Fixed DL3 creation complaining about mismatch in binning.

[jsitarek]

It sometimes happened that the IRF files have different bin ranges saved despite using the same input card. The difference is very small, comparable to the numerical accuracy of double precision. I saw it in the files processed in CNAF and my suspicion is that it comes from inhomogenuity of machines over which the code was run. Now if different binning is detected, it checks also if the relative difference exceeds somewhere 10^-15 and only then raises an error.

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 4 lines in your changes missing coverage. Please review.

Project coverage is 77.18%. Comparing base (76ab089) to head (8458bcc).
Report is 5 commits behind head on master.

Files with missing lines	Patch %	Lines
magicctapipe/io/io.py	0.00%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #273      +/-   ##
==========================================
- Coverage   77.27%   77.18%   -0.09%     
==========================================
  Files          22       22              
  Lines        2618     2621       +3     
==========================================
  Hits         2023     2023              
- Misses        595      598       +3     

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

[Elisa-Visentin]

Seems fine to me. Why the changes in pre-commit config?
As for tests, I will work again on coverage and new tests after finishing auto-mcp (not to have too many open projects), so we can probably neglect it

[jsitarek]

pre-commit was failing when I was trying to commit the change to github. For some reason it was trying to use python3.1, no idea why. This solved the problem (and allowed me to commit the changes) and I think it is harmless because we also require python 3.11 in the environment. But I am open to other solutions as well.

[Elisa-Visentin]

I was just curious (there were no failings in the other PRs, and you just added an if/else, so this is quite strange). BTW, if it works, it is fine