Skip to content

Files

Latest commit

 

History

History

Project

README.md

Pattern recognition WEB and API

This is a full-stack project for the subject of pattern recognition. The project is split into 3 parts:

  • backend - API and DB for feature set creation/deletion and execution of exploratory data analysis using PCA and hierarchical clustering
  • frontend - Interface for hosting a SPA and a wrapper for various web requests done from the frontend
  • nginx - Reverse proxy and load balancer for the backend and frontend and serving static files (CSS/JS)

For each part, there is a separate README file with more detailed information in the respective folder.

Running the code

You can run the project by running the following command in the root of the project:

docker compose build
docker compose up -d

To stop the project, run the following command in the root of the project:

docker compose down

The dashboard can be accessed on the following URL: http://localhost:5000/recognition-web/v1/dashboard

Basic project architecture

The project uses a microservices architecture with docker compose to run the services. It creates 4 containers that are interconnected with 2 networks. The containers are:

  • db - Postgres database for storing the feature sets (Network: db-api)
  • recognition-api - API for creating/deleting feature sets and running exploratory data analysis (Network: db-api, api-nginx)
  • recognition-web - SPA for the dashboard and a wrapper for the API requests (Network: web-nginx)
  • recognition-nginx - Reverse proxy and load balancer for the backend and frontend and serving static files (CSS/JS) (Network: web-nginx, api-nginx)

Dependencies used

Bellow is a list of dependencies used in the project by each part of the project:

  • Backend (API and DB)

    • API (Docker image: python:3.11-alpine)
      • Python (Version: >=3.9, <3.12)
      • asyncio (Version: ^3.4.3)
      • aiohttp (Version: ^3.8.1)
      • asyncpg (Version: ^0.26.0)
      • aiohttp-sqlalchemy (Version: ^0.34.0)
      • SQLAlchemy (Version: ^1.4.39)
      • numpy (Version: ^1.17.3)
      • scikit-learn (Version: ^1.2.0)
      • pytest (Version: ^6.2.5)
      • pytest-asyncio (Version: ^0.16.0)
      • testcontainers (Version: ^3.0.0)
    • DB (Docker image: postgres:15.1)
      • Postgres (Version: 15.1)

  • Frontend (Docker image: python:3.11-alpine)

    • Python (Version: >=3.9, <3.12)
    • asyncio (Version: ^3.4.3)
    • aiohttp (Version: ^3.8.1)

  • Nginx

    • Nginx (Version: 1.23)

References

[1] Markelle Kelly, Rachel Longjohn, Kolby Nottingham, The UCI Machine Learning Repository, https://archive.ics.uci.edu