Simulating Large Populations of Rare Binaries with POSYDON

A prominent roadblock in the study of binary stellar systems is efficiently generating large populations of rare binaries. We significantly improve the efficiency of this process in the POSYDON population synthesis code by implementing the STROOPWAFEL adaptive importance sampling algorithm. We use this algorithm to generate samples of rare binary systems like Double Compact Object (DCO) Mergers and Low-Mass X-ray Binaries (LMXBs) more efficiently enabling more detailed analysis of these systems. We find a 12x and 30x improvement in the sampling of 100 thousand binaries for LMXBs and DCO Mergers respectively.