tidymodels/recipes

Pipeable steps for feature engineering and data preprocessing to prepare for modeling

What it solves

It provides a standardized, pipeable way to perform feature engineering and data preprocessing for machine learning models, overcoming the limitations of traditional R formula methods for creating design matrices.

How it works

Users define a "recipe" using a sequence of steps (such as step_normalize) to specify how predictors and outcomes should be transformed. These steps are applied to the data to prepare it for modeling or visualization.

Who it’s for

Data scientists and analysts using R who need a structured approach to prepare data for machine learning.

Highlights

  • Uses a dplyr-like syntax for pipeable sequences of steps.
  • Alternative to traditional R formulas and model.matrix for creating design matrices.
  • Integrates with the broader tidymodels ecosystem.

Related

  • Project
    kedro-org/kedro

    A Python framework for creating production-ready data engineering and data science pipelines using software engineering best practices.

  • Project
    tidymodels/tune

    An R package for the tidymodels ecosystem that facilitates hyperparameter tuning for machine learning models.

  • Project
    Pipelex/pipelex

    A framework for building and running reusable, typed AI procedures called Methods, enabling declarative and repeatable AI pipelines across multiple LLM providers.

  • Project
    terrier-org/pyterrier

    A framework for building and evaluating information retrieval pipelines, supporting sparse and dense retrieval, neural reranking, and RAG use-cases.

  • Project
    facebookresearch/recipes

    A library of ready-to-run deep learning training recipes that provide production-ready baselines for common tasks to eliminate repetitive glue code.