jeff-regier/Celeste.jl

Implement asinh magnitudes in `AccuracyBenchmark`

Closed this issue · 0 comments

In scoring accuracy of reference band brightness, we look at error in units of both flux (nMgy) and mags (Pogson magnitudes). There seems to be some consensus that comparing on asinh magnitudes would be better. See #510 for some comments and this page for definitions and discussion.