Regression test currently fails, because I'm comparing against hardcoded arrays. I'll adjust those once we have converged to the right formula.

@mreineck what about writing a unit test that uses both this and blm_gauss to show that they are consistent?
@hivon any feedback on this?

Unit tests are always a good idea.
@zonca, @mreineck I enclose below my reply from Sep 4 about the numerical factors

-------- Forwarded Message --------
Subject: Re: [healpy/healpy] Try to make Gaussian blm consistent (PR #885)
Date: Mon, 4 Sep 2023 14:29:24 +0200
From: Eric Hivon [email protected]
To: reply+ABKVQDMCLQKEQUOTOD7QXQODALZ6HEVBNHHHAD7NIU@reply.github.com

Hi @mreineck,
this is the factor I use for E and B beams, based on the Challinor et al paper.
While it matters for theoretical studies, it won't matter as much for data analysis, where
the temperature and polarization gains and beams are calibrated separately.

Cheers,
Eric

On 04/09/2023 13:37, mreineck wrote:

Fixes #861 #861

I'd love to get some independent opinions on this, especially on the normalization factor for the polarised components.

@hivon https://github.com/hivon, do you think this is the right expression?

---

    You can view, comment on, or merge this pull request online at:

#885

    Commit Summary

• 4993ab3
  4993ab3
  try to address issue sphtfunc.blm_gauss() is not consistent with the rest of healpy #861

  File Changes
  

(1 file https://github.com/healpy/healpy/pull/885/files)

• M healpy/sphtfunc.py
  https://github.com/healpy/healpy/pull/885/files#diff-9e049750776ba9d889d8e303d1dfe3ef1cb9a95ffe6340f2a7438222579200e8
  (6)

  Patch Links:
  

• https://github.com/healpy/healpy/pull/885.patch

• https://github.com/healpy/healpy/pull/885.diff

—
Reply to this email directly, view it on GitHub #885, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABKVQDNKSDULNGYRLD3KQT3XYW4OHANCNFSM6AAAAAA4KJDASY.
You are receiving this because you were mentioned.Message ID: healpy/healpy/pull/[email protected]