How ralijaonafehizoro961-wq/immo-price-predictor works
Immo Price Predictor is a React site backed by a FastAPI house-price estimation API. The main user workflow is to enter house characteristics, load supported ZIP codes, submit the form, and view an estimated USD price. The API validates the input, derives house age, and uses a saved scikit-learn pipeline to predict log(price) before converting it back to dollars. The README also describes a notebook-based exploration, cleaning, feature-selection, training, and evaluation workflow. Notebook internals were not sampled, so their relationships are not inferred.
Main entry points: App.jsx Estimate form · main.py API and validation · main.jsx React entry · 02_cleaning.ipynb Data cleaning
Components 10
Web experience
- React entry
frontend/src/main.jsx - Estimate form
frontend/src/App.jsx
Prediction API
- API and validation
backend/app/main.py - Saved model
backend/models/house_model.joblib
Model development
- Data exploration
notebooks/01_exploration.ipynb - Data cleaning
notebooks/02_cleaning.ipynb - Feature selection
notebooks/03_features.ipynb - Model training
notebooks/04_model.ipynb - Model evaluation
notebooks/05_test.ipynb
Other components
- Home seeker
Connections 9
- Home seeker → React entry: opens site
README.md - React entry → Estimate form: renders
frontend/src/main.jsx - Estimate form → API and validation: requests ZIP codes
frontend/src/App.jsx - Estimate form → API and validation: submits estimate
frontend/src/App.jsx - API and validation → Saved model: loads pipeline
backend/app/main.py - Data exploration → Data cleaning
- Data cleaning → Feature selection
- Feature selection → Model training
- Model training → Model evaluation