Skip to content

GET_STARTED.md

Latest commit

 

History

History
173 lines (133 loc) · 5.28 KB

GET_STARTED.md

File metadata and controls

173 lines (133 loc) · 5.28 KB
Raw

🚀 Getting Started with SL-GPS Frontend GUI

Welcome! This guide will get you up and running with the SL-GPS graphical interface in 5 minutes.

Installation (2 minutes)

Step 1: Clone the Repository

git clone https://github.com/ctftamu/SL-GPS.git
cd SL-GPS

Step 2: Install Dependencies

# Install all requirements (including GUI)
pip install -r requirements.txt
pip install -r frontend/requirements.txt

That's it! You're ready to go.

Launch the GUI (1 minute)

python -m frontend

Your browser will automatically open to http://localhost:7860

You should see a beautiful interface with three tabs:

  • 📊 Generate Dataset
  • 🧠 Train Neural Network
  • 📖 Documentation

First Run: Quick Test (2 minutes)

Follow these steps to test everything:

1. Generate Dataset Tab

  • Click Upload Cantera Mechanism

    • For testing, use any Cantera YAML mechanism (e.g., gri30.yaml built-in)
    • Or download from Cantera

  • Keep default parameters OR for quick test change:

    • Number of Cases: 5 (instead of 100) - runs in ~2-5 minutes
    • Temperature: 1000-1500 K (instead of 800-2300)

  • Click 🚀 Generate Dataset

  • Wait for completion (you'll see green status messages)

2. Train Neural Network Tab

  • Species should auto-populate
  • Keep default architecture
  • Click 🚀 Train Neural Network
  • Wait for training completion

3. Check Results

  • Look for generated_data/ folder with:
    • model.h5 - Your trained neural network
    • scaler.pkl - Input normalizer
    • data.csv - Training data
    • species.csv - Species importance masks

Next Steps

After the quick test:

  1. Read the Documentation

  2. Run a Real Experiment

    • Use your own mechanism file
    • Set realistic parameters for your fuel/conditions
    • Generate full dataset (n_cases = 100-500)
    • Train with custom NN architecture if needed

  3. Use the Results

    • See SL_GPS.py to use model in adaptive simulations
    • See display_sim_data.py to visualize results

File You'll Need

Most important files:

File Purpose
frontend/app.py The GUI application
FRONTEND_QUICKSTART.md Quick reference (parameters, troubleshooting)
docs/frontend.md Complete frontend documentation
docs/setup.md Installation & troubleshooting
docs/api.md Full API reference

Common Questions

Q: What mechanism file should I use?
A: Any Cantera YAML file. Examples:

Q: How long does data generation take?
A: Depends on n_cases:

  • 5 cases: 2-5 minutes
  • 20 cases: 5-15 minutes
  • 100 cases: 30-60 minutes
  • 500 cases: 2-4 hours

Q: Can I customize the neural network?
A: Yes! See docs/code_structure.md and edit src/slgps/mech_train.py::spec_train() for custom layers.

Q: Port 7860 is in use
A: Run python -m frontend --server_port 7861 for a different port.

Q: How do I use my trained model?
A: Load model.h5 and scaler.pkl in src/slgps/SL_GPS.py for adaptive simulations.

Where to Go From Here

File Structure Overview

SL-GPS/
├── frontend/              ← GUI code (python -m frontend)
│   ├── app.py            ← Main application
│   ├── requirements.txt   ← GUI dependencies
│   └── README.md          ← Detailed frontend docs
├── docs/
│   ├── frontend.md        ← GUI guide
│   ├── setup.md           ← Installation
│   ├── api.md             ← API reference
│   └── ...
├── FRONTEND_QUICKSTART.md ← Quick reference
├── IMPLEMENTATION_SUMMARY.md ← What was built
└── src/slgps/
    ├── main.py           ← Data generation (called by GUI)
    ├── mech_train.py     ← NN training (called by GUI)
    └── SL_GPS.py         ← Run simulations with trained model

Success Checklist

  • ✅ Installed Python 3.8+
  • ✅ Cloned SL-GPS repository
  • ✅ Installed dependencies (pip install -r requirements.txt)
  • ✅ Installed frontend (pip install -r frontend/requirements.txt)
  • ✅ Launched GUI (python -m frontend)
  • ✅ Browser opened to http://localhost:7860
  • ✅ Uploaded a mechanism file
  • ✅ Generated a dataset
  • ✅ Trained a neural network
  • ✅ Found results in generated_data/

All done! You're now ready to use SL-GPS for chemistry reduction. 🎉

Need Help?

Happy chemistry reduction! 🧪