Add calibration with RMF file

[matteobachetti]

No description provided.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 96.42%. Comparing base (e0b10f1) to head (ba6cc47).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #804      +/-   ##
==========================================
+ Coverage   96.41%   96.42%   +0.01%     
==========================================
  Files          45       45              
  Lines        8989     9016      +27     
==========================================
+ Hits         8667     8694      +27     
  Misses        322      322              

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

[matteolucchini1]

Probably not a big deal for a function this simple, but would it not be a bit more robust to use "with pf.open... as lchdulist" and then not needing to worry about closing the file by hand?

[matteolucchini1]

Otherwise the response stuff lgtm

[matteobachetti]

@matteolucchini1 good point. I often do the explicit open/close because Astropy has a nonstandard behavior when associating the with statement with the default memory mapping of FITS files (see discussion here: astropy/astropy#7404). I modified as requested, but forcing memmap=False, to avoid that bug

[mgullik]

HI @matteobachetti,

Another useful addition to Stingray!

I was wondering if there is an option to load the rmf when Stingray load an event file. A rms_file variable set to None as default.

Another thing, is it worth adding a warning message when we load an event file, saying how the energies are calculated?
I guess that if rmf is not specified the conversion is done using default_nustar_rmf() right?
And why is file is named default_nustar_rmf()? I think it should be general, right?

p.s. one curiosity, which rmf have you chosen to add in the tests?

[matteobachetti]

@mgullik good catch on the default_nustar_rmf, it was a leftover from HENDRICS' documentation with no real effect.
I added the possibility to calibrate directly at read or initialization, I agree it's a simple and very useful addition.

[matteobachetti]

The test rmf file is just taken from an old NuSTAR observation, I don't remember the details 😅

[mgullik]

Thanks @matteobachetti, I can approve this PR.
I only have one last question, if the rmf is not specified what is the default method that Stingray uses to convert PI into Energies? If I'm not wrong the event file doesn't specified the energy of the single event. Wouldn't be good practice to specify, with a message when the event is read, which conversion Stingray uses?