skrub-data/skrub

Health v2 · maintenance only

Adoption not part of the score

1650 stars · 275 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

skrub is a Python library that facilitates machine learning with dataframes, providing preprocessing, encoding, and wrangling tools for tabular and dirty data. It integrates with scikit-learn-style pipelines to turn messy dataframes into ML-ready features.

Use cases

• preprocess messy dataframes for machine learning
• encode high-cardinality categorical columns
• clean dirty tabular data before training models
• build sklearn pipelines on pandas or polars dataframes
• join and featurize multiple tables for ML
• prepare text and categorical columns for scikit-learn estimators

When to choose

• you train ML models on tabular data stored in dataframes
• you need robust encoders for messy categorical or text columns
• you want scikit-learn-compatible transformers for data preparation
• you work with pandas or polars and need feature engineering helpers

When to avoid

• you need deep learning on images or raw text rather than tabular data
• you want a full AutoML or model-training framework
• your data is not dataframe/tabular in nature

Facets

library · maturity active

machine-learning data-science etl machine-learning data-science python dataframes data-preprocessing data-cleaning scikit-learn dirty-data tabular-data

Member repositories

For agents

markdown · JSON · MCP: product_card(name="skrub-data/skrub")