Struct AdaptiveRandomForest 

pub struct AdaptiveRandomForest { /* private fields */ }

Available on crate feature alloc only.

Adaptive Random Forest for streaming classification.

An ensemble of n_trees streaming learners, each trained on a Poisson-weighted bootstrap with a random feature subspace. ADWIN drift detection automatically resets individual trees when their error rate changes significantly.

Example

use irithyll::{AdaptiveRandomForest, StreamingLearner};
use irithyll::ensemble::adaptive_forest::ARFConfig;
use irithyll::learners::linear::StreamingLinearModel;

let config = ARFConfig::builder(5).lambda(6.0).build().unwrap();
let mut arf = AdaptiveRandomForest::new(config, || {
    Box::new(StreamingLinearModel::new(0.01))
});

arf.train_one(&[1.0, 0.0], 0.0);
let pred = arf.predict(&[1.0, 0.0]);

Implementations

impl AdaptiveRandomForest

pub fn new<F>(config: ARFConfig, factory: F) -> Self

where F: Fn() -> Box<dyn StreamingLearner> + Send + Sync + 'static,

Create a new ARF with the given config and learner factory.

The factory closure is called n_trees times to create the initial ensemble, and again whenever a tree is replaced after drift detection.

pub fn train_one(&mut self, features: &[f64], target: f64)

Train on a single sample.

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

Predict by majority vote across all trees.

Each tree casts a vote for a class (prediction rounded to nearest integer). The class with the most votes wins.

pub fn predict_votes(&self, features: &[f64]) -> Vec<(f64, u64)>

Vote counts per predicted class.

pub fn n_trees(&self) -> usize

Number of trees in the ensemble.

pub fn n_samples_seen(&self) -> u64

Total samples processed.

pub fn tree_accuracies(&self) -> Vec<f64>

Per-tree accuracy (correct / evaluated).

pub fn n_drifts_detected(&self) -> usize

Total number of drift-triggered tree replacements.

Trait Implementations

impl StreamingLearner for AdaptiveRandomForest

fn train_one(&mut self, features: &[f64], target: f64, _weight: f64)

Train on a single observation with explicit sample weight.

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

Predict the target for the given feature vector.

fn n_samples_seen(&self) -> u64

Total number of observations trained on since creation or last reset.

fn reset(&mut self)

Reset the model to its initial (untrained) state.

fn train(&mut self, features: &[f64], target: f64)

Train on a single observation with unit weight.

fn predict_batch(&self, feature_matrix: &[&[f64]]) -> Vec<f64>

Predict for each row in a feature matrix.

fn diagnostics_array(&self) -> [f64; 5]

Raw diagnostic signals for adaptive tuning.

fn adjust_config(&mut self, _lr_multiplier: f64, _lambda_delta: f64)

Apply smooth learning rate and regularization adjustments.

fn apply_structural_change(&mut self, _depth_delta: i32, _steps_delta: i32)

Apply structural changes at model replacement boundaries.

fn replacement_count(&self) -> u64

Total number of internal model replacements (e.g. tree replacements triggered by drift detection or max-tree-samples).

fn check_proactive_prune(&mut self) -> bool

Manually trigger a proactive prune check.

fn set_prune_half_life(&mut self, _hl: usize)

Dynamically set the contribution accuracy EWMA half-life.

fn readout_weights(&self) -> Option<&[f64]>

Return the readout weight vector for supervised projection, if available.

fn tree_structure(&self) -> Vec<(usize, usize, f64, f64, u64)>

Optional tree-level structure diagnostics.