Adding an implementation of lgamma for speed - #304

Merged

andrewdavidsmith merged 6 commits into

radmeth-builtin-lgamma

Oct 24, 2025

Commits on Oct 23, 2025

src/radmeth/radmeth_model.hpp: fixing the Apple floating point parsing
andrewdavidsmith
committed

Commits on Oct 24, 2025

src/radmeth/radmeth_nano.cpp and src/radmeth/radmeth_optimize_gamma.hpp: removing linear optimization functions and using only the fdf
andrewdavidsmith
committed
src/radmeth/radmeth_optimize_gamma.cpp: removed linear optimization functions and only using the fdf. Reordered some local functions. Added include for radmeth_lgamma and now using that instead of …
andrewdavidsmith
committed
src/radmeth/radmeth_lgamma.hpp: adding an lgamma implementation because the one in glibc is slow and I don't think we need all the branching, etc., which seemed to be pressuring icache
andrewdavidsmith
committed
src/radmeth/radmeth_lgamma.hpp: minor cleanup
andrewdavidsmith
committed
Makefile.am: adding the radmeth_lgamma.hpp source
andrewdavidsmith
committed