Reproducibility for Experimental Predictions
Make sources, equations, inputs, code, environment, expected outputs and failure criteria standard deliverables for every prediction.
Registry and comparison planning; no new canonical physical result or closed quantitative discriminator.
01Reproduction is more than drawing a curve
Raw-data reproduction, comparison-model recalculation, an independent ASD prediction and an apparatus fit are different outputs. Label every run. Synthetic data cannot enter an observational-support column.
02Minimum reproducible record
| File / record | Required content |
|---|---|
| README | Prediction ID, scope, commands, limits and expected result |
| equations.md | Equation IDs, source versions, inputs and approximations |
| input / provenance | Source, license, units, selection, raw/processed/synthetic labels |
| notebook + script | Common-input interfaces for both models; missing ASD output explicit |
| expected output + figure | Checkable values and figures, without presenting examples as experiments |
| environment | Python/software versions, dependencies and seeds if used |
| checksums | SHA-256 of inputs, code and expected outputs |
| decision protocol | Error budget, correlations, likelihood, rejection region and affected node |
03Frozen parameters and fair comparisons
Record which parameters come from independent measurements and which are fitted here. Separate calibration and test data. Both models use the same apparatus, selection and noise assumptions. If calibration depends on theory, model that explicitly instead of silently sharing a theory-reconstructed “raw field.” Retain failures, zero differences and non-identifiability.
04Scope of this release
The kit recalculates directional geometry, constant-energy standard muon survival, binned acceptance sums and a standard point-charge waveform through an assumed response. It supplies no novel ASD prediction, real-data fit or completed Ota/S301 reproduction. A formal data release is a separate later version.
Position within the framework
These links index inputs and research relationships. The completion status of each derivation is specified in its text. Untranslated supporting pages open in Chinese.
Revision history
Existing canonical module snapshots / 现有正式模块快照 · Predictions & Falsification v1.0 · review 2026-09-22
Download source and scope manifest
Based on versioned canonical modules and cited instrument literature. This experimental program adds operational templates, not new canonical field equations.
Knowledge scopeASD: Reproducibility for Experimental Predictions. ASD-EXP-TESTS-REPRODUCIBILITY-EN. Version 1.0.0; 2026-09-22. https://asd-physics.org/en/tests/reproducibility. Predictions & Falsification v1.0 · review 2026-09-22. Source EXP-7ca61744dcd2.
Continue with the assumptions, equations and sources.
These links carry this page and its source version into the research workspace. Each request pins its knowledge snapshot. Model availability is shown in the workspace; retrieval and private export remain available.