Ssc fsky #115
Ssc fsky #115
Conversation
…ust still add tests
| nfw = ccl.halos.HaloProfileNFW( | ||
| mass_def=mass_def, concentration=cM, fourier_analytic=True | ||
| ) | ||
| hmc = ccl.halos.HMCalculator( |
There was a problem hiding this comment.
We should move this out of this function. That way we can reuse whatever it's already computed in each hmcalculator. But we can open an issue later.
| from .covariance_builder import CovarianceFourier | ||
|
|
||
|
|
||
| class FourierSSCHaloModelFsky(CovarianceFourier): |
There was a problem hiding this comment.
This class is basically the same as FourierSSCHaloModel except for a few lines. I would not create a new class but check whether fsky is defined in the SSC configuration that you retrieve in the __init__. If fsky is there, then you compute the sigma2_B with the fsky method; otherwise with the mask.
Alternatively, I would add a method to the class that returns the sigma2_B, inherit from this class in your FourierSSCHaloModelFsky and overwrite that method.
There was a problem hiding this comment.
Right, I agree. I can make some modifications to reduce the code duplication. Should I make this on a new branch then?
There was a problem hiding this comment.
As you prefer. You can create a new branch and merge it onto this one or work directly on this one. I would probably do the later. It's an easy modification so in one or two commits should be done.
This PR includes a simple extension of the SSC covariance using the Fsky approximation. The user will no longer require a mask to compute this term to the non-Gaussian covariance, also enabling a full comparison between the approximation and a mode-coupling formalism.