muon isolation problem

[gioar]

Hi,

I am trying to reproduce the recent search by Atlas of events with three leptons and missing energy. I have generated samples of signals and WZ and ttW/Z backgrounds with Madgraph and used Delphes3.0.6 for detector simulation using a delphes card designed for Atlas modifying the default b-tagging efficiency in order to match the values reported on the paper of the collaboration.

For each of the samples generated, i.e both signal and background, the number of events results at least a factor five less than the expectation. In order to evaluate the events I employ offline cuts on the delphes output which is a .lhco file.

I think that the problem is the muon isolation, i.e. maybe too many of them are rejected by the isolation cuts. Most of the muons result close to a jet. This should be normal since they can be produced by the decay of b and c quarks and I have to reject them. However I have found many muons which have exactly the same eta and phi and pt parameters of a jet. Is this normal?

I attach the delphes card I am using and one of my output lhco files.

Thanks,
Giorgio

[Michele Selvaggi]

Hello,

thanks for testing Delphes 3.0, and for your questions.

There are probably two separate issues, the muon isolation parameters are not really tuned. What you could do is to relax the muon isolation criteria, for instance you could reduce the isolation cone (line 366) in the configuration card. By the way how do electrons behave? Do you get the correct number?

However I have found many muons which have exactly the same eta and phi and pt parameters of a jet. Is this normal?

this is not normal. It means that these are jets formed out of single muons.

Could you please add to the configuration card the following line and retry your analysis?
add InputArray MuonIsolation/muons muons

between line 476 and 477 in the card you attached.

Also, for future iterations, it would be useful for us to have a small sample of your madgraph generated sample (before Delphes) in hep format.

Thanks,

Michele

Replying to gioar:

Hi,

I am trying to reproduce the recent search by Atlas of events with three leptons and missing energy. I have generated samples of signals and WZ and ttW/Z backgrounds with Madgraph and used Delphes3.0.6 for detector simulation using a delphes card designed for Atlas modifying the default b-tagging efficiency in order to match the values reported on the paper of the collaboration.

For each of the samples generated, i.e both signal and background, the number of events results at least a factor five less than the expectation. In order to evaluate the events I employ offline cuts on the delphes output which is a .lhco file.

I think that the problem is the muon isolation, i.e. maybe too many of them are rejected by the isolation cuts. Most of the muons result close to a jet. This should be normal since they can be produced by the decay of b and c quarks and I have to reject them. However I have found many muons which have exactly the same eta and phi and pt parameters of a jet. Is this normal?

I attach the delphes card I am using and one of my output lhco files.

Thanks,
Giorgio