Property Price Prediction API Documentation

---

Table of Contents

1. Overview
2. Endpoints
   • 1. Predict Property Price
3. How to Run
4. Error Handling

---

Overview

This API is designed to predict property prices based on geographical coordinates. It utilizes a decision tree model to estimate property prices per square meter at a given latitude and longitude.

---

Endpoints

1. Predict Property Price

• URL: /predict_get
• Method: GET
• Query Params:
  • latitude=[float] (required)
  • longitude=[float] (required)
• Success Response: JSON object with the predicted price per square meter.
• Error Response:
  • 400 Bad Request if query parameters are missing or invalid
  • 503 Service Unavailable if the model is not loaded
  • 500 Internal Server Error for other issues
• Example: http://localhost:5000/predict_get?latitude=48.8566&longitude=2.3522

---

How to Run

• Ensure FastAPI, Uvicorn, and necessary dependencies are installed.
• Load the decision tree model named decision_tree_model.pkl.
• Run the server using uvicorn main:app --host=localhost --port=5000, with main.py being the script containing the FastAPI app.
• Access the API at http://localhost:5000.

---

Error Handling

• Custom error handling is implemented to manage internal server errors.
• The server returns appropriate error messages and HTTP status codes based on the nature of the error encountered.
• The API checks for the model's availability before making predictions, ensuring reliable responses.

---