You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, there are two kinds of interpolation table controlled by the UserParams field USE_INTERPOLATION_TABLES.
the Sigma(M) interpolation table
Conditional mass function integral tables.
Since 1. is used to construct 2. and both involve integrals, these become very slow if the flag is set to false, which makes testing the interpolation tables annoying.
We should add a third option to the flag:
"no-interp" (0 in the C backend): No interpolation tables used at all, for testing the sigma tables on limited runs
"sigma-only" (1 in the C backend): Use interpolation tables for sigma, but not the CMF, used for testing the CMF tables and/or slower but more accurate runs
"all-interp" (2 in the C backend): Use all interpolation tables.
While this change should be easy to make, we should wait for PRS #415 and #427 to avoid clashing, since this will change lines all over the place in the backend and frontend.
The text was updated successfully, but these errors were encountered:
Currently, there are two kinds of interpolation table controlled by the UserParams field USE_INTERPOLATION_TABLES.
Since 1. is used to construct 2. and both involve integrals, these become very slow if the flag is set to false, which makes testing the interpolation tables annoying.
We should add a third option to the flag:
While this change should be easy to make, we should wait for PRS #415 and #427 to avoid clashing, since this will change lines all over the place in the backend and frontend.
The text was updated successfully, but these errors were encountered: