abigiyacodehub/Forecasting-Financial-Inclusion-in-Ethiopia

# Forecasting Financial Inclusion in Ethiopia > Predicting the trajectory of financial inclusion in Ethiopia by combining sparse survey data with a rich catalogue of policy events, product launches, and regulatory changes — connected through a unified, schema-validated dataset. --- ## Project Overview Ethiopia's financial inclusion story is one of rapid but uneven progress. Account ownership grew from **22 % (2011) → 22 % (2014) → 35 % (2017) → 46 % (2021) → 49 % (2024)**, yet a 3-percentage-point gain over 2021–2024 sits alongside the explosive rise of Telebirr (launched May 2021) and the entry of Safaricom/M-Pesa (August 2023). Understanding *why* growth slowed — and what will drive it next — requires linking quantitative indicators to their causal events. This project builds that link using a **unified schema** that stores observations, events, targets, and causal impact_links in a single dataset, then applies forecasting techniques suited to limited, sparse time-series data. --- ## Repository Structure ``` . ├── data/ │ ├── raw/ # Unmodified source files │ │ ├── ethiopia_fi_unified_data.xlsx │ │ ├── reference_codes.xlsx │ │ └── Additional Data Points Guide.xlsx │ ├── processed/ # Cleaned & enriched outputs │ │ ├── ethiopia_fi_unified_data.csv # CSV export of raw data │ │ └── ethiopia_fi_enriched.csv # + new records from enrichment │ ├── data_enrichment_log.md # Audit trail for all additions │ └── README.md # Detailed schema documentation │ ├── notebooks/ │ ├── 01_data_exploration_enrichment.ipynb # Task 1 — load, explore, enrich │ └── 02_eda_financial_inclusion.ipynb # Task 2 — full EDA & insights │ ├── src/ │ ├── __init__.py │ ├── data_loader.py # Load, filter, and join dataset functions │ ├── schema_utils.py # Validation, record constructors, schema docs │ └── visualization.py # Reusable Matplotlib chart functi …