CosmoFit.DESSN5YRLikelihood

class CosmoFit.DESSN5YRLikelihood(cosmology, version='des-sn5yr', marginalize_offset=True)[source]

Bases: BaseLikelihood, AnalyticOffsetMixin

DES-SN5YR (Dark Energy Survey 5-year) Supernova likelihood.

Unlike PantheonLikelihood (which compares an apparent-magnitude-like observable, m_b_corr, to mu(z) + M_B), DES-SN5YR distributes the SN distance modulus directly (MU, computed assuming a fiducial H0=70). Comparing that to a model mu(z) computed at a different H0 leaves a constant offset – the same H0/absolute-calibration degeneracy Pantheon+ has, just expressed directly in distance- modulus space rather than magnitude space – which is marginalized over analytically the same way (see AnalyticOffsetMixin); this exactly matches the DES-SN5YR data release’s own reference likelihood (DES-Dovekie-SN_Likelihood.py, cov_log_likelihood).

Parameters:

  • cosmology – Cosmology model instance (LCDM, CPL, …).

  • version (str) – DES-SN5YR dataset version.

  • marginalize_offset (bool, optional) – If True (default), the constant distance-modulus offset (fully degenerate with H0) is marginalized over analytically. If False, cosmology.MB is added to the model as an explicit free/fixed nuisance parameter instead (mu_model = mu(z) + M_B).

Warning

Do not combine "pantheon" and "des_sn5yr" in the same Fitter fit. DES-SN5YR’s low-redshift “anchor” sample includes CfA3, CfA4, and Foundation supernovae (~11% of its 1820 SNe, IDSURVEY != 10) that are also compiled into Pantheon+ – fitting both at once double-counts those supernovae (correlated, non-independent data treated as independent), understating uncertainties and biasing any joint result. Use one SN Ia compilation per fit, not both.

References

Popovic et al. (2026), MNRAS, arXiv:2511.07517 – the DES-Dovekie reanalysis these files come from (byte-identical to 4_DISTANCES_COVMAT/DES-Dovekie_HD.csv and STAT+SYS.npz in des-science/DES-SN5YR): the same supernovae recalibrated with the Dovekie photometric calibration, 1820 SNe after selection.

It supersedes the original release (Vincenzi et al. 2024; DES Collaboration 2024, arXiv:2401.02929; 1829 SNe), which is still tagged as version 1.3 of that repository but is not what is bundled here. The two give different distances to the same supernovae, and so different w0-wa constraints – quote the right one.

__init__(cosmology, version='des-sn5yr', marginalize_offset=True)[source]
Parameters:

  • version (str)

  • marginalize_offset (bool)

Methods

__init__(cosmology[, version, ...])

best_fit_offset()

Best-fit additive offset that analytic marginalization would assign, given the current cosmology.

chi2()

Chi-square statistic.

log_likelihood()

Return log-likelihood.

marginalized_chi2()

Chi-square with the offset analytically marginalized out.

model()

Predicted distance modulus.

predictions()

Alias for model().

residuals()

Data minus model residuals.

summary()

Return a summary of the likelihood evaluation.

Attributes

n_data

Number of data points.

name_and_size

Name together with the number of data points.