add CatBoost as a classifier option

[gbeane]

Add CatBoost Classifier Support

Summary

Adds CatBoost as a third classifier option alongside Random Forest and XGBoost, providing users with a high-accuracy gradient boosting alternative.

Changes

Core Implementation:

• Added ClassifierType.CATBOOST enum value
• Imported CatBoostClassifier directly (no conditional handling needed - CatBoost doesn't require libomp)
• Created _make_catboost() factory with appropriate parameters:
  • verbose=False to suppress training output
  • allow_writing_files=False to prevent intermediate files
• Updated train(), predict(), predict_proba(), and sort_features_to_classify() methods to handle CatBoost's native NaN support and feature_names_ attribute

Documentation:

• Added comprehensive "Choosing a Classifier Type" section to user guide
• Includes pros/cons, performance characteristics, and usage recommendations for all three classifiers
• Quick comparison table to help users make informed decisions

Key Features

• ✅ Native NaN/missing value support (like XGBoost)
• ✅ No external dependencies (unlike XGBoost's libomp requirement)
• ✅ Higher accuracy potential than Random Forest
• ⚠️ Slower training time (documented in user guide)

Testing

Unit tests for CatBoost will be added after PR #253 (test reorganization) is merged.

Files Modified

• src/jabs/types/classifier_types.py - Added CATBOOST enum
• src/jabs/classifier/classifier.py - Implemented CatBoost support
• src/jabs/resources/docs/user_guide/gui.md - Added classifier comparison guide

[Copilot]

Pull request overview

This PR adds CatBoost as a third classifier option to JABS, providing users with a high-accuracy gradient boosting alternative that natively handles missing data without external dependencies like XGBoost's libomp requirement.

Key Changes:

• Implemented CatBoost classifier support with native NaN handling
• Added comprehensive classifier comparison documentation to the user guide
• Updated prediction methods to properly handle CatBoost's feature name attribute

Reviewed changes

Copilot reviewed 3 out of 4 changed files in this pull request and generated 2 comments.

File	Description
src/jabs/classifier/classifier.py	Added CatBoost factory function, imported CatBoostClassifier, updated train/predict methods to handle CatBoost's NaN support and feature_names_ attribute
src/jabs/types/classifier_types.py	Added CATBOOST enum value (referenced but not shown in diff)
src/jabs/resources/docs/user_guide/gui.md	Added detailed "Choosing a Classifier Type" section with pros/cons comparison table and recommendations for all three classifiers
pyproject.toml	Added catboost>=1.2.8 dependency

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Comment is misleading - the code replaces infinities with NaN (not 0), and then CatBoost/XGBoost handle both the original NaN values and the newly converted infinity→NaN values. Consider revising to: 'XGBoost and CatBoost can handle NaN, so convert infinities to NaN instead of 0'

[Copilot]

Comment is misleading - the code replaces infinities with NaN (not 0), and then CatBoost/XGBoost handle both the original NaN values and the newly converted infinity→NaN values. Consider revising to: 'XGBoost and CatBoost can handle NaN, so convert infinities to NaN instead of 0'