xgboost-tuner-pack 0.1.5

A hyperparameter tuner for XGBoost.

xgbtuner

A world-class, easy-to-use hyperparameter tuner for XGBoost models, built on top of scikit-learn’s GridSearchCV and RandomizedSearchCV. Supports both classification and regression out of the box, with robust error handling, comprehensive testing, and flexible customization.

Demo Usage: https://colab.research.google.com/drive/1eSbeIKlVeUo-0_PLUtr0lBi3WxlBG9st#scrollTo=yybTYeBVr5Nh

---

Features

• Grid Search & Random Search
  Choose between exhaustive grid search or randomized search for hyperparameter exploration.
• Classification & Regression
  One API for both objectives—simply set objective="classification" or "regression".
• Default & Custom Grids
  Sensible default parameter grid, plus ability to pass your own param_grid.
• Robust Input Validation
  Checks for array-like inputs, consistent lengths, and raises clear errors.
• Detailed Logging
  Built-in logging statements to trace tuning progress and errors.
• Fully Tested
  Over 10 unit tests cover edge cases, custom grids, list inputs, invalid configs, and more.
• Scikit-Learn Compatible
  Behaves like any estimator: .tune(), .predict(), and .predict_proba() for classification.

---

Installation

Install from PyPI:

pip install xgboost-tuner-pack