how to improve efficiency for large Ht multijet events?

Hi,

Is your 74 events, the events generated before matching or those after matching?

Cheers,

Olivier

On May 22, 2014, at 8:56 PM, Rakhi Mahbubani <email address hidden> wrote:

> New question #249120 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/249120
>
> Dear MG/ME team,
>
> I am trying to generate matched 4-jet events at a 100 TeV collider, and am using Ht bins (Htmin, Htmax) for the leading 4 jets to increase statistics at high Ht. I've set xqcut=qcut = Htmin/3 for pt-ordered showers with shower Kt = true. Starting Htmin = 2 TeV the generation statistics get rather poor (74 events generated for 50000 requested). Is there a good way to improve the efficiency of the generation, or is my only option to generate more events?
>
> Cheers,
>
> Rakhi
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

I get 77 events before matching, and 74 after. The log for the first gg -> gg process for examples states that none of the events pass the cuts.

I'm now really confused because when I use the same Ht cuts but turn off matching (ickkw=0, xqcut=0) and turn on minimal pt, eta and drjj cuts (ptmin=30 GeV, etajmax=5, drjjmin=0.3) I get huge cross sections, 2.8E4 pb, as compared with 0.004 pb in the matched example. It's almost like the xqcut I'm choosing is too large, and I'm cutting out all my phase space...

okay now I realise I was being a complete idiot - I was choosing xqcut = 660 (Htmin/3) for *4* jets with 2000 GeV < Ht(4jet) < 3000 GeV...if each jet has a minimum pt of 660 then I'm basically cutting out almost my entire Ht window!! I'm guessing I would need to choose xqcut < Htmin/4, say 450, to be on the safe side: would you agree?

Cheers,

Rakhi

> I'm guessing I would need to choose xqcut < Htmin/4,
> say 450, to be on the safe side: would you agree?

Yes I agree, but I’m not sure that using such large value for xqcut makes sense at all since this means that
your matching scale will be very very high and I doubt that you will have nice matching curve. But I let you check/validate that part.

Cheers,

Olivier

On May 27, 2014, at 6:31 AM, Rakhi Mahbubani <email address hidden> wrote:

> Question #249120 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/249120
>
> Status: Answered => Open
>
> Rakhi Mahbubani is still having a problem:
> Hi Olivier,
>
> I get 77 events before matching, and 74 after. The log for the first gg
> -> gg process for examples states that none of the events pass the cuts.
>
> I'm now really confused because when I use the same Ht cuts but turn off
> matching (ickkw=0, xqcut=0) and turn on minimal pt, eta and drjj cuts
> (ptmin=30 GeV, etajmax=5, drjjmin=0.3) I get huge cross sections, 2.8E4
> pb, as compared with 0.004 pb in the matched example. It's almost like
> the xqcut I'm choosing is too large, and I'm cutting out all my phase
> space...
>
> okay now I realise I was being a complete idiot - I was choosing xqcut =
> 660 (Htmin/3) for *4* jets with 2000 GeV < Ht(4jet) < 3000 GeV...if each
> jet has a minimum pt of 660 then I'm basically cutting out almost my
> entire Ht window!! I'm guessing I would need to choose xqcut < Htmin/4,
> say 450, to be on the safe side: would you agree?
>
> Cheers,
>
> Rakhi
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.