Diabetes Prediction Demo

React + TypeScript + Vite app for a browser-based diabetes prediction demo powered by a BigML decision tree.

Stack

• React 19
• TypeScript
• Vite
• Tailwind CSS
• shadcn/ui-style reusable components

Run locally

npm install
npm run dev

Build

npm run build

BigML model source

The app reads the model directly from Decision tree - code from BigML.txt and evaluates the exported nested if/else tree in the browser.

The public helper remains predictDiabetes(data), so the UI stays unchanged even if you swap in a new export later.

Project structure

• src/App.tsx: page layout, form state, submit handling
• src/components/: UI primitives and reusable form field
• src/data/diabetes-fields.ts: field metadata and biological ranges
• src/lib/validation.ts: parsing and input validation
• src/lib/predict-diabetes.ts: parser and evaluator for the BigML decision tree export

Deployment

GitHub Pages

This repository now includes .github/workflows/deploy.yml for automatic GitHub Pages deployment.

1. Open the repository on GitHub.
2. Go to Settings -> Pages.
3. Under Build and deployment, set Source to GitHub Actions.
4. Push to main and GitHub will build and publish the dist folder automatically.

Your Pages URL should be:

https://tom-kroupa.github.io/diabetesdatamining/

Vercel or Netlify

If you prefer a simpler UI and automatic preview URLs on every push:

1. Import the GitHub repository into Vercel or Netlify.
2. Keep the default Vite build settings (npm run build, output dist).
3. Each new push to GitHub will trigger a deployment automatically.

Note

The current repository includes the pasted BigML export file and evaluates that logic client-side.