Post date: Aug 27, 2013 11:13:13 PM
I edited the bgsr source code to avoid domain errors (e.g., expected allele frequency < 0) and numerical overflow. Some of these issues might have affected results even when they did not cause the program to crash (I need to re-run an analysis with simulated and real data). There are now fewer errors, but at least one numerical issue persists. The error occurs during the offspring genotype update.