SYNAPPS: Data-Driven Analysis for Supernova Spectroscopy

We introduce a new computer program, SYNAPPS, for forward-modeling of supernova (SN) spectroscopy data sets. SYNAPPS is a spectrum fitter embedding a highly parameterized synthetic SN spectrum calculation within a parallel asynchronous optimizer. This open-source code is primarily aimed at the problem of systematically interpreting large sets of SN spectroscopy data. While SYNAPPS should be useful to current SN spectroscopy efforts like the Nearby Supernova Factory, Lick Observatory Supernova Search, Palomar Transient Factory, Harvard Center for Astrophysics SN program, and so on, it could also benefit future similar efforts connected to the Dark Energy Survey, Panoramic Survey Telescope and Rapid Response System, or the Large Synoptic Survey Telescope. Smaller programs are also potential users. SYNAPPS illustrates the potential for data-driven discovery enabled by high-performance computing, where complex physical systems are directly constrained by large information-rich sets of scientific observations. Here, we discuss the motivation of our approach, outline the structure of the code, present some example calculations, and describe a few enhancements in terms of physics modeling, optimization, and computing that we will be pursuing for the future.