README.md

metadata

title: Weather Simulation & Traffic Risk
emoji: 🌦️
colorFrom: blue
colorTo: green
sdk: docker
pinned: false

🌦️ Weather-Traffic Risk Prediction Pipeline

This project implements a full E2E pipeline that fetches weather data, cleans it, performs feature engineering, and uses a Machine Learning model to predict traffic risk.

🚀 Components

1. Weather Pipeline (weather_pipeline.py):

   • Fetches real-time data from WeatherAPI.com.
   • Handles missing values and normalizes conditions (Rainy, Cloudy, etc.).
   • Calculates a "Risk Score" based on heuristics.
   • Implements caching (10 min) to optimize API usage.

2. ML Model (model.py):

   • Uses a Random Forest Classifier trained on synthetic data.
   • Predicts Low, Medium, or High traffic risk based on temperature, rainfall, and weather condition.
   • Includes a training script to generate baseline patterns.

3. FastAPI Server (app.py):

   • Exposes a REST API endpoint: /weather/{city}.
   • Returns structured JSON with weather details and ML risk prediction.
   • Visual alerts (Green/Yellow/Red) included in responses.

🛠️ Setup

1. Install dependencies:

   pip install -r requirements.txt
   

2. Configure API Key:

   • Open .env and replace your_api_key_here with your OpenWeather API key.

3. Run the Server:

   python app.py
   

4. Test the endpoint:

   • Open http://localhost:8000/weather/Dehradun in your browser.

📊 Output Format

{
  "city": "Dehradun",
  "weather": {
    "temperature": 28,
    "rainfall_1h": 12,
    "condition": "Rainy",
    "base_risk_score": 3
  },
  "ml_prediction": {
    "traffic_risk_level": "High",
    "confidence": 0.89,
    "status_color": "🔴 Red"
  }
}

🤗 Deploying to Hugging Face Spaces

This project is configured for easy deployment on Hugging Face using Docker.

1. Create a New Space:

   • Go to huggingface.co/new-space.
   • Select Docker as the SDK.
   • Choose the Blank template or upload your files directly.

2. Add Secret (API Key):

   • In your Space settings, go to Variables and secrets.
   • Add a new secret named WEATHER_API_KEY and paste your key.

3. Upload Files:

   • Upload app.py, weather_pipeline.py, model.py, requirements.txt, Dockerfile, and the .pkl files.
   • Hugging Face will automatically build and deploy your container on port 7860.

4. Test:

   • Access your API at https://huggingface.co/spaces/[YOUR_USERNAME]/[YOUR_SPACE_NAME]/weather/Dehradun.