How deepakns021/housing-price-analysis works
This repository analyzes North Andover home sales to estimate the relationship between housing features, incinerator proximity, and price, then exposes predictions through a Streamlit app. The documented workflow runs from source data through exploratory analysis and OLS modeling to an interactive prediction or near/far comparison. App and README evidence supports the runtime model-loading and prediction links; notebook internals were not sampled, so analysis details are limited to documented capabilities.
Main entry points: app.py Streamlit Interface · model.pkl Trained Model · incinerator.xlsx Home Sales · notebook.ipynb Exploratory Analysis
Components 11
Data Analysis
- Home Sales
data/incinerator.xlsx - Exploratory Analysis
notebook.ipynb - Proximity Analysis
notebook.ipynb
Price Modeling
- OLS Regression
notebook.ipynb - Trained Model
model.pkl
Prediction App
- Streamlit Interface
app.py - Feature Encoding
app.py - Price Prediction
app.py - Near-Far Comparison
app.py - Business Insights
app.py
Other components
- Stakeholder
Connections 19
- Home Sales → Exploratory Analysis: analyzed in
- Exploratory Analysis → Proximity Analysis: examines
- Home Sales → OLS Regression: modeled from
- OLS Regression → Trained Model: produces
- Stakeholder → Streamlit Interface: provides inputs
- Streamlit Interface → Trained Model: loads
- Streamlit Interface → Feature Encoding: encodes inputs
- Streamlit Interface → Price Prediction: requests prediction
- Feature Encoding → Price Prediction: supplies features
- Trained Model → Price Prediction: predicts
- Streamlit Interface → Near-Far Comparison: requests scenarios
- Feature Encoding → Near-Far Comparison: supplies scenarios
- Trained Model → Near-Far Comparison: predicts scenarios
- Price Prediction → Streamlit Interface: returns estimate
- Near-Far Comparison → Streamlit Interface: returns price impact
- Streamlit Interface → Business Insights: displays
- Business Insights → Stakeholder: shows findings
- Price Prediction → Stakeholder: shows estimate
- Near-Far Comparison → Stakeholder: shows comparison