Add Hemmah Pro IVS 2025 valuation module (data engine, ML, dashboard, PDF reports)

[Moshbbab]

Motivation

• Provide a self-contained IVS 2025‑compliant real-estate valuation module to support data ingestion, quality checks, modeling and formal reporting.
• Enable Arabic-friendly reporting and UI to produce officially formatted PDF valuation reports and signed outputs for local/regional workflows.
• Bundle ML workflows and explainability (SHAP) plus an interactive dashboard to streamline IVS 101/102/103/104/105 tasks in one entrypoint.

Description

• Add hemmah_pro_ivs_2025.py which implements HemmahDataEngine, HemmahMLEngine, HemmahDashboard, and HemmahReportGenerator and the IVS_2025_FRAMEWORK metadata.
• Implement robust data handling and feature engineering including multi-encoding CSV loading, completeness/uniqueness/outlier checks, price/area cleaning, price_per_sqm creation, location target encoding and area categorization.
• Add ML training and selection pipeline that trains Random Forest, XGBoost and Gradient Boosting models with cross-validation, stores best model and feature importance, supports SHAP explanations, predict and sensitivity_analysis helpers.
• Provide interactive Jupyter widgets dashboard for data/model/valuation/report flows and a PDF report generator with Arabic reshaping and RTL handling, automatic font fallback and a structured IVS report (scope, basis, data quality, methodology, valuation, sensitivity and compliance sections). The module includes Colab-friendly silent installs and Amiri font download logic.

Testing

• No automated tests were executed against the new module as part of this change.

---

Codex Task