Fork me on GitHub

Opened 10 years ago

Last modified 10 years ago

#401 new Task

jet reconstruction in central region

Reported by: Jessica Metcalfe Owned by:
Priority: critical Milestone:
Component: Delphes miscellaneous Version: Delphes 3
Keywords: Cc:

Description

Hi,

I am using the Delphes ATLAS card with pileup and I have modified the card for an "extended tracker" scenario where I have taken the tracking efficiencies, reconstruction efficiencies, pT and energy resolutions, etc. and extended the range of reconstruction out to |eta|<4. I am matching the two leading jets (in pT) to GenJets by a deltaR matching for a same-sign WW scattering sample that has a large forward jet signature. When I plot a scatter plot of the reconstructed jet pT versus the GenJet pT for well-matched jets, 0 < DeltaR(RecoJet-GenJet) < 0.1, I see more jets in the central region with a fairly sharp drop at |eta|=2.5. I do not understand why this is the case? Could it be due to the pileup subtraction in Delphes? or segmentation of the calorimeter? something else? I've attached the plot here. I'd greatly appreciate any help as soon as possible as we are trying to make a deadline for approval.
Thanks!
Jessica

Attachments (1)

nonPUtagRECOptvseta0_ssWWEW_0PU_8TeV_wdR_truth-match_0dR01.pdf (32.2 KB ) - added by Jessica Metcalfe 10 years ago.

Download all attachments as: .zip

Change History (3)

comment:1 by Michele Selvaggi, 10 years ago

Hi Jessica,

without knowing the specifics of your analysis, I will have to guess. From your explanation I understand that the leading jet should lie mostly in the forward region, and you don't match efficiently there.

One simple reason could be that you are using a very aggressive matching requirement, especially when compared to the calo resolution in the forward region. Also the fact that no charged pile-up subtraction is performed outside the tracker region could lead to a large smearing of the jet direction. What looks strange though, is that you seem to have a good matching in the "old tracker region" leading to me to suspect that the rho pile up subtraction is performed in a wrong eta range.

Could you attach the card you are using, together with the version? Also, if still possible time wise, I would strongly suggest to switch to the latest version of Delphes (3.2.0) , since several improvements on pile-up subtraction have been made recently.

Michele


comment:2 by Michele Selvaggi, 10 years ago

Hi Jessica,

nothing strikes me as being wrong in the pile up subtraction section on your card.
What you might be due to the calorimeter cells becoming of size 0.2 starting from 2.5, which means as I said previously that your matching criterion is too aggressive. Could you try reproducing it the plot with DR ~ 0.3-0.4?

Michele

Note: See TracTickets for help on using tickets.