increasing xqcut parameter

Hi,

Do you speak about the cross-section after of before matching?
Did you set icckw to 1? the pdf reweighing to T?

Cheers,

Olivier

On Dec 10, 2013, at 6:56 PM, Andrew Levin <email address hidden> wrote:

> New question #240589 on MadGraph5:
> https://answers.launchpad.net/madgraph5/+question/240589
>
> Why does increasing the xqcut parameter (from 0 to 20) cause the cross section of this
>
> define l+ = e+ mu+ ta+
> define l- = e- mu- ta-
> generate p p > w+ w+ p p QED=4 QCD=99, w+ > l+ vl
> add process p p > w- w- p p QED=4 QCD=99, w- > l- vl~
>
> to increase?
>
> This is a minimum cut, so I would expect the cross section to decrease.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5.

icckw is set to 0 and pdfwgt is set to T

ok,

so I think I have found the reason.
What is the value of ptj and the value of mjj?
and what is the value of auto_ptj_mjj

if auto_ptj_mjj is on True and xqcut >0
then the value of ptj and mjj are automatically modified, which can therefore increase the cross section if the new value are lower than the old one.

Cheers,

Olivier

ptj is set to 10

you mean mmjj not mjj right?

mmjj is set to 100
auto_ptj_mjj is set to T

It says in the madgraph runcard this

# (turn off for VBF and single top processes)

I don't really understand why this should be turned off for only VBF and single top processes.

Do you have a recommendation for if we should turn this on or off for our process?

Should clearly be on OFF for your process as this is stated to doing it.
For more deep explaining why I refer to Simon.

Cheers,

Olivier

Hi,

I have recalculated the cross section with auto_ptj_mjj set to F, but actually this makes the cross section even higher than it was with auto_ptj_mjj set to T.

Andrew

Here are the cross sections I get:

auto_ptj_mjj set to T, xqcut set to 0 ----> 0.01906 ± 1.5e-05

auto_ptj_mjj set to T, xqcut set to 20 ----> 0.02301 ± 1.7e-05

auto_ptj_mjj set to F, xqcut set to 20 ----> 0.02628 ± 2e-05

Here is the run_card.dat I am using (for the last case):

http://t3serv001.mit.edu/~anlevin/tmp/run_card.dat

Hi Andrew and Olivier,

Poking this old thread out of curiosity :-)
Would the must-choice of auto_ptj_mjj = F for VBF and single top understood? From Simon?
Also I have a sideline question:
What auto_ptj_mjj would we set if we do the same final state with VVjj-QCD instead of VVjj-EWK?
And what would the auto_ptj_mjj value would be if we do them both together at a time or for VVjj-Interference term only by specifying the interference at ME-square level.
Many thanks.

best,
Shu

Dear Shu,

> Would the must-choice of auto_ptj_mjj = F for VBF and single top understood? From Simon?

The reason is actually is that you have jet not related to QCD interactions. Therefore those jets are not include in the matching/merging.
Setting auto_ptj_mjj=T means that the cut on those jet will not be the same when you change the value of xqcut and therefore you can not compare the various number.

>What auto_ptj_mjj would we set if we do the same final state with VVjj-QCD instead of VVjj-EWK?

in that case auto_ptj_mjj=T makes a lot of sense. Now auto_ptj_mjj=F is not wrong (at least when ptj is smaller than xqcut)

>And what would the auto_ptj_mjj value would be if we do them both together at a time or for VVjj-Interference term only by specifying the interference at ME-square level.

In that case, auto_ptj_mjj=F should be used. If you want to use ptj bigger than xqcut, then I will think more in details in that case to see if this makes sense or not.

Cheers,

Olivier

Thanks a lot, Olivier.
So in principle, if we don't use xqcut>0., we are safe towards auto_ptj_mjj choice regardlessly I think?
Btw, would xqcut=0. be ok with VVjj-QCD, EWK and Inteference? Many thanks.

best,
Shu

Hi Shu,

So you run with ickkw=0, xqcut=0 then in that case “auto_ptj_mjj” is irrelevant for the computation.

Cheers,

Olivier
> On Nov 24, 2016, at 17:27, Shu Li <email address hidden> wrote:
>
> Question #240589 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/240589
>
> Shu Li posted a new comment:
> Thanks a lot, Olivier.
> So in principle, if we don't use xqcut>0., we are safe towards auto_ptj_mjj choice regardlessly I think?
> Btw, would xqcut=0. be ok with VVjj-QCD, EWK and Inteference? Many thanks.
>
> best,
> Shu
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.