Reproducibility Resources and Templates

01Download and run

ASD Experimental Reproducibility Kit v1.0

After extraction, use Python 3.10 or later:

python run.py --output results
python verify.py

Calculations use only the standard library. The notebook calls the same core functions and requires your own Jupyter environment. The numerical script reads bundled synthetic inputs and writes JSON and CSV. Expected figures are bundled; regenerating them uses optional matplotlib 3.10.8 and figure.py. ASD outputs and predicted differences are explicitly null. Nothing automatically requests paper data or contacts authors.

02Reproducible workbench content

The clock model computes a user-selected background-axis projection. The muon model first computes standard single-path survival, then sums SYNTHETIC bin weights and acceptance. The charge model computes a normalized standard point-source field with a Gaussian response. Defaults match the website. README records units, precision, truncation and implementation version; expected outputs and a numerical summary are retained.

03Before importing real data

Four CSV schema templates cover clocks, muons, high-γ fields and S301, with units, separate covariance, observed/model/synthetic labels, zero points and reference frames. Real inputs need sources, licenses and SHA-256. Synthetic muon bins must not be treated as a production spectrum or measured counts.

04Versions and contributions

Bind each run to a Prediction ID and canonical source snapshot, retaining Python version and code checksums. Model changes require new expected outputs and ledger entries while preserving old versions. Follow the reproducibility standard. This release provides standards and templates; the full data release remains a later phase.