Struct DriftDetector 

pub struct DriftDetector { /* private fields */ }

Drift detector using Z-score analysis.

Detects performance regressions using sliding window baseline comparison. Based on statistical process control principles.

Example

use jugar_web::loadtest::DriftDetector;

let mut detector = DriftDetector::new(10, 2.0);

// Calibrate with baseline samples
detector.calibrate(&[1.0, 1.1, 0.9, 1.0, 1.05, 0.95, 1.0, 1.0, 0.98, 1.02]);

// Normal observation
let result = detector.observe(1.0);
assert!(result.is_normal());

// Anomalous observation (way outside normal range)
let result = detector.observe(10.0);
assert!(result.is_anomaly());

Implementations

impl DriftDetector

pub fn new(window_size: usize, z_threshold: f64) -> Self

Create a new drift detector.

Arguments

• window_size - Number of samples in the sliding window
• z_threshold - Z-score threshold for anomaly detection (typical: 2.0-3.0)

pub fn with_drift_threshold(self, percent: f64) -> Self

Set the drift threshold percentage.

pub fn calibrate(&mut self, samples: &[f64])

Calibrate the detector with baseline samples.

This establishes the “normal” baseline for comparison.

pub fn is_calibrated(&self) -> bool

Check if the detector has been calibrated.

pub fn observe(&mut self, frame_time_ms: f64) -> AnomalyResult

Add a sample and check for anomaly.

pub fn detect_drift(&self) -> Option<DriftReport>

Detect sustained drift (window mean significantly different from baseline).

pub fn window_stats(&self) -> Option<(f64, f64)>

Get current window statistics.

pub fn reset(&mut self)

Reset the detector (clear window, keep calibration).

pub fn full_reset(&mut self)

Full reset (clear window and calibration).

Trait Implementations

impl Debug for DriftDetector

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

Formats the value using the given formatter.