Open tglauch opened 3 years ago
It is actually one of the reason I wrote Legend.py. In LDDE, luminosity dependent density evolution, evolution depends on both L and z. Legend did something similar:
If I perform a normalization to the 1D redshift distribution, it should be able to perform just as FIRESONG.
The way that FIRESONG is currently implemented this is not super important, but in general, it would be nice if FluxPDF could also deal with non-factorable luminosity functions, i.e., those that do not fulfill p(z, L) = p(z) * p(L). Technically, this simply requires us to sum over all combinations of luminosity and redshift. We did have a working implementation for it at some point, but I fear we have lost it during migration so FIRESONG. Hence, a re-implementation would be necessary.