GNILC dust, new small scales implementation

[zonca]

• notebook from @giuspugl
• fmt: black formatting
• executed notebook from @giuspugl
• notebooks not used anymore in gnilc dust
• generate artifacts and templates
• add documentation to gnilc notebooks
• new inputs for d11
• d11 support galactic mask and new implentation of gnilc dust
• docs: link to new notebooks
• cleanup cell
• slurm files to run notebooks
• set new template name in d9 and d10
• set lmax to 2.5 nside

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[zonca]

See https://github.com/galsci/pysm/issues/101

[zonca]

ok, I have modified giuspugl notebook to have a fixed seed and used it to compare with my 2-step implementation (first generate inputs in spherical harmonics space, basically inputs to d11, then from that create the d10 templates), for reference:

https://gist.github.com/c5f71634f4466a52945eb22b0044a3c2 https://gist.github.com/3bf14002bb54d2959c9cb02de48674fe

[zonca]

I have re-created the templates also for Td and beta using the new modulation maps and also applying modulation outside of the GAL080 mask.

[zonca]

See the new d9 and d10 templates in:

https://portal.nersc.gov/project/cmb/pysm-data/dust_gnilc/

the inputs to d11:

https://portal.nersc.gov/project/cmb/pysm-data/dust_gnilc/raw/

also accessible directly at NERSC at:

/global/project/projectdirs/cmb/www/pysm-data/dust_gnilc

[zonca]

unit tests are passing. Now testing at NERSC the one that requires a lot of memory and is disabled here:

pysm3/tests/test_dust_pysm3.py::test_d10_vs_d11

[zonca]

d10 vs d11 passes at NERSC both at 353 and 857