Weather Forecasting System (ML + FastAPI + Docker + CI)

Production-grade weather prediction pipeline using machine learning, real-time API data collection, FastAPI-based model deployment, and automation with GitHub Actions — all powered by CSV file storage only.

---

1. Project Motivation

This project simulates an end-to-end ML production system under minimal infrastructure constraints (no databases). It focuses on reproducible engineering workflows, live data integration, and clean, testable deployment.

---

2. Features

• Hourly live weather data ingestion using the Open-Meteo Archive API.
• Supervised ML model predicting next-hour temperature
• FastAPI for inference
• Dockerized app containerization
• CI pipeline with Pytest and Docker build check
• CSV-based storage (no SQL/NoSQL database used)
• Logging and error handling with loguru and Pydantic validation

---

3. Project Structure

weather-predictor/
├── app/
│   ├── __init__.py
│   ├── main.py              # FastAPI app
│   └── model.py             # ML model loader and predictor
├── data/
│   ├── historical_weather.csv
│   ├── DATA_FRAME.csv
│   └── live_weather.csv
├── model/
│   └── weather_pipeline.pkl # Trained model artifact
├── scripts/
│   ├── fetch_and_save.py    # Historical data collector
│   ├── load_data.py         # Data loader
│   ├── feature_engineer.py  # Feature extraction
│   ├── train_model.py       # Model training script
│   ├── evaluate_model.py    # Evaluation metrics
│   └── live_collector.py    # Live data appender
├── tests/
│   ├── test_data_validation.py
│   └── test_train_model.py
├── .github/workflows/ci.yml # GitHub Actions CI
├── Dockerfile
├── requirements.txt
├── .dockerignore
└── README.md

---

4. Tech Stack

• ML: scikit-learn, pandas, joblib
• API: FastAPI, Pydantic, Uvicorn
• Data Collection: requests, tenacity, dotenv
• Visualization: seaborn, matplotlib
• Automation: GitHub Actions, Pytest
• Packaging: Docker

---

5. Local Execution(FastAPI)

5.1 Environment Setup

conda create -n weatherapi python=3.12 -y
conda activate weatherapi

5.2 Clone the Repository

git clone https://github.com/Amaan-developpeur/WeatherPredictionAPI.git
cd WeatherPredictionAPI

5.3 Install Dependencies

pip install -r requirements.txt

5.4 Run the API Server

uvicorn app.main:app --reload

Open: http://127.0.0.1:8000/docs

---

6. Setup Instructions

6.1 Clone the Repository

git clone https://github.com/Amaan-developpeur/weather-predictor.git
cd weather-predictor

6.2 Install Requirements

pip install -r requirements.txt

6.3 Environment Variables

Create a .env file with:

OPENWEATHERMAP_API_KEY = "https://archive-api.open-meteo.com/v1/archive" 
[To do API Call]

---

7. Running Tests

pytest -v

---

8. Docker Usage

docker build -t weather-api .
docker run --rm -p 8000:8000 weather-api

Open: http://localhost:8000/docs

---

9. Docker CLI Output

This shows the container running and exposing the FastAPI app:

---

10. Continuous Integration

• Workflow: .github/workflows/ci.yml
• Stages: checkout → setup Python 3.11 → install dependencies → run tests → (optional) Docker build

11. API Usage

11.1 Example Input

{
  Directly hit the execute buttoon, it will fetch the live data and do prediction about the next hour.

11.2 Example Output

{
  "predicted_temperature": 32.1
}

Visit Swagger docs at: http://localhost:8000/docs

---

12. Images

12.1 Swagger UI

This shows the interactive API documentation served at /docs:

12.2 Model Prediction Output

Example response after hitting the /predict-live endpoint:

---

13. License

This project is licensed under the MIT License.
You are free to use and modify.