Struct HistGradientBoostingRegressor 

#[non_exhaustive]
pub struct HistGradientBoostingRegressor { /* private fields */ }

Histogram-based Gradient Boosting for regression.

Uses pre-binned features and O(256) histogram scans for split finding, delivering 5-10× speedup over standard GBT on large datasets. This is the same algorithmic approach as LightGBM/XGBoost/CatBoost, implemented in pure Rust with no external BLAS dependency.

Example

use scry_learn::dataset::Dataset;
use scry_learn::tree::HistGradientBoostingRegressor;

let features = vec![vec![1.0, 2.0, 3.0, 4.0, 5.0]];
let target = vec![2.0, 4.0, 6.0, 8.0, 10.0];
let data = Dataset::new(features, target, vec!["x".into()], "y");

let mut model = HistGradientBoostingRegressor::new()
    .n_estimators(100)
    .learning_rate(0.1)
    .max_leaf_nodes(31);
model.fit(&data).unwrap();

let preds = model.predict(&[vec![3.0]]).unwrap();
assert!((preds[0] - 6.0).abs() < 1.0);

Implementations

impl HistGradientBoostingRegressor

pub fn new() -> Self

Create a new regressor with default parameters.

Example

use scry_learn::tree::HistGradientBoostingRegressor;

let model = HistGradientBoostingRegressor::new()
    .n_estimators(200)
    .learning_rate(0.05);

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

Set number of boosting rounds (default: 100).

pub fn learning_rate(self, lr: f64) -> Self

Set learning rate / shrinkage (default: 0.1).

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

Set maximum number of leaf nodes per tree (default: 31).

This controls tree complexity. LightGBM default is 31.

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

Set minimum samples required in a leaf (default: 20).

pub fn max_depth(self, d: usize) -> Self

Set maximum tree depth (default: 8). Acts as a secondary depth limit.

pub fn max_bins(self, bins: usize) -> Self

Set maximum number of bins (2..=256, default: 256).

pub fn l2_regularization(self, l2: f64) -> Self

Set L2 regularization (default: 0.0).

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

Set random seed (default: 42).

pub fn fit(&mut self, data: &Dataset) -> Result<()>

Train the histogram-based gradient boosting regressor.

pub fn predict(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict values for new samples.

features is row-major: features[sample_idx][feature_idx].

pub fn feature_importances(&self) -> Result<Vec<f64>>

Feature importances (total gain, normalized).

pub fn n_trees(&self) -> usize

Number of trees in the ensemble.

pub fn n_features(&self) -> usize

Number of features the model was trained on.

pub fn learning_rate_val(&self) -> f64

Learning rate value.

pub fn init_prediction_val(&self) -> f64

Initial (base) prediction value.

pub fn tree_node_views(&self) -> Vec<Vec<HistNodeView>>

Convert internal HistTree nodes to public HistNodeView arrays for ONNX export. Bin thresholds are converted to raw feature thresholds using the binner.

Trait Implementations

impl Clone for HistGradientBoostingRegressor

fn clone(&self) -> HistGradientBoostingRegressor

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Default for HistGradientBoostingRegressor

fn default() -> Self

Returns the “default value” for a type.

impl PipelineModel for HistGradientBoostingRegressor

fn fit(&mut self, data: &Dataset) -> Result<()>

Train the model on a dataset.

fn predict(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict on row-major feature matrix.

impl Tunable for HistGradientBoostingRegressor

fn set_param(&mut self, name: &str, _value: ParamValue) -> Result<()>

Apply a named hyperparameter.

fn clone_box(&self) -> Box<dyn Tunable>

Clone this model into a boxed trait object.

fn fit(&mut self, data: &Dataset) -> Result<()>

Train on a dataset.

fn predict(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict on row-major features.