Module drift 

Drift Detection Module

Implements Jidoka (Automation with Human Touch) for detecting when the process is out of control and signals for help (Retraining).

Provides statistical tests for detecting data drift and concept drift:

• Kolmogorov-Smirnov test (continuous features)
• Chi-square test (categorical features)
• Population Stability Index (PSI)

Structs

DriftDetector

Drift detector with statistical tests and callbacks

DriftResult

Drift detection result

DriftSummary

Summary of drift detection results

Enums

DriftTest

Statistical test for drift detection

Severity

Severity levels for drift

Functions

bin_counts

Count samples in bins defined by edges

chi_square_p_value

Approximate chi-square p-value using Wilson-Hilferty approximation

erf

Error function approximation

ks_p_value

Approximate p-value for KS statistic using Kolmogorov distribution

Type Aliases

CategoricalBaseline

Baseline data for categorical features (histogram per feature)

DriftCallback

Callback type for drift events (Andon Cord)