Replies: 1 comment 4 replies
-
|
One fundamental question for me is whether the perturbations API will normally be a separate object or assumed to be part of the cosmology object. The advantage of the former is that it becomes easy to code up something like growth function and an Eisenstein-Hu power spectrum and call that your perturbations, independent of background. Emulators fall into the same category. The disadvantage is that Boltzmann codes like CAMB or CLASS would provide both background and perturbations API in one, requiring either wrapping one of their instances in two objects, or passing the same object twice. |
Beta Was this translation helpful? Give feedback.
4 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
We should start thinking about a cosmoAPI for perturbations. Let's open a couple of threads in this discussion with the relevant questions.
Beta Was this translation helpful? Give feedback.
All reactions