Metadata-Version: 2.4
Name: pyg-captum-shap
Version: 0.1.0
Summary: A robust SHAP explainer wrapper for PyTorch Geometric models.
Author-email: Author Name <author@example.com>
Project-URL: Homepage, https://github.com/username/pyg-captum-shap
Project-URL: Bug Tracker, https://github.com/username/pyg-captum-shap/issues
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Scientific/Engineering :: Chemistry
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: torch>=2.0.0
Requires-Dist: torch_geometric>=2.0.0
Requires-Dist: captum>=0.6.0
Dynamic: license-file

# PyG-SHAP

A robust wrapper bridging the Euclidean sampling mechanisms of Captum with the non-Euclidean batching of PyTorch Geometric (PyG). Designed specifically for complex QSAR modelling, including multi-task toxicity prediction and molecular property prediction (e.g., DPD-Cancer models achieving target MCCs of 0.6010).

## Installation
```bash
pip install pyg-captum-shap
