Metadata-Version: 2.4
Name: light-mlt
Version: 0.1.1
Summary: Lightweight preprocessing and reversible Modular Linear Tokenization (MLT) utilities for categorical and continuous data.
Author-email: Tcharlies Schmitz <tcharliesschmitz@gmail.com>
License: MIT
Project-URL: Homepage, https://pypi.org/project/light-mlt/
Project-URL: Repository, https://github.com/tcharliesschmitz/light-mlt
Project-URL: Bug Tracker, https://github.com/tcharliesschmitz/light-mlt/issues
Project-URL: Documentation, https://github.com/tcharliesschmitz/light-mlt#readme
Description-Content-Type: text/markdown

# 🧩 light-mlt


# 🧩 light-mlt

[![PyPI version](https://img.shields.io/pypi/v/light-mlt?color=orange)](https://pypi.org/project/light-mlt/)
[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)
[![GitHub](https://img.shields.io/badge/github-light--mlt-black?logo=github)](https://github.com/tcharliesschmitz/light-mlt)



---

**Lightweight preprocessing and reversible Modular Linear Tokenization (MLT) utilities for categorical and continuous data.**

---

## ✨ Overview

`light-mlt` implements **Modular Linear Tokenization (MLT)** — a deterministic and reversible method for encoding high-cardinality categorical identifiers into compact numerical vectors.

Unlike hashing or one-hot encodings, **MLT guarantees bijective mappings**, provides **explicit dimensionality control**, and integrates seamlessly with **machine learning pipelines**.

It was developed as part of applied research on scalable tokenization and efficient preprocessing for tabular and recommendation systems.

---

## 🚀 Installation

```bash
pip install light-mlt
