EFlow and leptons

[Jan Hajer]

Hi,

For my analysis, I am using the EFlow variables as an input for jet clustering. Additionally, I want to know about the leptons in each event. Usually the leptons I am interested in, are not recorded because they are not isolated enough. Therefore, I am reducing the Isolation radius in the delphes card.
Now I am afraid that I am introducing a double counting of leptons.
Once as leptons (in the lepton branch of delphes after reducing the isolation radius) and a second time as part of the jets.
Is this true? And is there a straight forward way to remove the leptons from the EFlow branches?

Best,
Jan

[Michele Selvaggi]

Hi Jan,

in principle there should be no-double counting, since both the Lepton and the jet are fed to the UniqueObjectFinder, which should precisely take care of vetoing overlapping objects. This said, what will happen in your example is that since jets have priority over isolated electrons, photons and muons (isolated in the sense that they passed the isolation criteria, which in your case is very loose) in the UniqueObjectFinder, all those leptons that are present in jets will be vetoed. If you want to give priority to leptons over jets, you can revert the order that is used for supplying those collections to the UniqueObjectFinder.

Hope this helps,
Michele

[Jan Hajer]

Hi Michele,

thanks for the explanation. I am working with highly boosted leptonic top jets and I have already noticed that I can keep either the lepton or the bottom from the top decay depending on the looseness of the isolation criteria. But it is my goal to work with both, therefore I was using the eflow branches instead of the jet branch. Now I am afraid that the lepton is not deleted from the eflow branches. Hence, I think I am introducing the double counting by using the lepton and eflow branches together. Is it possible to apply something like the UniqueObjectFinder also to the eflow branch?

Best,
Jan

[Michele Selvaggi]

Hi Jan,

you can then simply store in the output the leptons and jets before they passed the UniqueObjectFinder. This way you will have double-counting, but you can then simply remove manually jets that are made of the prompt-lepton from the W decay by looping over the jets constituents to identify that fake-jet (see examples/Example3.C for an example on how to access jet constituents).

Cheers,
Michele

[Jan Hajer]

Hi Michele,

i am familiar with example 3. I will try this technique.

Thanks,
Jan