ChimezieFavour/AnalystLab-Africa-ETL-Weather-Pipeline

# ETL Pipeline: Real-Time Weather Data Analysis **AnalystLab Africa Internship Program | Batch A | Week 7** **Author:** Chimezie Favour Chidinma **Date:** June 2026 ## Project Overview This project builds a simple ETL (Extract, Transform, Load) pipeline using real-time weather data from the OpenWeather API. The pipeline automatically collects live weather data for 5 cities around the world, cleans and transforms it into a structured format, and stores it for analysis. ## Data Source - **API:** OpenWeather Current Weather API - **Website:** openweathermap.org - **Cities covered:** Lagos, London, New York, Tokyo, Nairobi ## ETL Process ### Extract - Connected to the OpenWeather API using Python's requests library - Sent HTTP GET requests with an API key and city name as parameters - Received raw JSON responses containing nested weather data for each city ### Transform - Parsed nested JSON fields into a flat structured format - Extracted 14 relevant fields including temperature, humidity, wind speed, and timestamp - Converted UNIX timestamps to readable datetime format - Verified zero missing values and correct data types across all columns ### Load - Saved the cleaned dataset in three formats: - weather_data.csv — for general use and sharing - weather_data.xlsx — for Excel-based reporting - weather_data.db — SQLite database for structured querying ## Tools Used - Python - requests - pandas - matplotlib - seaborn - sqlite3 - openpyxl - Jupyter Notebook ## Key Findings - Warmest city: London (25.1°C) — surprisingly warmer than Lagos on the day of extraction - Coolest city: New York (20.6°C) - Most humid: Lagos (91%) — consistent with its coastal tropical climate - Least humid: Nairobi (55%) — reflecting its drier highland climate - Windiest city: London (5.36 m/s) - Weather condition: All 5 cities recorded cloudy conditions at time of extraction ## What I Learned - How to authenticate and connect to a real-world REST API - How to parse nested …