# Forecasting Financial Inclusion in Ethiopia
Selam Analytics — forecasting system for **Access** (Account Ownership) and **Usage** (Digital Payment Adoption) in Ethiopia, aligned with the World Bank Global Findex framework.
## Overview
- **Access**: Share of adults (15+) with an account at a financial institution or using mobile money (past 12 months).
- **Usage**: Share of adults who made or received a digital payment (past 12 months).
The system:
1. **Loads and enriches** the unified financial inclusion dataset (observations, events, targets, impact links).
2. **Analyzes** patterns and event–indicator relationships.
3. **Models** how events (product launches, policy, infrastructure) affect inclusion via `impact_link` records.
4. **Forecasts** Access and Usage for 2025–2027 (trend + event-impact adjustments).
5. **Presents** results in an interactive dashboard.
## Data
- **Unified dataset**: `data/raw/ethiopia_fi_unified_data.xlsx`
- Sheet **ethiopia_fi_unified_data**: `observation`, `event`, `target`.
- Sheet **Impact_sheet**: `impact_link` (event → indicator effects, lag, magnitude).
- **Reference codes**: `data/raw/reference_codes.xlsx`.
- **Enrichment guide**: `data/raw/Additional Data Points Guide.xlsx` (alternative baselines, direct/indirect indicators, market nuances).
Enrichment (Task 1) adds:
- 2011 Account Ownership (14%) for continuity.
- Digital Payment Adoption Rate (Usage) for 2021 and 2024.
- Placeholder structure for direct/indirect indicators (agent density, ATM/branch density, smartphone penetration, mobile internet) to be filled from IMF FAS, GSMA, ITU, NBE.
## Setup
```bash
# From project root; use a virtual environment if possible
python3 -m venv .venv
source .venv/bin/activate # or .venv\Scripts\activate on Windows
pip install -r requirements.txt
```
Dependencies include `pandas`, `openpyxl`, `numpy`, `scikit-learn`, `statsmodels`, `dash`, `plotly`, and `python-docx` for report generation.
## Usage
### Load and enrich data
`` …