Running and Checking Experiments

Use these guides for environment setup, numerical evidence, figure reproduction and shared benchmark suites. A guide supports experiments and is not counted as an experiment.

Choose a workflow

• Reproducing a Figure — Reconstruct a computational figure from its model, numerical outputs, plotting instructions, and checks without confusing visual agreement with scientific validation.
• Running an Experiment — Download a complete computational experiment, create its declared environment, run the checks, and interpret the resulting evidence.
• Reproducibility Checklist — A release audit for computational claims, environments, parameters, validation tests, retained data, figures, provenance, and licenses.
• Reproducibility Benchmarks — A versioned acceptance suite for hydrogen, helium, H₂⁺, Rabi, optical Bloch, Jaynes–Cummings, and Doppler-cooling computational artifacts.

Interpret the evidence

The Labs catalog distinguishes downloadable pilot packages from other existing script-backed investigations. Each investigation states its own assumptions and retained checks. Follow Running an Experiment for the shared execution workflow and the Reproducibility Checklist for evidence and error budgets.

References

• National Academies of Sciences, Engineering, and Medicine, Reproducibility and Replicability in Science, National Academies Press, 2019, doi:10.17226/25303 — interpreting reproducibility and independent scientific evidence.