How nourah-alotaibi/gp works
Project Type and Purpose
• This is a full-stack web application for recommending epilepsy treatment plans using a machine-learning model.
• The React-based frontend provides an interface for doctors to upload patient data (Excel/CSV) and view predictions.
• The Flask-based backend exposes REST endpoints, handles file uploads, runs preprocessing and the ML inference pipeline, and returns treatment recommendations.
File Structure Analysis
• Top-level:
• backend/: Contains Flask API (app.py, minimal_server.py, simple_server.py), ML utilities (severity_engine.py, create_sample_data.py), tests, environment scripts, and an uploads/ folder for incoming files.
• frontend/: Contains React app, package.json, node_modules/.
• data/: Holds training and sample datasets (Excel, CSV).
• venv/ and backend/venv: Python virtual environments.
• Key patterns:
• REST API layer (Flask app).
• ML “engine” module (severity_engine.py + data extractors).
• File-based I/O for uploads and sample data.
• Frontend-backend separation via HTTP (CORS configured).
Architecturally Relevant README Insights
• Backend runs on http://localhost:5000, React on http://localhost:3000.
• Emphasis on file‐format requirements, optional sample generation, and troubleshooting.
• No external databases—data is loaded from data/ and persisted temporarily to backend/uploads/.