trellixvulnteam / compas-racial-bias_71to

Microsoft Azure Hackathon: COMPAS Racial Bias and Score Prediction (4th place; Honorable Mentions on Microsoft Blog)

About COMPAS Racial Bias

Azure Hackathon: COMPAS Racial Bias and Predict Scores

Packages and Platforms

1. Python 3.5
2. Streamlit for providing the interactive web app template.
3. Google Colab Notebooks for building the models and predicting the scores.
4. Flask and Microsoft Azure to deploy the model and website.

Installation

Create virtual environment

If running on Python 3, the venv is built-in: python3 -m venv env and then activate the environment: source env/bin/activate Install dependencies (details shown below). pip3 install -r requirements.txt

Run streamlit app

streamlit run app.py

Dependencies

Add or remove dependencies

We use "pipreqs" to generate requirements.txt which contains the required requirements for the deployment.

Add/remove dependencies at any point by updating requirements.txt (Python deps) or packages.txt (Debian deps) and doing a git push to your remote repo. This will cause Streamlit to detect there was a change in its dependencies, which will automatically trigger its installation.

It is best practice to pin Streamlit version in requirements.txt. Otherwise, the version may be auto-upgraded at any point without your knowledge, which could lead to undesired results (e.g. when we deprecate a feature in Streamlit).