Reference library

ML engineering pipelines

Feature prep, batch inference, model-serving hooks, and production ML workflow glue.

3 matches

Detect Concept Drift in Python with a Simple Statistical Test

Detect concept drift by comparing the mean of recent data against a reference distribution using a z-score-like threshold.

concept drift statistics ml monitoring

Python

import random
import statistics

def detect_drift(recent, reference, threshold=1.5):
    ref_mean = statistics.mean(reference)
    ref_std = statistics.stdev(reference)
    
    recent_mean = statistics.mean(recent)
    drift_score = abs(recent_mean - ref_mean) / (ref_std if ref_std > 0 else 1)
    
    drifted = drif…

15 0 Open

ML engineering pipelines medium

How to Detect Data Drift with PSI in Python

Calculate the Population Stability Index (PSI) in Python to compare expected vs actual distributions and detect data drift in machine learning pipelines.

data drift psi monitoring

Python

import numpy as np

def calculate_psi(expected, actual, buckets=10):
    """Calculate Population Stability Index (PSI) between two distributions."""
    # Create bucket edges based on expected distribution percentiles
    edges = np.percentile(expected, np.linspace(0, 100, buckets + 1))
    edges[-1] = np.inf  # Ensur…

13 0 Open

ML engineering pipelines easy

How to Trigger Model Retraining on Drift in Python

Automatically detects accuracy drift in a mock ML model and triggers retraining when performance falls below a threshold.

ml drift-detection retraining

Python

import random
import time

class MockModel:
    def __init__(self, name):
        self.name = name
        self.accuracy = 0.85
        self.version = 1

    def train(self, data_size):
        # Simulate training time and accuracy improvement
        time.sleep(0.1)
        drift = random.uniform(-0.02, 0.02)
       …

16 0 Open

Browse by section

Each section groups closely related Python snippets.