Running with pileup randomly give spurious results

[Prashant Saraswat]

I've been running Delphes with pileup on some several samples such as WW to dileptons. About 20-50% of the time that I run in this way, the output file gives absolutely no events with leptons at all; in fact the events look like pure pileup with no trace of the hard event, just a lot of low-pT jets. The rest of the time the output seems completely fine (to the point that I successfully validate against some LHC data). The spurious output seems to occur entirely randomly; i.e. I can run Delphes multiple times on the exact same sample with the same card, sometimes getting good results and sometimes getting the strange pileup-only output. I've tried this with both the .pileup file available from the Delphes site as well as one I generated myself using Pythia8; it seems so far that using the latter gives a higher chance of the spurious output, but it can occur with both. I can provide the event files, cards etc. I'm using if necessary.

[Alexandre Mertens]

Hello,

Yes, can you provide a detector card and an small event file, please?

Cheers,

Alexandre

[Pavel Demin]

Could you, please provide a small (10-100 events) Delphes input file, your cards, and a small Delphes output file that contains some problematic events?

[Prashant Saraswat]

Includes a PYTHIA .hep file and the output from two attempts at running Delphes with pileup-- one run giving reasonable results and one giving apparent pure pileup events.

[Prashant Saraswat]

Hi Alexandre, Pavel,

I've attached a zip with the relevant files-- I ran Delphes twice on the Pythia HEP file using the given Delphes card and the .pileup file provided on this website. One run gave me "good" output where most events had one or more leptons as expected for WW --> ll, and one run gave me the "failed" output with just a lot of soft jets.

Thanks for looking into this,
Prashant

[Prashant Saraswat]

Hi,

Any word on whether you guys can reproduce this bug? It would be useful to know even if this is just an issue on my system.

Thanks,
Prashant

[Michele Selvaggi]

Hi Prashant,

sorry for the late reply.
We can reproduce the bug, and we are investigating, we will let you know as soon as we know better,

Thanks for spotting this,

Michele

[Pavel Demin]

Hi Prashant,

Sorry for answering you so late.

Contrary to Michele, I can't reproduce the problem with Delphes 3.0.12.

What Delphes version are you using?

I've run Delphes 100 times using the following commands:

for i in `seq 0 99`
do
  rm test.root
  ./DelphesSTDHEP delphes_card_PileUp_PS.tcl test.root ww_ll_test_pythia_events.hep
  root -l -b -q counter.C'("test.root")' >> counter.txt
done

and I've never obtained less than 8 leptons (electrons + muons) per .root file.

The counter.C macro is in the attachments.

My .pileup file is from WorkBook/PileUp (​http://cp3.irmp.ucl.ac.be/downloads/MinBias.pileup.gz).

Regards,

Pavel

[Michele Selvaggi]

Hi Prashant,

there was apparently a bug which appears only on some systems.
Pavel has just released version 3.1.2 could you try with that please?

Michele