About tau tagging in Delphes.

[Jung Chang]

Dear Sir,

Recently, I used the Delphes with MG5 and tried to reconstruct tau. I had generate a simple processes wo test the tau tagging in Delphes. p p > W+ h, h > tau+ tau-, W+ > l+ \nu_l, and plot the IM_tau, tau, hope to see a resonance peak at mH. It seems after delphes can not reconstruct IM_tau,tau with a peak at mH. I had fixed some value with tau tagging, i.e. set TauPTMin 40.0, set tagging efficiency of tau as 1. But still not work. So I just want to ask, how can I do with Delphes for the tau? Thanks a lot!

[Michele Selvaggi]

Hi,

do you mean that you do not see a peak, or that the peak is shifted?
As you may know, the energy of a reconstructed jet is often below that of its parton-level counterpart.
In particular, a tau-jet is composed of the visible decay products of the original tau lepton, therefore a significant fraction of energy is missing due un-detected neutrinos.
To correct for this effect, you might wan't to play with the jet energy scale correction:

module EnergyScale JetEnergyScale {
   set InputArray FastJetFinder/jets
   set OutputArray jets
	
   # scale formula for jets
   set ScaleFormula {1.08}
}

In this simple example (already included in the default CMS and ATLAS cards), there is a constant energy scale (x1.08) applied to all jets.
You can start by changing this value. You can also apply a (pt,eta) dependent correction. If you want to apply an energy scale only to tau-jets, I suggest you first create a collection that filters out non-tau-jets and then apply the energy scale.

To create a module that creates a collection of tau-tagged jets only, I suggest you have a look at the existing modules StatusPidFilter or ConstituentFilter. Do no hesitate to ask for help if you have any issue with this.

Michele

[Jung Chang]

Hi Michele,

Thanks a lot for the fast and detailed reply!
The peak is shifted. I got it, and will try it! Thanks!!