Integrate halo model for intrinsic alignments in 3x2-point systematics. #400
Conversation
# Conflicts: # firecrown/likelihood/gauss_family/statistic/source/weak_lensing.py
# Conflicts: # examples/des_y1_3x2pt/des_y1_3x2pt_PT.py # examples/des_y1_3x2pt/des_y1_cosmic_shear_TATT.py # firecrown/likelihood/gauss_family/statistic/two_point.py # firecrown/modeling_tools.py
# Conflicts: # firecrown/modeling_tools.py
…insic alignments. - Currently, the IG contributions do not match.
firecrown/modeling_tools.py
Outdated
| @@ -25,13 +25,21 @@ def __init__( | |||
| pt_calculator: Optional[pyccl.nl_pt.EulerianPTCalculator] = None, | |||
| pk_modifiers: Optional[Collection[PowerspectrumModifier]] = None, | |||
| cluster_abundance: Optional[ClusterAbundance] = None, | |||
| hm_definition: Optional[pyccl.halos.MassDef] = None, | |||
There was a problem hiding this comment.
I think passing the HMCalculator object directly (or a wrapper around it), similar to the PTCalculator, would make the API more future proof and also abstract away these CCL details.
|
Hi @marcpaterno and @vitenti, given the new update of Firecrown with factories, I would like to update this PR and try to finalise it. Is this a lot of work, and would I need to do this by myself? Edit: I already had a quick look at the new code and solved most of the conflicts. I would still need to test if my developed code runs okay with the new firecrown structure but any comments on this are welcome. |
# Conflicts: # examples/des_y1_3x2pt/des_y1_cosmic_shear_TATT.py # firecrown/likelihood/gauss_family/statistic/source/source.py # firecrown/likelihood/gauss_family/statistic/source/weak_lensing.py # firecrown/likelihood/gauss_family/statistic/two_point.py # firecrown/modeling_tools.py
|
@chrgeorgiou, please go ahead and finish your commits that will solve the conflicts. Please also run the pre-commit checks, which will let you locally see the same results that the GitHub CI will produce. You can do this with If you are struggling too much to get the checks passed, please go ahead and push the commits anyway, and we may be able to help solve some outstanding issues. |
This PR integrates the halo model for intrinsic alignments as a potential systematic for 3x2-point likelihoods. It adds a new
WeakLensingSystematicclass calledHMAlignmentSystematicand the appropriate behaviour for when a tracer has this systematic attached to it. It also extends theModelingToolsfor computing halo model-related quantities. Finally, it computes the correct power spectrum given the input tracer.