How mainakadak02/cardioguard-api works
Heart Guard is a Flask backend that accepts heart-health features and coordinates a risk prediction, returning a risk level and probability. For high-risk results, it ranks nearby hospitals using the submitted coordinates and a CSV-loaded hospital list. The repository also includes a script that trains and saves the classifier. The README gives no further product details; this map is based on the sampled source, and does not assume an unshown client or training-data file.
Main entry points: app.py Prediction route · heart_model.pkl Trained classifier · train_model.py Model training · hospitals.csv Hospital records
Components 9
Prediction API
- Prediction route
app.py - Feature framing
app.py
Risk Assessment
- Trained classifier
heart_model.pkl - Risk threshold
app.py
Hospital Lookup
- Hospital loader
app.py - Hospital records
hospitals.csv - Distance ranking
app.py
Model Training
- Model training
train_model.py
Other components
- API caller
Connections 9
- API caller → Prediction route: submits features
app.py - Prediction route → Feature framing: frames features
app.py - Prediction route → Trained classifier: requests probabilities
app.py - Prediction route → Risk threshold: applies threshold
app.py - Risk threshold → Prediction route: returns risk result
app.py - Hospital loader → Hospital records: reads records
app.py - Prediction route → Distance ranking: requests nearby results
app.py - Distance ranking → Prediction route: returns top five
app.py - Model training → Trained classifier: saves model
train_model.py