nnpdf icon indicating copy to clipboard operation
nnpdf copied to clipboard

Implementation of LHCB WPWM 5.02TeV

Open andrpie opened this issue 3 months ago • 1 comments

Implementation of the 2509.18817 future test dataset as per #2368.

The implementation and methodology was discussed with @kamillaurent on a call.

Points worth mentioning:

  • implemented two observables: differential cross functions in muon transverse momentum, one for W+, another for W-. Not sure whether "DIFWP", "DIFWM" are good choices for observable names
  • HEPData files are rather unconventionally named - instead of table numbers, there are names that only differ by sign: "+", "-". Thus the table fields in metadata.yaml contain these signs.
  • just to double check: is "DY_CC_PT" the right process type for both of the observables?
  • I've noticed that some other implementations translate pb to fb - is it necessary?
  • The luminosity uncertainty of 2% is implicitly mentioned in the paper and added to the central values. It's classified as MULT, LHCBLUMI5P02TEV (correlated across the experiment)

andrpie avatar Nov 06 '25 15:11 andrpie

Points worth mentioning:

* implemented two observables: differential cross functions in muon transverse momentum, one for W+, another for W-. Not sure whether "DIFWP", "DIFWM" are good choices for observable names

That's fine with me. Perhaps slightly better DIF_WP and DIF_WM, though I have no strong preferences.

* just to double check: is "DY_CC_PT" the right process type for both of the observables?

Yes.

* I've noticed that some other implementations translate pb to fb - is it necessary?

In principle it is not. Because we want to match data to FK table predictions, the unity of measure is dictated by the default unity with which grids (and FK tables) were generated. So for the time being I would not convert pb to fb. We can reassess this when we will have theoretical predictions.

* The luminosity uncertainty of 2% is implicitly mentioned in the paper and added to the central values. It's classified as MULT, LHCBLUMI5P02TEV (correlated across the experiment)

Very good. We don't have any other LHCb measurements at 6.02 TeV, so you have total freedom in the choice of the label for the correlated luminosity uncertainty.

enocera avatar Nov 09 '25 20:11 enocera