Struct DecomposedPrediction 

pub struct DecomposedPrediction {
    pub location_base: f64,
    pub scale_base: f64,
    pub location_contributions: Vec<f64>,
    pub scale_contributions: Vec<f64>,
}

Available on crate feature alloc only.

Decomposed prediction showing each tree’s contribution.

Fields

location_base: f64

Base location prediction (mean of initial targets).

scale_base: f64

Base scale prediction (log-sigma of initial targets).

location_contributions: Vec<f64>

Per-step location contributions: learning_rate * tree_prediction. location_base + sum(location_contributions) = μ.

scale_contributions: Vec<f64>

Per-step scale contributions: learning_rate * tree_prediction. scale_base + sum(scale_contributions) = log(σ).

Implementations

impl DecomposedPrediction

pub fn mu(&self) -> f64

Reconstruct the final μ from base + contributions.

pub fn log_sigma(&self) -> f64

Reconstruct the final log(σ) from base + contributions.

pub fn sigma(&self) -> f64

Reconstruct the final σ (exponentiated).

Trait Implementations

impl Clone for DecomposedPrediction

fn clone(&self) -> DecomposedPrediction

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for DecomposedPrediction

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

Formats the value using the given formatter.