🏢 Real Estate Investment Dashboard (India Housing Data)

This project is a Streamlit-based interactive dashboard built to analyze real estate property data across India.
It helps users explore price trends, location insights, amenities impact, property features, and investment-related factors through clean visualizations.

1. Overview
   This project builds a complete data analytics dashboard to explore:

• Property prices
• Size distribution
• Locality/City/State comparisons
• Amenities, owner type, BHK, facing, furnishing
• Multiple filters: State, City, Furnishing, Facing, Amenities, Property type
• Interactive Plotly charts (bar, pie, heatmaps, comparisons)

The cleaned dataset and engineered features are created using eda_cleaning_pipeline.py, which includes:

• Missing/duplicate handling
• Numeric coercion
• Feature engineering (Price per SqFt, Age, Density Scores, Amenities Count, Furnishing Flag, Facing Simplification, Ready-to-Move Flag)
• Output: cleaned_india_housing.csv + preprocessor.joblib

---

2. Data Cleaning Pipeline (eda_cleaning_pipeline.py)

This script:

• Loads india_housing_prices.csv
• Renames and standardizes columns
• Removes duplicates
• Converts numeric text → numbers
• Handles missing values (median imputation for numeric, "Unknown" for categorical)
• Drops numeric columns that are fully empty
• Recomputes missing Price_per_SqFt
• Creates several engineered features:
  • Price_per_SqFt_final
  • Age_of_Property
  • School_Density_Score
  • Hospital_Density_Score
  • Amenities_count
  • Is_Fully_Furnished
  • Is_Ready_to_Move
  • Facing_simple
• Saves:
  • cleaned_india_housing.csv
  • preprocessor.joblib

---

3. Dashboard (app.py)

The Streamlit dashboard provides:

Key Features

• Full-width dashboard layout
• 5 KPIs
• 4 analysis pages:
  1. Price & Size Analysis
  2. Location Analysis
  3. Relationship & Correlation
  4. Ownership / Investment / Amenities

Tech Highlights

• Multi-dropdown filters with “Select All” logic
• Plotly visualizations
• Clean UI with HTML/CSS adjustments
• Custom re-usable function to render charts in two-column layout

Run the dashboard using:

streamlit run app.py

---

📁 Project Structure

📦 Real_Estate │ ├── app.py # Streamlit dashboard ├── eda_cleaning_pipeline.py # Data cleaning & preprocessing script ├── train_models.py # (Optional) ML training script │ ├── india_housing_prices.csv # Raw dataset (not included) ├── cleaned_india_housing.csv # Cleaned dataset (generated) ├── preprocessor.joblib # Saved preprocessing pipeline ├── investment_clf.joblib # (Optional) Trained classifier ├── price_reg.joblib # (Optional) Trained regressor ├── feature_names.joblib # OHE-expanded feature names │ ├── README.md └── requirements.txt

🛠️ Installation & Setup

1️⃣ Clone the repository git clone https://github.com/your-username/real-estate-dashboard.git cd real-estate-dashboard

2️⃣ Install dependencies pip install -r requirements.txt

3️⃣ Run data cleaning pipeline python eda_cleaning_pipeline.py

This generates: cleaned_india_housing.csv preprocessor.joblib

4️⃣ Launch Streamlit dashboard streamlit run app.py

📊 Data Used The dataset contains cleaned housing information including:

Price (Lakhs) Price per SqFt Size in SqFt Amenities count Age of property Furnished status BHK Facing direction Ownership & availability details Location: State, City, Locality Nearby hospitals & schools Public transport accessibility

🧹 Cleaning & Feature Engineering All cleaning happens in eda_cleaning_pipeline.py, which:

Removes duplicates Converts numeric columns Handles missing values Creates new columns: Price_per_SqFt_final School_Density_Score Hospital_Density_Score Amenities_count Is_Fully_Furnished Facing_simple Is_Ready_to_Move Saves cleaned dataset

🎯 Purpose of This Project This dashboard was created to:

Visualize property trends Understand location-based pricing Compare amenities impact on price Explore investment-related metrics No machine learning models are included.

🧑‍💻 Tech Stack Python 3.10+ Streamlit Pandas Plotly Express NumPy Scikit-learn Seaborn & Matplotlib (used inside cleaning script)