NMMA fits for SN 2021ugl (NMMA–Astro-COLIBRI paper)

Reproducibility package for the Results section of "NMMA–Astro-COLIBRI: An Automated Light-Curve Classification Service for Supernovae in the Multi-Survey Era". Contains the nested-sampling posteriors, best-fit parameters, and figures for all four analysis configurations used in the paper's case study (SN 2021ugl / ZTF21abotose), plus the scripts needed to reproduce them from scratch.

• data/SN2021ugl_NMMA_posteriors/, posterior samples, best-fit parameters, and figures for all four configurations:

  • early_time_with_upper_limits/, early_time_detections_only/ (first ~6 days of photometry)
  • full_baseline_with_upper_limits/, full_baseline_detections_only/ (full 47-day baseline)

  See data/SN2021ugl_NMMA_posteriors/README.md for the exact file layout.

• scripts/plot_compare_models_public.py, overlays best-fit light curves from several model runs on one figure (standard matplotlib styling; this is a de-branded, standalone version of the script used to generate the multi-model comparison figures in the paper — see "Relation to the Astro-COLIBRI service" below).

• Makefile — builds MultiNest/PyMultiNest from source and installs the Python environment with uv.

make setup                     # compiles MultiNest + PyMultiNest, uv sync
make fit MODEL=v19-1993j-corr CONFIG=full_baseline_with_upper_limits

This replays the exact lightcurve-analysis command recorded in that model's *_manifest.json. See docs/INSTALL.md for prerequisites and troubleshooting.

The NMMA–Astro-COLIBRI service (https://astro-colibri.science/, docs at https://nmma.live/) that produced these fits runs the same underlying NMMA analysis but renders its own figures with the Astro-COLIBRI app's visual theme and branding. If you just want to submit your own fits through the web interface, no local installation is needed, use the live service directly. The plot_compare_models_public.py script here is a standalone reimplementation using plain matplotlib defaults, so this repository can be run and modified independently without depending on, or redistributing, the Astro-COLIBRI application's front-end styling code, useful if you want to reproduce this paper's specific results offline, or build on the analysis pipeline yourself (see docs/INSTALL.md).

Code is released under the MIT License. Data in data/ is released under CC-BY-4.0, please cite the paper (see CITATION.cff) if you reuse it.

See CITATION.cff, or cite the paper and/or the archived Zenodo release directly: 10.5281/zenodo.21771188 (concept DOI, always resolving to the latest version: 10.5281/zenodo.21771187).

@misc{kiendrebeogo2026nmmaastrocolibriautomatedlightcurvesupernovae,
      title={NMMA-Astro-COLIBRI: An Automated Light-Curve Supernovae Classification Service in the Multi-Survey Era},
      author={R. Weizmann Kiendrebeogo and
              Bernardo Cornejo Avila and
              Sofia Bisero and
              Maxime Cellier and
              Antoine Ciric and
              Ilja Jaroschewski and
              Henrik Rose and
              Alexandre Saint-Paul and
              Fabian Schüssler},
      year={2026},
      eprint={2608.17568},
      archivePrefix={arXiv},
      primaryClass={astro-ph.HE},
      url={https://arxiv.org/abs/2608.17568},
}

@dataset{kiendrebeogo_2026_21771188,
  author       = {Kiendrébéogo, Ramodgwendé Weizmann and
                  Cornejo Avila, Bernardo and
                  Bisero, Sofia and
                  Cellier, Maxime and
                  Ciric, Antoine and
                  Jaroschewski, Ilja and
                  Rose, Henrik and
                  Saint-Paul, Alexandre and
                  Schüssler, Fabian},
  title        = {NMMA fits for SN 2021ugl (NMMA–Astro-COLIBRI paper)},
  month        = aug,
  year         = 2026,
  publisher    = {Zenodo},
  version      = {v1.0.0},
  doi          = {10.5281/zenodo.21771188},
  url          = {https://doi.org/10.5281/zenodo.21771188},
}