sagemaker-scikit-learn-extension

What it is and what it does

SageMaker Scikit-Learn Extension is a collection of scikit-learn-compatible estimators and transformers built to fill gaps in scikit-learn's standard library and support AWS SageMaker's Autopilot automated machine learning service. It provides specialized tools for dimension reduction, feature extraction from datetime and time-series data, text vectorization, missing-value imputation with custom masking logic, and categorical encoding schemes. The package is designed as a repository for estimators that don't meet scikit-learn's strict inclusion criteria but are useful in production ML pipelines.

The project is dormant—the last release was 2022-02-18, and it has not been updated since then. It is tested on Python 3.7 only. The package integrates with SageMaker's training containers and includes utilities for feature and target transformation, making it a legacy tool primarily suited for maintaining existing SageMaker workflows rather than new projects.

Use it for:

• Preprocess tabular data for SageMaker Autopilot by applying imputation and categorical encoding transformers.
• Extract numeric features from datetime columns for time-aware machine learning models.
• Handle extreme values and heavy-tailed distributions with specialized transformers.
• Encode categorical features using supervised or similarity-based encoding methods.
• Perform dimension reduction on sparse matrices without dense conversion.
• Extract features from time-series sequence data using provided extractors.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Extends scikit-learn with additional estimators and preprocessing tools designed to support SageMaker Autopilot, including dimension reduction, feature extraction, imputation, and encoding transformers.

No, unless you are maintaining an existing SageMaker Autopilot workflow. The package is dormant (last release 2022-02-18), has high install friction (requires conda for mlio version 0.7), and is tested only on Python 3.7. Compatibility with modern Python and scikit-learn versions is unverified. For new projects, use scikit-learn directly or actively-maintained alternatives.

Install

sagemaker-scikit-learn-extension on PyPI

pip install sagemaker-scikit-learn-extension

uv add sagemaker-scikit-learn-extension

poetry add sagemaker-scikit-learn-extension

Installing sagemaker-scikit-learn-extension

Before you install

High install friction: the package is distributed as a source tarball and requires conda to install an optional but commonly-needed dependency (mlio version 0.7). Maintenance is dormant—last release was 2022-02-18, with no updates since then.

License in practice

Licensed under Apache License 2.0 (permissive). You may use, modify, and distribute this package freely in commercial and private projects, provided you include a copy of the license.

Quickstart

# Install from pip
pip install sagemaker-scikit-learn-extension

# For I/O functionality, also install mlio via conda
conda install -c mlio -c conda-forge mlio-py==0.7

# Example usage with imputation and preprocessing tools
pip install --upgrade .[test]

mlio version 0.7 is only available via conda, not pip. Package is tested on Python 3.7; compatibility with newer versions is unspecified.

Verify before relying

• Whether the package works reliably on Python versions newer than 3.7.
• Current compatibility with recent scikit-learn versions, given dormant maintenance since 2022-02-18.
• Whether all submodules install without additional manual steps beyond the documented conda requirement.

Package facts

Evidence: sagemaker-scikit-learn-extension-2.5.0.tar.gz

Keywords: ML, Amazon, AWS, AI, SKLearn, Scikit-Learn

Tags

More Artificial Intelligence packages

Further reading