Pythia 8 with aMC@NLO

Hi Harikrishnan,
let me first understand some details of what you did, point by point.

1) The sample you used was a Les Houches file with p p > w+ w- and no
decays of the w's. You shower it with Py8 and require HepMC (i.e. no
other user written analysis), right?

2)

.. sorry it was incomplete

2) what do you mean by Py8 standalone on amc@nlo generated events?

3) what is the difference between this point and point 1) ?

Thanks.
Cheers.
Paolo

ok i should have been more clear
1) generate p p > w w [QCD] ie. @NLO then i do madspin and then pythia8 with only .hep production no other analysis routine

2)take the lhe events after mad decay, and then run a modified form of main41.cc inside pythia8/examples which corresponds to reading an lhe file and showering it and printing out .hep. By standalone i mean i take the lhe file and shower it outside amc@nlo by myself.

3)same as 2 but without maddecay and standalone pythia does the decay.

Harikrishnan

Hi Harikrishnan,
ok now I understand better, thanks.

Since the problem is mainly at the decayed level, could you first see what happens for
point 1) if you do not decay the W's, and use pythia8 inside MG5_aMC@NLO?
You can even do a simpler thing: try p p > W+, then p p > W+ + MadSpin, and then p p > e+ ve
(or whatever decay product you would put in MadSpin), and see what is the fraction
of unshowered events in the hepMC file.
If everything is fine for this processes, then it could be something related to the presence
of at least two bosons, and I would ask if you could do the same for p p > w+ w-, p p > w+ w-
+ MadSpin, p p > e+ ve mu- vm~. (Of course you can generate these processes with [real=QCD]
instead of [QCD], since these are only tests, not physical runs, and you may not want to waste
time in computing loops).

Could you also try running with version 2.1.2_beta2? There have been some fixes here
and there linked to the lepton masses used in the shower, maybe this can help.

For points 2) and 3) I don't really know, it some pythia8 internal thing, so it would be
convenient for you to ask the pythia8 authors directly.

Cheers.
Paolo

Hi,
I showered just p p > w+ without madspin and still there is a huge percentage of unshowered events with Pythia8!
i used generate p p > w+ [QCD]...
Am still trying to get beta2 to work but am having some issues interfacing it with Pythia8 but I ll troubleshoot that later!
Sincerely
Harikrishnan

Hi Harikrishnan,
by unshowered events you mean events that are discarded by the shower
or events that in the HepMC file have exactly the same kinematics as they
have in the LH file?
How do you check they are unshowered?
I'm asking since when we use the analyses for top drawer, we never notice
this behaviour, and your answer could help understanding where the problem
is.
I have another request: are you asking for scale and/or PDF variations in the
run_card? Could you try without scale/PDF variations and see if you get a large
fraction of unshowered events?
Thanks.
Cheers.
Paolo

Hi Paolo,
I do not request ion scale/pdf variations.
By unshowered I mean, in a W+ process the pt_W remains zero even after showering.
I convert the hep file produced into a root file using delphes and then analyze it using the root software. I find that the unshowered events have ~ 10-12 particles(including initial-intermediate-final state) while the showered events have ~ 100 -250 particles. Both the lack of any hadrons or additional particles and pt_W remaining 0 made me decipher that those events were not being showered and just converted directly into a hep file by Pythia8

Hari

Hi Harikrishnan,
ok, your explanation is very clear, thanks.
I’ve asked some advise to one of the Pythia8 authors, since
I have never met this feature using the built-in analyses we
have for top-drawer.
It may be some problem in the output to hepevt or to HepMC, I
don’t know.
I’ll try to get back to you as soon as I know something more.
Cheers.
Paolo

On 03 Jul 2014, at 08:22, Harikrishnan Ramani <email address hidden> wrote:

> Question #250764 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/250764
>
> Status: Needs information => Open
>
> Harikrishnan Ramani gave more information on the question:
> Hi Paolo,
> I do not request ion scale/pdf variations.
> By unshowered I mean, in a W+ process the pt_W remains zero even after showering.
> I convert the hep file produced into a root file using delphes and then analyze it using the root software. I find that the unshowered events have ~ 10-12 particles(including initial-intermediate-final state) while the showered events have ~ 100 -250 particles. Both the lack of any hadrons or additional particles and pt_W remaining 0 made me decipher that those events were not being showered and just converted directly into a hep file by Pythia8
>
> Hari
>
> --
> You received this question notification because you are a direct
> subscriber of the question.

Hi Harikrishnan,

in order to try understanding this problem better, it would be
great if you could do a couple of things.
First, shower your sample and print 100 events on the log file
(by selecting ‘maxprint = 100’ in shower_card.dat).
Then, attach the log file (MCatNLO/RUN_PYTHIA8_**/mcatnlo_run.log),
to this thread.

Stefan Prestel, the Pythia8 author I asked to, thinks this may
be due to hadronisation somehow. Could you also shower once without
hadronisation ('hadronize = F' in shower_card.dat: check in the
corresponding mcatnlo_run.log that this indeed turns hadronisation
off) and see if the percentage of unshowered events goes down?

Cheers.
Paolo

On 04 Jul 2014, at 15:46, Paolo Torrielli <email address hidden> wrote:

> Hi Harikrishnan,
> ok, your explanation is very clear, thanks.
> I’ve asked some advise to one of the Pythia8 authors, since
> I have never met this feature using the built-in analyses we
> have for top-drawer.
> It may be some problem in the output to hepevt or to HepMC, I
> don’t know.
> I’ll try to get back to you as soon as I know something more.
> Cheers.
> Paolo
>
>
> On 03 Jul 2014, at 08:22, Harikrishnan Ramani <email address hidden> wrote:
>
>> Question #250764 on MadGraph5_aMC@NLO changed:
>> https://answers.launchpad.net/mg5amcnlo/+question/250764
>>
>> Status: Needs information => Open
>>
>> Harikrishnan Ramani gave more information on the question:
>> Hi Paolo,
>> I do not request ion scale/pdf variations.
>> By unshowered I mean, in a W+ process the pt_W remains zero even after showering.
>> I convert the hep file produced into a root file using delphes and then analyze it using the root software. I find that the unshowered events have ~ 10-12 particles(including initial-intermediate-final state) while the showered events have ~ 100 -250 particles. Both the lack of any hadrons or additional particles and pt_W remaining 0 made me decipher that those events were not being showered and just converted directly into a hep file by Pythia8
>>
>> Hari
>>
>> --
>> You received this question notification because you are a direct
>> subscriber of the question.
>