samoraowino/co2-emissions-africa-pipeline

# CO₂ Emissions Data Pipeline for African Countries This project presents an end-to-end data engineering pipeline focused on analyzing and storing carbon emissions data for African nations. It includes data ingestion, cleaning, normalization into a relational database, and exploratory data analysis (EDA) through visualizations. ## Project Summary - **Objective**: To build a clean and structured data pipeline for CO₂ emissions data in Africa and generate meaningful insights. - **Data Source**: Publicly available CSV dataset on CO₂ emissions. - **Tools Used**: Python, SQLite, pandas, matplotlib, seaborn, DBeaver. ## Project Components ### 1. Data Ingestion - Read raw CSV file using `pandas`. - Initial inspection and schema review. ### 2. Data Cleaning & Transformation - Handled missing values. - Converted column types (e.g., years to integers). - Created new columns such as CO₂ emissions per capita. ### 3. Database Design (Normalization) - Structured the database into 3 normalized tables: - `Country`: Stores country name, code, and sub-region. - `Year`: Holds the distinct years. - `Emissions`: Main fact table with foreign keys to `Country` and `Year`. ### 4. SQL Integration - Created and populated tables using SQLite. - SQL dump provided (`co2_emissions_dump.sql`) for reproducibility. ### 5. Exploratory Data Analysis (EDA) Visualizations created include: - CO₂ Emissions Over Time (per country) - Emissions Per Capita Over Time - GDP vs CO₂ Emissions Scatter Plot - Top 5 CO₂ Emitters in Latest Year - Emissions by Sub-Region ### 6. Data Ethics Reflection - Addressed underreporting biases. - Justified cleaning assumptions and per capita normalization. - Reflected on the importance of transparent and equitable data use.