Hamza098-code/Morocco-Water-Stress-Analysis

Data-driven analysis of water scarcity in Morocco. Utilizes Python for ETL/Viz and R for Linear Regression to forecast strategic water reserves through 2028. # 🇲🇦 Morocco Regional Water Stress Analysis & Prediction (2021-2028) ## 📌 Executive Summary This research project presents a longitudinal analysis of water stress evolution in Morocco (**2021-2025**), with statistical forecasting extending to **2028**. Developed as a technical portfolio for the **CERN Summer Student Programme 2026**, this repository demonstrates a **hybrid workflow** combining Python's data engineering capabilities with R's statistical precision. The study aggregates multi-source hydrological data to quantify the impact of the recent drought cycle and models the recovery trajectory following the **January 2026 hydro-climatic inflection point**. --- ## 📂 Repository Structure The project follows a modular structure to ensure reproducibility and separation of concerns: ```text Morocco-Water-Stress-Analysis/ ├── data/ │ ├── raw/ # Historical datasets (2014-2025) │ ├── processed/ # Cleaned regional metrics (Granular CSVs) │ └── predictions/ # R-generated Forecasts (2026-2028) & Performance Metrics ├── notebooks/ │ └── Water_Stress_Analysis.ipynb # Python Pipeline (ETL, EDA, Visualization, Reporting) ├── reports/figures/ # Visualization Gallery (PNG) │ regional_dashboard.png │ risk_ranking_2026.png │ national_trajectory.png ├── scripts/ │ └── regression_models.R # R Script for Pure Regression Modeling ├── requirements.txt # Python dependencies ├── install.R # R package installation script └── README.md # Documentation ``` --- ## 💻 Tech Stack: A Hybrid Approach To achieve optimal model accuracy, this project leverages the strengths of both ecosystems: | Component | Technology | Role in Project | | :--- | :--- | :--- | | **Data Engineering** | **Python** (Pandas) | Cleaning 10+ years of daily telemetry data and feature engineering. | | **Statistical Modeling** | **R Language** | Execution of pure linear regression algorithms for precise tr …