Antares

Explainable AI through decision trees.

What is it?

Antares allows maintainters of AI systems to offer end-users explanations for classification and regression results.

Setup

API Setup with Pipenv

Install pipenv.
Run pipenv install from the project's root.
Activate the virtualenv before developing using pipenv shell

Secret Key setup

Copy the config.ini.example file and create a config.ini file. These secrets are only valid for the dev environment.

Setup Redis

Ensure you have a version of Redis installed on your computer. You can find a guide on how to install here. If you are on OSX we recommend using homebrew, a guide can be found here.

Once installed you can start your Redis server using this command redis-server /usr/local/etc/redis.conf and can ensure it is running via redis-cli ping. If you get the response PONG the server is running!

Finally, ensure the redis python client is pointing to the right server. If you haven't configured anything in a custom manner for the installation, the default configuration should be good. Otherwise you will have to change the port, and password field.

Setup Web Server

Run yarn from the webapp directory and issue yarn start to start the development server at https://localhost:3000

Tests

The API uses pytest to run tests. From the project's root, run:

    pytest

API Documentation

The API documentation can be found here

Linting

ESLint

To run ESLint use yarn lint.

coshx/antares