First release

📦 pretab v0.0.1 – Initial Release

We’re excited to introduce pretab, a flexible and extensible preprocessing library for tabular data, built on top of scikit-learn.

🚀 Features

• 🔢 Numerical preprocessing via:

  • Discretization with custom binning
  • Tree-based transformations (PLE, CustomBin)
  • Spline expansions: Cubic, Natural, P-splines, Thin Plate Splines, Tensor Product Splines
  • Neural-inspired bases: RBF, ReLU, Sigmoid, Tanh
  • All available sklearn transformations:
    • Standard, MinMax, Quantile, and Robust scaling
    • Polynomial features, Box-Cox, and Yeo-Johnson transforms

• 🧠 Categorical preprocessing via:

  • Ordinal, one-hot, and float encodings
  • Language embeddings (optional)

• 🧱 Integrated with sklearn's ColumnTransformer and Pipeline

• 🔍 Detailed feature metadata via get_feature_info()

• 📦 Easy integration of embeddings alongside transformed features

💬 Community

This is an open project – contributions, feature requests, and bug reports are welcome via GitHub!