| SNe Ia are a key standard candle for measuring cosmological distances. To derive accurate distances, SN datasets require precise calibration, modellable selection functions, and a cosmological analysis incorporating multiple types of uncertainties. In this talk, I will describe a recent example of a SN cosmological dataset: the SNfactory sample taken with the SNIFS instrument on the UH 2.2m. These unique data posed several challenges to be useful for cosmology. I will describe how Bayesian analysis frameworks are necessary to extract all the information from SN datasets, from SNfactory to future surveys especially. My new UNITY 1.5 Bayesian framework regresses on the true amount of dust extinction and intrinsic color for each supernova and has a more precise and flexible model for selection effects. This model enables UNITY to recover the selection function of a SN survey with no prior information, although more precise constraints can be realized with knowledge of the survey. With data and model in hand, I show distance measures and their implications for SNe and cosmology. |