Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Implementation of the Magnitude Kernel and the RSD Kernel in jax_cosmo #55

Open
wants to merge 7 commits into
base: master
Choose a base branch
from

Conversation

bhorowitz
Copy link
Collaborator

I implemented the new kernels to the best of my knowledge, based on the formulas here: https://arxiv.org/pdf/1812.05995.pdf

One thing I had to do for the RSD version is change the kernel to take a z1 variable (corresponding to z_{ell+1}).

NOTE! I wasn't able to get any of these functions working in CCL so I don't really have anything to compare directly against. In particular the s(z) function for the mag_bias I have no idea what a reasonable version would look like...

@bhorowitz
Copy link
Collaborator Author

It would be nice to not bother calculating the z1 component unless has_RSD is enabled, but idk how to check that from the angular cl function.

@EiffL EiffL self-requested a review June 11, 2020 22:26
@EiffL
Copy link
Member

EiffL commented Jun 12, 2020

So I was looking into the RSD part first, I see the issue with z1, and I managed to compare to CCL quickly, here are clustering cls with and without RSD:
image

image
So not exactly that ^^' but I suspect it might just be a scaling factor.

I see what you have implmented, and by curiosity I went to see how CCL does it, and they explicitly keep track of bessel derivatives in the integral: https://github.com/LSSTDESC/CCL/blob/60704b5245689e178a5e32fc0504156d8bb7ac1c/pyccl/tracers.py#L324
So, may have to adopt a similar mechanism somehow, so that we don't have to carry this z1 variable when not needed.
I'm thinking more about this ^^' Will check the mag kernel first

@bhorowitz
Copy link
Collaborator Author

bhorowitz commented Jun 16, 2020

EDIT: I fixed a few more bugs, and RSD now has very minimal effects on this example. Can you send me the pyccl output and I can play around some more?

@EiffL
Copy link
Member

EiffL commented Jul 21, 2021

@all-contributors please add @bhorowitz for code

@allcontributors
Copy link
Contributor

@EiffL

I've put up a pull request to add @bhorowitz! 🎉

@dkirkby
Copy link
Collaborator

dkirkby commented Nov 3, 2022

What is the status of this PR @EiffL?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants