Explainable Investment Analytics Framework (EIAF)

EIAF is a lightweight Python framework for generating audit-ready explainability artifacts for tabular investment / credit-risk models built with scikit-learn.

It produces:

• Global explanations (permutation importance)
• Local “what-if” explanations for individual records
• Reason codes (human-readable drivers)
• Stability metrics (PSI drift checks)
• A Model Card (Markdown) + a structured artifact bundle (JSON)

Why this exists

In regulated finance, teams often need consistent, repeatable explainability outputs for:

• model risk management (MRM)
• governance reviews
• audit traceability
• stakeholder communication

EIAF focuses on standardized artifacts, not just plots.

Install (editable)

pip install -e ".[dev]"


## How to cite this framework

If you use EIAF in research, documentation, or production workflows, please cite:

**Explainable Investment Analytics Framework (EIAF)** — Deepak Saxena, v0.1.0, 2026.  
GitHub repository: `https://github.com/saxenade/explainable-investment-analytics`

### BibTeX
```bibtex
@software{saxena_eiaf_2026,
  author  = {Saxena, Deepak},
  title   = {Explainable Investment Analytics Framework (EIAF)},
  year    = {2026},
  version = {0.1.0},
  url     = {https://github.com/saxenade/explainable-investment-analytics}
}