Oddity Demo

Web app showcasing a demo of time series anomaly detection with the Oddity Engine

The web app was built with Streamlit and deployed to a google cloud kubernetes cluster. The cluster may or may not be currently still available, so follow the instructions below to run the demo locally with a docker container.

Instructions

Running Locally

• Ensure Docker is installed on your computer

• Clone or download this github repository and cd into it

• Build the docker container with:

   docker build -t oddity-demo/app:latest .

• Then run the container lockally with:

   docker run -p 8501:8501 oddity-demo/app:latest

The demo should now run locally on port 8501

Important Links

The following are some important links for more information:

PyPi: https://pypi.org/project/oddity/

Oddity: https://github.com/Lleyton-Ariton/oddity

Oddity Engine (Rust): https://github.com/Lleyton-Ariton/oddity-engine

Extra

For some extra information on time series data/anomaly detection, you can check out my medum article series Houston, we have a problem.