Evidence and reproducible code for Thermo-Axiomatic Geometry at the Lβ′/LβI Boundary in Hydrated DPPC: A Conditional Geometric–Kinetic Correspondence
Description
This paper-specific archive contains 47 research case records spanning calculations, estimator and null-model audits, mathematical derivations, candidate screening, and literature/source assessment. The 47 records are not 47 experiments, independent validations, or successful applications. The archive also contains file-level evidence maps for 19 cases, provenance-labelled transcribed source-table inputs, dated reconstruction code and outputs, a preserved historical implementation and test suite, and a 60-row DPPC source audit. No new experimental measurements are reported. Copyrighted source publications and PDFs are not redistributed. The available record does not identify a termination coordinate for the hydrated-DPPC Lβ′/LβI boundary; the associated article concludes that the proposed cross-arm correspondence is presently non-evaluable rather than positive or negative.
Files
Steps to reproduce
Download and extract DPPC_TAG_Mendeley_v1_20260917.zip, then open a terminal in the extracted DPPC_TAG_Mendeley_v1 directory. Python 3.12 is recommended; the archive was clean-tested with Python 3.12.14. Create an isolated environment and install the exact dependencies: macOS/Linux: python3 -m venv .venv .venv/bin/python -m pip install -r requirements.txt Windows: py -3.12 -m venv .venv .venv\Scripts\python.exe -m pip install -r requirements.txt First verify file integrity, record counts, evidence mappings, the approved article title, and the absence of transient internal paths: macOS/Linux: .venv/bin/python scripts/verify_package.py Windows: .venv\Scripts\python.exe scripts\verify_package.py The expected final line is: PASS: package hashes, counts, mappings, title and public-path checks Then run the complete reproduction workflow: macOS/Linux: .venv/bin/python scripts/reproduce_all.py Windows: .venv\Scripts\python.exe scripts\reproduce_all.py The wrapper works in a new local_runs/<UTC timestamp>/ directory, so preserved outputs are not overwritten. It runs the source-table reconstruction; 47 numerical regression checks (these are distinct from the 47 research case records); synthetic method audits; the F04 arithmetic reconstruction; the A05/D05/E07 available-case checks; and the 89-test historical implementation suite. It then compares 22 declared regenerated outputs byte-for-byte with the preserved outputs. The expected final status is PASS. Full task logs and run_summary.json are written under the timestamped local_runs directory. After the dependencies have been installed, the workflow requires no network access and no source PDFs. Reproduction begins from the included provenance-labelled transcribed rows. Copyrighted source publications are not redistributed, and no new experimental measurements are reported.