Add cref

[jsitarek]

I found in GDAF specification:
https://gamma-astro-data-formats.readthedocs.io/en/latest/irfs/irf_axes/index.html
(see also the example in https://gamma-astro-data-formats.readthedocs.io/en/latest/general/fits-arrays.html#id1)
that the axis order of IRFs is not fully fixed, but only recommended, and that the tools using the IRFS should read it from CREF header key, which so far was not filled in GDAF I/O classes of pyirf.
I added it for all the IRFs in gadf.py file, and added a test function to validate that all the shapes are consistent

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (e591335) 95.36% compared to head (31a668b) 95.40%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #275      +/-   ##
==========================================
+ Coverage   95.36%   95.40%   +0.03%     
==========================================
  Files          60       60              
  Lines        3109     3135      +26     
==========================================
+ Hits         2965     2991      +26     
  Misses        144      144              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[maxnoe]

Hi @jsitarek,

Thanks, this in general looks fine, but given the discussion in open-gamma-ray-astro/gamma-astro-data-formats#102 and the fact that Gammapy only supports the recommended axis order and the general let's say "under specification" of this in the GADF docs, I am not sure we gain something from it...

[jsitarek]

Thank you @maxnoe for a quick review. The issue that you mentioned is still open since 5 years and in the meantime some of the instruments included this CREF in their output IRF files, so even if gammapy is not using it (yet?) I think it makes sense to have it in the file. This also gives information to people checking the irf file what axis order was really used, and allows consistency checks.

[maxnoe]

@jsitarek Ok. Let's include the change for automatically determining the column index and then include it.

I'll also open an issue with Gammapy.

[maxnoe]

There seems to be an issue with the docs builds unrelated to your changes, I'll investigate and come back yo you

[jsitarek]

There seems to be an issue with the docs builds unrelated to your changes, I'll investigate and come back yo you

Hi @maxnoe, any update on the docs issue? If it is unrelated to these changes, maybe this PR can be merged already?

[maxnoe]

Sorry, looking into it, could reproduce locally. The issue is the downloading using a raw cell.