search

simonsobs / fgspectra

Foreground SED and power spectrum library for building cross frequency power spectrum models
BSD 3-Clause "New" or "Revised" License
8 stars 7 forks source link

Default params #24

Open mabitbol opened 4 years ago

mabitbol commented 4 years ago

Did we ever define a set of defaults? For example for nu_0s? Or do we leave this completely to the user?

dpole commented 4 years ago

This is up to us. I prefer to leave them unspecified because those of, e.g., a power law can be very different depending on the component you want to describe. This said, if there is consensus on a specific component having the same set of defaults in most of the case of use, we can decide to set parameters to some defaults.

msyriac commented 4 years ago

Sorry to resurrect this, but I think it's useful to have a set of defaults corresponding to choices made in important papers like Dunkley et. al. This would allow for painless quick loading of foreground spectra for people who are trying to do quick science exploration.

dpole commented 4 years ago

This issue actually referred to how to algorithmically handle defaults, so maybe we should move your question to a new issue. In brief, the reason we didn't assign defaults is that many classes are used in different contexts (think of the power law). Something we can do is to implement models that are aliases with the parameters fixed. @msyriac @xzackli @mabitbol what do you think?

mabitbol commented 4 years ago

Aliasing some classes with preset parameter values might make sense. Although the FG parameters vary somewhat with sky patch / fraction, so we'll have to make some choices there.

dpole commented 4 years ago

10 x 👍 for this statements. This links to the general open question of having an official sky mask.