Live Webservice

https://floodapi.onrender.com/docs

Flood Risk Prediction API

This API is built using FastAPI to predict flood risks based on various environmental and infrastructural factors. The model used for prediction is loaded from a pickle file and uses features like Monsoon Intensity, Urbanization, Climate Change, and more.

Features

• Load a pre-trained model from a specified path.
• Predict flood risk based on input data with 20 different features.
• Structured logging for better tracking and debugging.
• CORS support for local development and external connections.

Getting Started

Prerequisites

Ensure you have the following installed:

• Python 3.7+
• FastAPI
• Uvicorn
• Pydantic
• Pandas
• Pickle
• Python-dotenv

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/flood-risk-prediction-api.git
   cd flood-risk-prediction-api

2. Install dependencies:

   pip install -r requirements.txt

3. Create a .env file in the root directory and add the following environment variables:

   MODEL_PATH=path/to/your/FloodRisk.pkl
   HOST=127.0.0.1
   PORT=8000

4. Place your trained model (e.g., FloodRisk.pkl) in the specified path or update the MODEL_PATH in the .env file.

Running the API

Run the API using Uvicorn:

uvicorn main:app --reload