Maurizio Viviani — Reproducible computational research notes
October 11, 2026

Articles
https://quantum.robotics.it/research/qaoa-shot-budget.html
https://quantum.robotics.it/research/calibration-before-quantum-claims.html

Download reproduce_research.py and run:
python -m pip install numpy==2.3.5 matplotlib==3.10.8
python reproduce_research.py --output-dir results

The script generates research-results.json, two CSV tables, and SVG/PNG figures.
It uses an ideal four-qubit QAOA statevector and synthetic binary classification data.
Seeds, graph edges, angle-grid settings, sample sizes, temperature-grid settings and metric definitions are included in the source and JSON record.
The published runtime versions are in research-results.json. Small last-digit differences can occur across numerical environments.
The sampling study fixes the angles before drawing measurement counts. The calibration study fits its temperature only on the validation sample.
All data are synthetic; no quantum hardware, real particle-detector data or private records are used.