lambda.

fn clone_fresh(&self) -> Box<dyn LeafModel>

Create a fresh (zeroed / re-initialized) clone of this model’s architecture.

fn clone_warm(&self) -> Box<dyn LeafModel>

Create a warm clone preserving learned weights but resetting optimizer state.