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
First, thank you for all the work building this amazing package.
I read the cookbooks on
how to paint the momentum with a fixed line-of-sight (L.O.S.), and
how to compute multipoles of the density power-spectrum in a realistic survey with varying L.O.S.
I would like to:
compute multipoles of the momentum power-spectrum with a varying L.O.S.
Looking at the code of ConvolvedFFTPower, it looks like it really requires a FKPCatalog or FKPCatalogMesh, which assume that I want the power-spectrum of F = w * (n_g - alpha * n_r).
The solution for my problem would be to be able to provide F = w * n_g * radial_velocity, and then give it to ConvolvedFFTPower for compute the multipoles with a varying LOS. This new definition of F does not require a random catalog, but I need the varying LOS capability of ConvolvedFFTPower.
Do you have any suggestions on solutions or alternative hacks I could use? I am new to this package, so apologies if the solution is obvious.
Many thanks
The text was updated successfully, but these errors were encountered:
A very quick suggestion is to define a different CatalogMesh class, for example, FKPVelocityCatalogMesh. The implementation can be based on (mostly copying) FKPCatalogMesh. In FKPVelocityCatalogMesh.to_real_field, however, you can paint F = w * n_g * radial_velocity, rather than painting the traditional FKP field.
Because your data doesn't come with randoms, I would also suggest __init__ of FKPVelocityCatalogMesh consumes a Catalog object, instead of a FKPCatalog object (which is used by a FKPCatalogMesh.__init__).
First, thank you for all the work building this amazing package.
I read the cookbooks on
I would like to:
Looking at the code of ConvolvedFFTPower, it looks like it really requires a FKPCatalog or FKPCatalogMesh, which assume that I want the power-spectrum of F = w * (n_g - alpha * n_r).
The solution for my problem would be to be able to provide F = w * n_g * radial_velocity, and then give it to ConvolvedFFTPower for compute the multipoles with a varying LOS. This new definition of F does not require a random catalog, but I need the varying LOS capability of ConvolvedFFTPower.
Do you have any suggestions on solutions or alternative hacks I could use? I am new to this package, so apologies if the solution is obvious.
Many thanks
The text was updated successfully, but these errors were encountered: