Add Agora and Half dome SZ models and comparison notebook

[zonca]

No description provided.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[zonca]

see the comparison notebook: https://github.com/zonca/pysm/blob/6cc039cf5320e09740beee9082033fedd45e8a71/docs/preprocess-templates/verify_templates/compare_websky_agora_sz.ipynb

[zonca]

As discussed on the call we would like to have one more person take a look at this before merging if you are interested please reply on the issue. The idea would be to execute the Notebook yourself and look for inconsistency. It is also useful to double check that the documentation is accurate.

[eunseongleee]

I will work on this!

[eunseongleee]

I encountered a small issue when generating the maps, as the notebook could not find the appropriate data directory in which to save the files. I resolved this by adding os.makedirs(os.path.dirname(cache_file), exist_ok=True) before writing the maps, although this may simply be due to the notebook being located temporarily in a different folder. Aside from this, I have reviewed and executed the notebook and can confirm that it runs correctly with no inconsistencies. The documentation describing the different simulations appears accurate.

[alexlague]

I can also check the notebook once the kSZ model from half-dome is available (looking at issue 204, it may not yet be finalized).

[zonca]

@alexlague I would like to merge this first, so if you can review the notebook now I think it is better

[alexlague]

@zonca Eunseong's fix also worked for me and the notebook executed without issues.

[zonca]

ok, can you do a PR with that fix?

[zonca]

@eunseongleee @alexlague we have a PySM call tomorrow, can you coordinate between the 2 of you and prepare a presentation for 10/15 min about the new components? either 1 of you can present or both, you can do slides or just walk through notebooks. We want to give an overview of the new components to the group and gather some feedback.

[alexlague]

@zonca Sure, we can share a few slides. Eunseong volunteered to cover tSZ, and I can do kSZ. (Currently working on the PR mentioned above.)

[zonca]

discussed today during the panexp call. added a documentation fix in 336da2b mentioning the monopole, the fact that they are different realization and the large scale features in websky kSZ.

@seclark @brandonshensley this is ready to merge, can you please take a final look at the docs?
https://github.com/galsci/pysm/pull/244/changes#diff-d606c1bcd72689c8bb949dcdafe4bcd0420da5cbaa7daf9614f0757734cffaf7

[zonca]

Docs preview for the updated SZ section: https://pysm3--244.org.readthedocs.build/en/244/models.html#sunyaevzeldovich-emission

[brandonshensley]

Docs preview for the updated SZ section: https://pysm3--244.org.readthedocs.build/en/244/models.html#sunyaevzeldovich-emission

Looks great! Just a few notes on the docs:

1. It looks like the original paper calls the sims "Websky" (not WebSky). Should we change throughout?
2. Is tsz1 lensed or unlensed? The docs note this for some of the other models, thought not tsz4 or ksz1.

[mremazei]

It might be useful to mention in the docs (if this is not already done somewhere) that some of these SZ models are correlated with other components of the PySM, such as some of the CIB and CMB lensing models. This is the case, I believe, for the Websky-based templates, while other templates correspond to independent realizations of the LSS.

It might be worth warning users that if they run PySM simulations using the lensed CMB realization from Websky but the tSZ model from Agora (for example), they should not expect correlations between them, as they come from different realizations of the LSS.

[seclark]

Agreed with @mremazei -- it might be worth an explicit "recipe" for which components are correlated with one another (especially because it's not as simple as different components with the same number).

Also it looks like the presets.cfg link points to the wrong place?

[zonca]

Thanks for the reviews. Addressed the doc comments in this PR:

• Renamed docs prose from WebSky to Websky
• Fixed the presets.cfg link in the model summary
• Clarified the realization / lensing status of tsz1, tsz4, and ksz1
• Added guidance on which Websky, Agora, and HalfDome components are correlated and which combinations should be treated as independent realizations
• Added the SZ comparison notebook to the docs toctree