QuantAdvantage-MaxCut

This repository contains code, scripts, and data used to generate, solve, and analyze Max-Cut instances for benchmarking classical and quantum algorithms. It accompanies a research paper (link to be added here). The project includes instance generation, exact and approximate solvers, Goemans–Williamson (GW) SDP utilities, QAOA experiments, and analysis/plotting tools.

Contents overview

• Aggregate_plot/ — scripts for processing and aggregating QAOA experiment outputs across the entire dataset and plotting results with confidence intervals.
• Generate_Instances/ — tools to generate random graphs, compute exact Max-Cut values, compute Goemans–Williamson expectations, and filter instances whose GW expectation meets a specified approximation threshold.
• Goemans-Williamson/ — GW SDP scripts and plotting helpers.
• Selected_Instances/ — a curated set of example .gml graph instances and the Selected_Instances_Info/ folder with per-instance precomputed parameters.
• QAlgorithm/ — QAOA script.
• Run_QAOA/ — runner script for launching single-instance or batched QAOA experiments.
• Plots/ — directory containing plots of QAOA and GW performance on Max-Cut instances.
• Process_stats/ — scripts to aggregate experiment outputs, compute percentiles, and plot results with confidence intervals.
• Results/ — recommended directory for saving raw experiment outputs.

Paper link and citation

TODO: Add link to the research paper and citation instructions here.