Getting Started

Please reference our project repository to view the codebase for our 2024-2025 UC San Diego Data Science Capstone Project: An Adversarial Framework for Mitigating Gender Bias in Coronary Heart Disease Prediction.

1. Cloning Code

• Clone repository into your working directory: git clone https://github.com/patsals/CHD-adversarial-nn.git
• Enter the repository directory CHD-adversarial-nn

2.a. Setting up Environment (Windows Users)

• Install Linux using Windows Subsystem for Linux (WSL) in powershell: wsl --install
• Install Linux Distribution System through Microsoft store: Ubuntu 22.04.5 LTS
• Open Ubuntu and wait for download to complete
• Once download complete enter a Username and Password as prompted
• Open VS Code then click on the two caret mark icon on the bottom left of window:
• Select the “Connect to WSL Using Distro…”
• You should see Ubuntu 22.04.5 LTS and select that option
• Proceed to 2.b instructions

2.b. Setting up Environment (Linux/Mac OS users)

• Note: TensorFlow may not be available for the latest version of Python. As of the development of this project, we are using Python 3.11.
• Create a working environment: python3 -m venv venv
• Activate the working environment: source venv/bin/activate
• Download the dependencies: pip install -r requirements.txt

3. Downloading & Processing Data

• Run the data setup script to extract, transform, and load all of the data locally:
  • python data_setup.py for windows users
  • python3 data_setup.py for Linux/Mac OS users
  • adding the --skip_download flag skips the requesting/downloading of files
• Run the data processing script to handle null/missing values:
  • python data_process.py for windows users
  • python3 data_process.py for Linux/Mac OS users
• Note: The setup script makes requests to multiple url endpoints at www.sciencedirect.com to download the files. It is important to lookout for any message logs that do not show [SUCCESS] as an output — this indicates an error that needs to be resolved.