DAEMONFLUX: DAta-drivEn MuOn-calibrated atmospheric Neutrino Flux

In this paper, we present a refined calculation of the atmospheric neutrino flux spanning from GeV to PeV energies. Our method, Daemonflux, utilizes data-driven inputs and incorporates adjustable parameters to take their uncertainties into account. By optimizing these parameters using a combination of muon data and constraints from fixed-target experiments, we achieve uncertainties in the calculated neutrino fluxes of less than 10% up to 1 TeV, with neutrino ratios constrained to below 10%. Our model performs particularly well at energies below 100 GeV, where the smallest errors are obtained. We make our model available as a software package that provides access to predictions of fluxes, ratios, and errors, including the covariance matrix obtained from the fit.