Constraining dark energy models using Jackknife and Bootstrap resampling
Analyses of type Ia supernovae have helped us shed light on the existence and nature of dark energy. Most of these analyses have relied on Bayesian techniques. In this work, we employ resampling techniques, namely Jackknife and Bootstrap, together with generalised least squares, to analyse supernova data. We first calibrate these techniques using near-ideal mock data and versions of the PantheonPlus data, and compare their performance with Bayesian methods. We find that with near-ideal mock data, Jackknife can yield better constraints. We then apply these methods to constrain parameters of flat $\Lambda$CDM, $\Lambda$CDM, flat $w$CDM, $w$CDM, and flat $w_0\,w_a$CDM models from the PantheonPlus and SH0ES (PPS) data. We observe that constraints obtained with different techniques for three- and four-parameter models are largely consistent. We also find that the Hubble tension is less significant with constraints from the Jackknife. For instance, we find that Jackknife estimates $h\,=\,0.743\,\pm\,0.029$ from PPS data, making the Planck value well within $3\sigma$. Moreover, we estimate $h\,=\,0.678\,\pm\,0.052$ from PantheonPlus data when we only consider sources with $z > 0.01$, in which case there is no tension with the Planck estimate. These results highlight the importance of using multiple techniques while analysing data and warrant further investigation.