CO2 Leaderboard

Overview

CO2 Leaderboard is a full-stack application that tracks and visualizes carbon emissions across different states. It provides insights into historical trends, future predictions, and regional comparisons to monitor progress toward environmental goals.

Tech Stack

• Backend: Flask, Pandas
• Frontend: React, TypeScript, Material-UI
• Machine Learning: PyTorch for neural network-based predictions

Installation & Setup

Backend Setup

1. Install dependencies:

   pip install flask flask-cors pandas torch torchvision torchmetrics

2. Run the Flask API:

   python app.py

Frontend Setup

1. Navigate to the frontend directory (if applicable) and install dependencies:

   npm install

2. Start the frontend:

   npm run dev

File Descriptions

Backend Files:

• app.py - Flask API to serve emissions data.
• co2pred.py - PyTorch model for CO2 emission predictions.
• test_viz.py - Visualization script to test and compare model results.

Frontend Files:

• App.tsx - Main React application file.
• main.tsx - Renders the React app.
• Dashboard.tsx - Displays US map of emissions.
• USMap.tsx - Interactive map for emissions data.
• ChartPage.tsx - Displays emissions trends for selected states.
• EmissionChart.tsx - Chart component for emissions visualization.

Data & Predictions:

• 5 Year Predictions.csv - Predicted CO2 emissions for future years.
• table1_train.xlsx - Training dataset for ML model.
• table1_test.xlsx - Test dataset for ML model.
• table1_copy.xlsx - Additional dataset for predictions.

How to Use

1. View emissions data by running the Flask API (app.py).
2. Interact with the dashboard by starting the React frontend.
3. Analyze predictions using co2pred.py to generate forecasts.
4. Explore visualizations via the ChartPage.tsx and EmissionChart.tsx components.

License

This project is licensed under the MIT License. See LICENSE for more details.