This codebase supports the work "Variational Quantum Optimization of Nonlocality in Noisy Quantum Networks."

Preprint: https://arxiv.org/abs/2205.02891

The goal of this codebase to make our numerical methods accessible, transparent, and reproducible. All scripts are designed to be run on a laptop computer within a few hours. A few scripts run quantum circuits remotely on IBM quantum computers, which may require special access to IBM devices.

This supplementary codebase is largely built upon two Python packages:

• qNetVO: the Quantum Network Variational Optimizer
• PennyLane: a cross-platform library for differentiable programming of quantum computers.

This supplementary codebase is primarily written in Python. The two exceptions are found in the ./script/proof_numerics/ directory where a Julia script and Mathematica notebook are found. These scripts are not used for data collection, but to verify the correctness of results presented in the work.

• ./script : This directory contains all scripts for data collection and plot creation. All data is saved to the ./data directory.

• ./data : This directory contains all data collected using the scripts. Data is organized first by the network topology, then by the script that collected the data. Within each folder, script data includes JSON and PNG files. The JSON files contain raw data collected during optimizations. The PNG files contain a rough plot of the collected optimization data. Each file is named by the ansatz for the optimization and a datetime identifier.

• ./src : This directory contains helper methods for collecting, writing, reading, analyzing, and plotting data.

See CITATION.bib for bibtex.

For convenience and reproducibility code should be run using the conda development environment. The Anaconda distribution of Python ensures a consistent development environment. Follow the Anaconda installation instructions to set up the conda command line tool for your operating system. The conda tool creates the dev environment from the environment.yml file. For more details on how to use conda see the managing environments page in the conda documentation.

To create the dev environment, navigate to the root directory of the vqo-nonlocality-noisy-quantum-networks repository and follow these steps.

1. Create the vqo-nonlocality-dev conda environment:

(base) $ conda env create -f environment.yml

2. Activate the vqo-nonlocality-dev conda environment:

(base) $ conda activate vqo-nonlocality-dev

All code in this project is autoformatted using black. After setting up the development environment, run:

(vqo-nonlocality-dev) $ black -l 100 src script