Struct DdmTest 

pub struct DdmTest<A: Float + Send + Sync> { /* private fields */ }

DDM (Drift Detection Method), Gama et al. 2004.

Tracks the running error probability p_i and its binomial standard deviation s_i = sqrt(p_i (1 - p_i) / i), remembers the minimum of p_i + s_i seen so far, and signals a warning at p_i + s_i >= p_min + 2 s_min and drift at p_i + s_i >= p_min + 3 s_min. The reported statistic is the number of s_min units the current error level sits above p_min, and the p-value is the one-sided normal tail of that z score.

Implementations

impl<A: Float + Send + Sync> DdmTest<A>

pub fn new(sensitivity: f64, significance_level: f64) -> Result<Self, String>

Creates a DDM detector. sensitivity scales the published 2-sigma / 3-sigma warning and drift levels, so a smaller value makes the detector fire earlier.

pub fn is_warning(&self) -> bool

Whether DDM is currently in its warning zone (between the 2-sigma and 3-sigma levels).

Trait Implementations

impl<A: Float + Default + Clone + Send + Sync + Sum> StatisticalTest<A> for DdmTest<A>

fn test_for_drift( &mut self, reference: &[A], current: &[A], ) -> Result<DriftTestResult<A>, String>

Performs the statistical test for drift

fn update_parameters(&mut self, performance_feedback: A) -> Result<(), String>

Updates test parameters based on historical performance

fn reset(&mut self)

Resets the test state