[REDMINE1D-12] [RM-6489] handle the use of fft processing for QSO Created: 04/Jun/21  Updated: 13/Jun/23  Resolved: 13/Jun/23

Status: Done
Project: 1D Redmine
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Normal
Reporter: Redmine-Jira Migtation Assignee: Redmine-Jira Migtation
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified


 Description   

Created on 2021-05-10 13:54:42 by Didier Vibert. % Done: 100

now that fftprocessing is a parameter of the method templatefitting : it can be activated for qso and stars .

But, since #5989, the zstep needs to be derived from the input spectrum sampling (when fftprocessing is activiated), it is the case for galaxies.
Thus the zstep needs to be the same for all objects. But the zgrid does not need to be aligned since the templates for each objects are different, the template sampling can be different.

And symmetrically, if the input spectrum in not log-sampled (this is not the expected behavior inside an integrated pipeline where the 2d processing should handle the log-sampling), it has to be resampled, for each different zsteps => better have a common zstep for galaxy/stars/qso

TODO: handle those aspects. note, in the current state, fftprocessing can be activated for stars and qso, and the behavior is uncertain (it may crash... or lead to incorrect results).
=> at the bare minimum: render fftprocessing impossible with stars & qso : DONE
=> or handle it correctly

 Comments   
Comment by yuki.moritani [ 13/Jun/23 ]

Comment by Didier Vibert on 2021-09-01 10:38:29:
TODO:

  • still forbid fftprocessing for stars. (may be implemented later on)
  • enable fft processing for qso objects for methods templatefittingsolve or linemodel

=> handle properly, the zstep, zrange, and templates/spectra logrebining.

  • if input spectra log sampled:
    zstep is imposed by the spectra sampling. then same behavior as for galaxy, zgrid start comes from parameters and zgrid end becomes the first above zend parameter
    -> rebin the qso templates accordingly, with same method as for galaxies, using lambda_range, to keep template sampling between spectra, even if the spectra are truncated
  • if input spectra has to be rebined & if the qso zstep and galaxy zstep are different, then the there should be in principle 2 rebinned spectra. Instead, we retain the smalled zstep of the two (with a warning) , and use one rebinning with a common zstep. But the zranges can still be different.
Comment by yuki.moritani [ 13/Jun/23 ]

Comment by Mira Sarkis on 2021-09-17 08:47:10:
Merge requests:
https://gitlab.lam.fr/CPF/cpf-redshift/-/merge_requests/209

Changing enablestellarsolve to enablestarsolve , results in merge requests for:

Comment by yuki.moritani [ 13/Jun/23 ]

Comment by Pierre-yves Chabaud on 2021-10-01 12:34:24:
Merged into @develop@ : @11c5021@

Generated at Sat Feb 10 15:28:23 JST 2024 using Jira 8.3.4#803005-sha1:1f96e09b3c60279a408a2ae47be3c745f571388b.