learning-machines-drift 0.0.2

A Python package for monitoring dataset drift in production ML pipelines.

Learning Machines

A Python package for monitoring dataset drift in production ML pipelines.

Built to run in any environment without uploading your data to external services.

Background

More background on learning machines.

Getting started

Requirements

• Python 3.9

Install

To install the latest version, run the following:

pip install -U learning-machines-drift

Example usage

A simple example along with the below:

from learning_machines_drift import Dataset, Display, FileBackend, Monitor, Registry
from learning_machines_drift.datasets import example_dataset

# Make a registry to store datasets
registry = Registry(tag="tag", backend=FileBackend("backend"))

# Save example reference dataset of 100 samples
registry.save_reference_dataset(Dataset(*example_dataset(100, seed=0)))

# Log example dataset with 80 samples
with registry:
    registry.log_dataset(Dataset(*example_dataset(80, seed=1)))

# Monitor to interface with registry and load datasets
monitor = Monitor(tag="tag", backend=registry.backend).load_data()

# Measure drift and display results as a table
Display().table(monitor.metrics.scipy_kolmogorov_smirnov())

Development

Install

For a local copy:

git clone [e-mail]:alan-turing-institute/learning-machines-drift
cd learning-machines-drift

To install:

poetry install

To install with dev and docs dependencies:

poetry install --with dev,docs

Tests

Run:

poetry run pytest

pre-commit checks

Run:

poetry run pre-commit run --all-files

To run checks before every commit, install as a pre-commit hook:

poetry run pre-commit install

Other tools

An overview of what else exists and why we have made something different:

• Cloud based

  • Azure dataset monitor

• Python

  • Evidently
  • whylogs

• ML pipelines: End to end machine learning lifecycle

  • MLFlow

What LM does differently

• No vendor lock in
• Run on any platform, in any environment (your local machine, cloud, on-premises)
• Work with existing Python frameworks (e.g. scikit-learn)
• Open source