shap

shap

Model interpretability using SHAP (SHapley Additive exPlanations) — feature importance, waterfall/beeswarm/bar plots, bias analysis.

Setup

cd shap && python3 -m venv .venv && source .venv/bin/activate && pip install shap numpy scikit-learn -q

Usage

python3 scripts/demo.py --help
python3 scripts/demo.py --example tree --format json

Dependencies

shap, numpy, scikit-learn

Tested with

• Agno (dry-run): ✅