Example3.C of Delphes example code

[Abhishek Iyer]

Dear Madam/Sir,
I have a question on line 185 of Example3.C. The link is below
https://cp3.irmp.ucl.ac.be/projects/delphes/browser/git/examples/Example3.C

While summing over the pT of the constituents of the jet, the momenta for the particle is added separately along with the Tower and track elements. My query is regarding the over counting of the jet constituent momenta as the particle may also get registered as a Track or Tower.

So while going over the constituents of the jet, is it necessary to check if the Jet constituent is a particle or not but just check if it is a track or tower. I want to extract all the jet constituent four vectors and use them for further clustering with a different algorithm. As a result I want to make sure, I am not missing out on any information.

thanks for your time
Abhishek

[Michele Selvaggi]

it is true that this code assumes the loop is occurring over particle-flow jets (the default Jet collection produced in CMS card).
In that case there is no double counting given that constituents are particle flow objects

[Abhishek Iyer]

Thanks.

What we observe is the following. We count the total number of constituents for the leading Jet.
Then we individually count the Particle, Track and Tower constituents. We find that the sum of Particle + Track +Tower does not always add up to the total number of constituents in the leading jet.
Since these three classes are mutually exclusive sets we believe they should match.

Are we missing something here?

thanks

Replying to mselvaggi:

it is true that this code assumes the loop is occurring over particle-flow jets (the default Jet collection produced in CMS card).
In that case there is no double counting given that constituents are particle flow objects

[Michele Selvaggi]

If you are looking at reconstructed jets there should be no constituent of type Particle in the constituents.
Why do you expect the sum of Particle + Track +Tower to add up to the total number of constituents in the leading jet? I don't understand ...

[Abhishek Iyer]

Thank you for your clarifications. We got a bit confused earlier but now it seems fine.
When using Example3.C we are having no issues in getting constituents of a jet in NO pileup scenarios.

We now introduce pileup in our events. To obtain the pileup subtracted jets, we are using the default
ATLAS Pileup card. We use the following lines,

add Branch TrackMerger/tracks Track Track
add Branch Calorimeter/eflowTracks EFlowTrack Track
add Branch Calorimeter/eflowPhotons EFlowPhoton Tower
add Branch Calorimeter/eflowNeutralHadrons EFlowNeutralHadron Tower .

Now, using the same code Example3.C, if we try to print the four momenta of the constituents and check whether it is a tower
or a track element we are encountering a segmentation error. Note that, we did not encounter this problem in no pileup scenario.
Are we missing something here. We are using Delphes 3.3.2.

thanks