This repository provides an implementation of privacy-aware Gaussian processes built on PyTorch.

• Python >= 3.10
• Dependencies are managed through pyproject.toml

To reproduce the experiments, run the following commands step by step in your terminal:

1. Clone the repository or download the folder, ensure you are in the privacygp directory

2. Create and activate a virtual environment:

python -m venv .env
source .env/bin/activate

3. Upgrade pip and setuptools to ensure compatibility:

pip install --upgrade pip setuptools wheel

4. Install the package and all dependencies:

pip install .

5. Run the experiments

• To reproduce the example 1 in Figure 1, run the following command

  python experiments/example.py

• To reproduce the satellite simulation in Figure 2, 3, 4 and Table 2, run the following command

  python experiments/satellite_simulation.py

• To reproduce the satellite simulation with zero-mean GP in Figure 5, run the following command

  python experiments/zeromean_satellite_simulation.py

• To reproduce the real-world application on the PUMS Data provided by the U.S. Census Bureau in Figure 6, 7 and Table 3, run the following command

  python experiments/census.py