Missing photon in delphes root file

[Daniele]

Dear Delphes author,
I have a problem quite similar to the one of
https://cp3.irmp.ucl.ac.be/projects/delphes/ticket/323
but there is still something not clear to me.

I am generating, via MG, a test process p p > z h, z > e+ e-, h > a a.
I am passing then the output file to pythia and then to Delphes.
At parton level I am requiring a photon pT of at least 40 GeV and 0.4 of isolation between photon and leptons.
Up to pythia level I have all my photons in the relative root files, but when I pass the pythia output to delphes and analyse the delphes root file I find that basically all of my photons are disappeared and that the remaining photon have at the most a pT of 20 GeV, no matter the number of events I generate.
I understood from the other topic that this problem could be related to an isolation criteria, but I am a bit confused that none of the photon that I generate at parton level, with pT>40 GeV, manage to pass this criteria. Is there something I am missing?

I am attaching here, as an example with 100 of events, the root file that I get from MadGraph, Pythia and Delphes, as well as the delphes card I am using.

Thanks for any help you can give me

Daniele

[Michele Selvaggi]

Hi Daniele,

sorry for the late reply.
It looks indeed like photons fail to pass the isolation criteria, since they are present in the EFlowPhoton collection.

Could you please attach the *.hep file produced by !MG5 please?

Thanks
Michele

[Daniele]

Hi Michele,
No worries.
For some reasons I have erased the old files, so I produced new ones.
I am attaching here the files corresponding to the ones of my previous post, together with the *hep files.
I manage to obtain the photons in my root file by commenting line 413 of the delphes_card

add InputArray UniqueObjectFinder/photons

but of course I understand that this is not the proper way to solve the problem, was more for kind of a test.

Daniele

[Michele Selvaggi]

Hi Daniele,

I just ran Delphes-3.1.2 standalone on your *.hep file and your card and everything looks fine, I have photons in the Photon collection.

Did you try with a fresh Delphes version?

You should make sure you properly set the ROOT environment variable first.
And then you can follow this quick tutorial to run Delphes standalone on your hep file.

https://cp3.irmp.ucl.ac.be/projects/delphes/wiki/WorkBook/QuickTour

Cheers,
Michele

[Daniele]

Dear Michele,
Thanks for the reply.
I have just download the latest version of Delphes from the website and, compile it

make -j 4

and run it not in MadGraph by

./DelphesSTDHEP delphes_card.tcl output.root tag_1_pythia_events.hep

but if I open the root file and I go, for example, in the Photon PT branch, I can see only two photons...
What you mean by "make sure you properly set the ROOT environment variable first"?
I think on my computer ROOT works fine, I never had any problem.
Moreover if I comment the following line in the card (line 413)

add InputArray UniqueObjectFinder/photons

I get a nice pT spectrum for the photon.
Do you think that my problem could be related anyway with the ROOT environment anyway? It looks strange to me, since I would say that I would get the same results without photons with both the delphes_card.tcl (with and without that line commented...)
Do you have any idea?

[Michele Selvaggi]

Hi Daniele,

can you try to add the following lines in tree writer section of the card, rerun Delphes?

add Branch PhotonEfficiency/photons PhotonEff Photon
add Branch PhotonIsolation/photons PhotonIso Photon

This way you will have three photon collections:

PhotonEff : after efficiency cuts
PhotonIso : after efficiency cuts + isolation
Photon : after efficiency cuts + isolation + overlap with jets removal

and please tell me how do the PhotonEff and PhotonIso collections look like.

This said,I fail to see how you can have photons in your output if you comment out the line you mentionned.

Thanks,
Michele

[Daniele]

Hi Michele,

This is kind of strange.
I took the delphes_card without the comment of the line I mentioned and just add at the end the line you told me, and the photon are present in all the three collections (as you can see from output2.root in attach). If I don't add those line I obtain output.root (still in attach) where there are no photons. Does this makes sense?

Regarding commenting the lines I thought that, commenting that line, I was just relaxing to the maximum the criteria of identifing a photon, so as all the photons were identified (extrapolating, maybe wrongly, the discussion in https://cp3.irmp.ucl.ac.be/projects/delphes/ticket/323).

This said I am a bit puzzled on what I have just found...

Thanks

Daniele

[Michele Selvaggi]

Hi Daniele,

what I meant is that you should not even have the Photon collection in the output if you comment that line, as in the treewriter you don't modify collections, you simply decide what you write in the output tree.

I fail to see which collection you were looking at before (when you commented that line), since the Photon collection simply could not be present in the output. Maybe you were looking at EFlowPhotons?

Starting from what you have now, you are telling me that if you comment back:

add Branch PhotonEfficiency/photons PhotonEff Photon
add Branch PhotonIsolation/photons PhotonIso Photon

the "Photon" collection gets back to being empty, while with those two other collections being in the output it is not??

If that is the case then I am lost I must say..

Michele

[Daniele]

Hi,
But the line I commented

add InputArray UniqueObjectFinder/photons

is not in the tree writer part, but in the Scalar HT merger part (line 413 of the card I have attached). So I don't see why the photon collection shouldn't be present in the output if I comment it.
I will agree with you if I would comment line 554, in the tree writer

add Branch UniqueObjectFinder/photons Photon Photon

and in fact commenting this line I don't have the photon collection in the root file.

This said I made some other tests with different cards, and I try to explain the situation, I have partly to correct my statements.

If I use the delphes_card_ATLAS.tcl which I find in the example directories without modifing it, everything works fine. I have my photons in the photon collection and I see the correct pT distribution.

If I use the delphes_card_CMS.tcl which I find in the example directories without modifing it, I don't have photons in the photon collection ( I find just two photons). Are these two cards drammatically different in this?
Moreover, and this I agree with you is crazy, if I add those two lines in the tree writer part, in the photon collection the photon appears and I see the nice pT distribution.
I know is absurd and it also puzzle me, but this is what I find...