LSosna/SPARC_Compactness_N123: SPARC Compactness Canonical Dataset (v3.2)

**SPARC Compactness Analysis v3.2 (Canonical Sample, N=123)** This version supersedes v3.1 and provides ApJ submission–ready data, analysis scripts, and figures for the canonical SPARC galaxy sample. It includes updated compactness scaling parameters, full reproducibility scripts, and verified Bayesian model comparison results. **Key Updates**• Updated compactness slope: s = 0.718 ± 0.020 (EIV/ODR regression)• Updated mass–size exponent: α = 0.282 ± 0.020• Hierarchical scatter decomposition: σ_obs = 0.211, σ_meas = 0.094, σ_int = 0.189 dex• Full Bayesian model comparison: ΔBIC = +263 653 (RAR strongly favored)• ApJ-safe language and reproducibility compliance• SHA-256 checksums included for all core assets **Included Files**Canonical parameters (`canonical_FINAL_v3.2.json`);Canonical dataset (`SPARC_Canonical123.csv`);Reproducibility code (`SPARC_Compactness_Analysis_v3.2.py` & `.ipynb`);Verification (`BIC_Verification_summary.json`);Figures 1–4 + S1 (`scatter_decomposition.png`);Integrity (`CHECKSUMS.txt`). **Supplementary Improvements**Intrinsic scatter decomposition (σ_int = 0.189 dex);Updated figures 1–4 + new Figure S1;Clean reproducibility notebook (Colab);Fixed random seed and supplementary paths. **Author:** Lukas Sosna — Independent Researcher **ORCID:** https://orcid.org/0009-0004-1266-9729 **License:** Creative Commons Attribution 4.0 (CC BY 4.0) **Reproducibility Instructions**To reproduce all figures and tables, run:```bashpython SPARC_Compactness_Analysis_v3.2.py```Requires: Python ≥ 3.10 with NumPy, SciPy, pandas, matplotlib.