PWGHF: Implement efficiency computation macro for D2H #3616
Conversation
PWGHF/D2H/Macros/efficiency.py
Outdated
| for i_bin in range(1, n_bins+1): | ||
| h_eff.SetBinContent(i_bin, teff.GetEfficiency(i_bin)) | ||
| # TH1 can't handle asymmetric errors so we take the max | ||
| err_max = max(teff.GetEfficiencyErrorLow(i_bin), teff.GetEfficiencyErrorUp(i_bin)) |
There was a problem hiding this comment.
I'm wondering in which case it can happen that the efficiency has an asymmetric error. If I am not wrong, the binomial uncertainty is always symmetric right?
Just to be sure that I'm not overlooking anything, the code looks fine.
There was a problem hiding this comment.
Hi Mattia, thanks for the reply. Indeed, I must say I could not think of a case where this would happen, but I put it as extra safety after reading about it in a ROOT forum. But, in our use case, I don't think that we truly need it, if I am not misssing something.
There was a problem hiding this comment.
Hi Mattia, thanks for the reply. Indeed, I must say I could not think of a case where this would happen, but I put it as extra safety after reading about it in a ROOT forum. But, in our use case, I don't think that we truly need it, if I am not misssing something.
Hi @AlexBigO , exactly! This is my point.
What about adding a warning when this happens? Something like:
if abs(teff.GetEfficiencyErrorLow(i_bin) - teff.GetEfficiencyErrorUp(i_bin))/teff.GetEfficiencyErrorLow(i_bin) > 1e-05: print a warning
Something similar to this. What do you think?
There was a problem hiding this comment.
Ah, this is better indeed! Thanks, I implemented it, and will push it along with future comments.
I just put a "sensitivity" of 1e-02, below this, I get a warning.
mfaggin
left a comment
mfaggin
left a comment
There was a problem hiding this comment.
thanks a lot @AlexBigO , good that this macro can become available for everybody! For me it looks fine, but I leave the time to @fcatalan92 to have a look.
I just have 1 question for my understanding, see the dedicated comment.
Best regards.
Mattia
…en converting TEfficiency to TH1, and add pt binning option
|
Some comments on my latest commit:
|
|
It looks good to me. @fcatalan92 can you have a look as well? |
…ds, try solving MegaLinter errors
|
Sorry for the many attempts to solve the MegaLinter issues, the |
AlexBigO
requested review from
adubla,
alibuild,
fcatalan92,
fcolamar,
fgrosa and
mmazzilli
as code owners
AlexBigO
requested review from
NicoleBastid,
deepathoms,
gtaillepied and
nzardosh
as code owners
) * Implement efficiency computation macro for D2H * Add new line at the end of config file * Fix bug in TH2 projection, add safety for asymmetric error setting when converting TEfficiency to TH1, and add pt binning option * Add missing loop on all pt bins after rebinning * Remove scaling after rebinning (not nedded), re-organise helper methods, try solving MegaLinter errors * Add safety for canvas file extension and try to solve mypy errors * New attempt to solve mypy errors * Try to fix type of save_canvas() last argument + import-error on modules * Add copyright + attempt to solve mypy error * Attempt to solve pyright error #1 * Second attempt to solve pyright error * Apply linters automatic fixes * Apply Vit's comments * Remove empty line between ROOT and style_formatter imports * Fix typo found by Fabio --------- Co-authored-by: Alexandre Bigot <abigot@sbgat402.in2p3.fr>
) * Implement efficiency computation macro for D2H * Add new line at the end of config file * Fix bug in TH2 projection, add safety for asymmetric error setting when converting TEfficiency to TH1, and add pt binning option * Add missing loop on all pt bins after rebinning * Remove scaling after rebinning (not nedded), re-organise helper methods, try solving MegaLinter errors * Add safety for canvas file extension and try to solve mypy errors * New attempt to solve mypy errors * Try to fix type of save_canvas() last argument + import-error on modules * Add copyright + attempt to solve mypy error * Attempt to solve pyright error #1 * Second attempt to solve pyright error * Apply linters automatic fixes * Apply Vit's comments * Remove empty line between ROOT and style_formatter imports * Fix typo found by Fabio --------- Co-authored-by: Alexandre Bigot <abigot@sbgat402.in2p3.fr>
4 participants
Hello @fcatalan92 @mfaggin !
Following our discussion via email, here is a proposition of macro for efficiency computation.