California Housing Price Prediction API

This project is a FastAPI-based application that predicts housing prices in California based on various features. The application is containerized using Docker and includes monitoring with Prometheus and Grafana.

Installation

Prerequisites

• Docker
• Docker Compose

Clone the Repository

git clone https://github.com/yourusername/california-housing.git
cd california-housing

Build and Run the Docker Containers

1. Build the Docker images: docker-compose build
2. Run the docker containers: docker-compose up -d
3. The FastAPI application will be accessible at http://localhost:8000
4. Prometheus will be accessible at http://localhost:9090
5. Grafana will be accessible at http://localhost:3000

Usage

1. Start the FastApi Server python src/app.py

2. Open your browser and go to http://127.0.0.1:8000/docs to access the swagger API documentation.

3. Make a post request to /predict endpoint to get the prediction with the required fields

Project Structure

• src/: Source code for the application
  • app.py: Main application file
  • model.py: Model loading and saving functions
  • data_download.py: Script to download and load data
• models/: Directory to store the trained models
• data/: Directory to store data files
• requirements.txt: List of dependencies
• Dockerfile: Dockerfile to build the Docker image
• docker-compose.yml: Docker Compose file to run the application and monitoring services
• .gitignore: Git ignore file to exclude unnecessary files

Monitoring

The application includes monitoring with Promethius and Grafana

• Promethius is configfured to scrape metrics from the FastApi Application
• Grafana is configured to visualize the metrics captured by Prometheus

Deploying to AWS

To deploy to the Docker container to AWS, you can use AWS Elastic Beanstalk, ECS (Elastic Container Service) or EKS (Elastic Kubernetes Service).

1. Install AWS Elastic Beanstalk CLI pip install awsebcli

2. Initialize the EB application eb init -p docker your-app-name

3. Create elastic beanstalk environment to deploy the application eb.create your-env-name

4. Open the app in web browser eb open

Replace yourusername, your-app-name, and your-env-name with appropriate names for your application and environment. This will deploy your Docker container to AWS Elastic Beanstalk and make your application accessible via a public URL.

License This project is licensed under the MIT License.