Duplicate particle in event?

[Jory]

Hi,

What does it mean when two particles (a tau and a jet) have the exact same eta, phi, and pt? Is this one and the same object tagged twice?

I obtained an event (LHCO output) from Delphes that looks like this:

0 2031 0
1 0 -0.739 -0.393 101.061 0.000 0.000 0.000 0.529 0.000 0.000
2 0 -0.739 -0.480 37.818 0.000 0.000 0.000 41.028 0.000 0.000
3 3 -0.748 -0.391 188.461 17.540 -3.000 0.000 1.360 0.000 0.000
4 4 -0.748 -0.391 188.461 17.540 6.000 0.000 1.360 0.000 0.000
5 4 -2.247 2.007 49.863 0.072 0.000 0.000 inf 0.000 0.000
6 4 -0.304 2.985 33.419 7.101 4.000 0.000 1.310 0.000 0.000
7 6 0.000 2.962 97.941 0.000 0.000 0.000 0.000 0.000 0.000

and also from another run:

0 5986 0
1 3 0.825 0.307 160.036 12.440 1.000 0.000 62.720 0.000 0.000
2 4 0.825 0.307 160.036 12.440 1.000 0.000 62.720 0.000 0.000
3 4 2.697 -2.531 49.239 12.318 6.000 0.000 0.930 0.000 0.000
4 4 -3.399 -1.834 42.792 5.383 0.000 0.000 7.323 0.000 0.000
5 4 -2.777 2.994 23.154 4.639 0.000 0.000 7.914 0.000 0.000
6 6 0.000 2.910 79.908 0.000 0.000 0.000 0.000 0.000 0.000

Thanks,
Jory

[Pavel Demin]

Hi Jory,

If you are using Delphes 2, then it's a jet identified as a tau-jet that is written twice.

Normally, in Delphes 3, tau-jets are written to LHCO only as taus and don't appear as jet.

However, I'm not sure which approach is better.

In the LHCO format jets have a field for b-tag but not for tau-tag.

Regards,

Pavel

[Jory]

Hi Pavel,

What do you mean 'a tau-jet that is written twice'? Usually, so it appears, taus are written once as a tau and not once more as a jet.

Why are these taus written twice (once as tau and once as jet)?

I would say a jet identified as a tau jet is written as a tau (?).
I would say when a tau appears as a jet you would write a jet, not a tau, since you detect a jet (?).

How often does this double writing occur?
Does it occur only with taus?

Thanks,

Jory

[Pavel Demin]

Just to be sure. What version are you using?

[Jory]

Delphes 2.0.3.

[Pavel Demin]

In Delphes 2, in the Delphes.cpp file around lines 711-713, there is the following code:

sorted_jets=JETRUN->RunJets(input_particles, TrackCentral,NTrackJet,EHADEEM,list_of_active_towers,NCALO);

JETRUN->RunJetBtagging(treeWriter, branchJet,sorted_jets,NFCentralQ,NTrackJet,EHADEEM,NCALO);

JETRUN->RunTauJets(treeWriter,branchTauJet,sorted_jets,towers, TrackCentral,NTrackJet,EHADEEM,NCALO);

Both jets and tau-jets are using the same input array 'sorted_jets'.

To answer your questions:

• in Delphes 2, this double writing occurs for all taus;
• in Delphes 2, isolated electrons and photons are also reconstructed as jets.

In Delphes 3, there is a module called UniqueObjectFinder that addresses this double counting problem and root2lhco converter does not write taus as jets.

Would it be possible for you to switch to Delphes 3?

[Jory]

Yes I could switch to Delphes 3, but if I understand it correctly I should ignore the taus in all LHCO events from Delphes 2 since they are already written there as jets.

[Pavel Demin]

If you don't need taus for your analysis you can ignore them.

With the LHCO files generated with Delphes 2, you'll still have a similar double counting problem for electrons and photons. So, you'll need to filter out jets with low had/em fraction.