GAL-7A SPARC Rotation-Curve Fits: Matched 143-Galaxy Subset and Full 175-Galaxy Reproducibility Package

This dataset and reproducibility package supplements the preprint “A One-Parameter Geometric IR-Transition Model for SPARC Rotation-Curve Phenomenology” (DOI: 10.5281/zenodo.21510387). The package provides GAL-7A rotation-curve predictions and validation diagnostics for the full sample of 175 SPARC disk galaxies, comprising 3389 valid rotation-curve points, as well as for the matched 143-galaxy negligible-bulge subset considered by Glass, Zimmerman, and Yahalom (2024). The baseline protocol uses fixed stellar mass-to-light ratios, Upsilon_disk = 0.5 and Upsilon_bulge = 0.7, together with one global galaxy-scale infrared acceleration parameter. No galaxy-by-galaxy chi_gal, radial chi_gal(r), per-galaxy mass-to-light optimization, or additional transition-shape parameter is used. The full-sample global value is chi_gal = 9.078706 × 10^-11 m s^-2. Under the matched fixed-baryonic-input statistical weighting protocol, the weighted RMS is 0.130819 dex for GAL-7A and 0.177232 dex for a single global constant-acceleration benchmark. In the N_MC = 500 Monte Carlo analysis, the median global scale is 8.992754 × 10^-11 m s^-2, with a 95% interval of [6.250027 × 10^-11, 1.384605 × 10^-10] m s^-2. The deposit contains a technical validation report, individual rotation-curve atlases for the matched 143-galaxy subset and the full 175-galaxy sample, point-level predictions, galaxy-level metrics, reproducibility code, documentation, and SHA-256 integrity checks. This package does not claim statistical superiority over RAR/MOND or over the published galaxy-by-galaxy retarded-gravity fits. It establishes that GAL-7A can be evaluated on the same 143-galaxy subset and on the full 175-galaxy SPARC sample using one global acceleration scale and fixed baryonic inputs. It further demonstrates that a single global constant acceleration addition is not empirically equivalent to the GAL-7A transition under the same restricted protocol.