Struct DefectPredictor 

pub struct DefectPredictor { /* private fields */ }

Defect prediction model using Random Forest

Predicts defect category (0-9) from commit features

Implementations

impl DefectPredictor

pub fn new() -> Self

Create new predictor with default parameters

pub fn with_params(n_trees: usize, max_depth: usize) -> Self

Create predictor with custom parameters

pub fn train(&mut self, features: &[CommitFeatures]) -> Result<()>

Train model on labeled features

Arguments

• features - Training features with defect_category labels

pub fn predict(&self, features: &CommitFeatures) -> Result<u8>

Predict defect category for new features

Uses k-NN approximation (k=5) for Phase 1 Full Random Forest in Phase 2 with aprender

pub fn predict_proba(&self, features: &CommitFeatures) -> Result<Vec<f32>>

Predict probabilities for all categories

pub fn params(&self) -> (usize, usize)

Get model parameters

pub fn is_trained(&self) -> bool

Check if model is trained

Trait Implementations

impl Default for DefectPredictor

fn default() -> Self

Returns the “default value” for a type.