Algorithm when pdfwgt == True

Asked by Junho Choi on 2018-07-07

Hi,

I study DrellYan ee process,

p p > e+ e-
p p > e+ e- j
p p > e+ e- j j

.

When I set

True = pdfwgt

in runcard,

ME level xsec values are different between v260 & v261 MG5.

v260 : 3.286e+03 +/- 7.039e+00

v261 : 3.875e+03 +/- 9.902e+00

As far as I know, if the "pdfwgt == True" , ISR is considered to set (x,Q^2). I mean if there's ISR, x, Q scale is changed and the change is considered when pdfwgt == True.

I guess algorithm related to pdfwgt = True is changed between v260 & v261. Is it true?

If you have any unclear thing in my description, please let me know.

Best regards,

Junho Choi

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
2018-07-12
Last query:
2018-07-12
Last reply:
2018-07-11

Hi,

If you look at the UpdateNotes of 2.6.1 you will see the following:

> OM: Fix a bug in matching/merging forbiding the pdf reweighting for some processes (since 2.4.0)

So this difference is indeed expected and you should discard result from 2.6.0.

Cheers,

Olivier

> On 7 Jul 2018, at 08:13, Junho Choi <email address hidden> wrote:
>
> New question #670715 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/670715
>
> Hi,
>
> I study DrellYan ee process,
>
> p p > e+ e-
> p p > e+ e- j
> p p > e+ e- j j
>
> .
>
> When I set
>
> True = pdfwgt
>
> in runcard,
>
> ME level xsec values are different between v260 & v261 MG5.
>
> v260 : 3.286e+03 +/- 7.039e+00
>
> v261 : 3.875e+03 +/- 9.902e+00
>
>
>
>
> As far as I know, if the "pdfwgt == True" , ISR is considered to set (x,Q^2). I mean if there's ISR, x, Q scale is changed and the change is considered when pdfwgt == True.
>
>
> I guess algorithm related to pdfwgt = True is changed between v260 & v261. Is it true?
>
>
> If you have any unclear thing in my description, please let me know.
>
>
> Best regards,
>
> Junho Choi
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Junho Choi (junhochoi) said : #2

Thanks Olivier Mattelaer, that solved my question.

Junho Choi (junhochoi) said : #3

Thanks!

I should have read the node. Sorry.

Best regards,

Junho

Qiang (qliphy) said : #4

Hi Olivier,

Sorry but I have other related questions on this point:

1. In 2.6.1, it seems there is no "pdfwgt" option in the run_card.dat. If we add by hand by accident "True = pdfwgt", will it overwrite the setting and play any role?

2. In CMS, we checked and found the difference at ME level before shower is large at 10%, as Junho mentioned above. However, after the merging, the difference is much smaller at 1% level. So we want to get comment from you whether pdfwgt really play curcial role here, or, is the related bug a big effect?

Thanks a lot!

Best,
Qiang

Hi,

> 1. In 2.6.1, it seems there is no "pdfwgt" option in the
> run_card.dat. If we add by hand by accident "True = pdfwgt", will it
> overwrite the setting and play any role?

I do not remember why I set this as an hidden parameter but ok.
The default value for this parameter is "True" so setting this line or not will not change the behavior.
If you set it to False, then it will change the behaviour and yes it will be taken into account.

Note that you can always write all the hidden parameter of the run_card via the command: "update to_full"

> 2. In CMS, we checked and found the difference at ME level before
> shower is large at 10%, as Junho mentioned above. However, after the
> merging, the difference is much smaller at 1% level. So we want to get
> comment from you whether pdfwgt really play curcial role here, or, is
> the related bug a big effect?

The fact that you have a 1% difference after merging is actually a good news.
My understanding for such difference is that most of the event with a large impact for the PDf reweighting
are going to be veto by MLM merging, this should make sense since this indicate that you have a quite high scale difference between the radiation and the hard-scale (and therefore a high probability to be veto-ed)

Cheers,

Olivier

> On 27 Jul 2018, at 15:02, Qiang <email address hidden> wrote:
>
> Question #670715 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/670715
>
> Qiang posted a new comment:
> Hi Olivier,
>
> Sorry but I have other related questions on this point:
>
> 1. In 2.6.1, it seems there is no "pdfwgt" option in the
> run_card.dat. If we add by hand by accident "True = pdfwgt", will it
> overwrite the setting and play any role?
>
> 2. In CMS, we checked and found the difference at ME level before
> shower is large at 10%, as Junho mentioned above. However, after the
> merging, the difference is much smaller at 1% level. So we want to get
> comment from you whether pdfwgt really play curcial role here, or, is
> the related bug a big effect?
>
> Thanks a lot!
>
> Best,
> Qiang
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Qiang (qliphy) said : #6

Hi Olivier,

Thanks for the answers! I have some further questions:

(1)
> If you set it to False, then it will change the behaviour and yes it will be taken into account.

So the default in 261 for pdfwgt is "True", do you think "False" is also a good option?

(2)
And I am now a bit confused about the following in UpdateNotes of 2.6.1:
> OM: Fix a bug in matching/merging forbiding the pdf reweighting for some processes (since 2.4.0)

Is this a bug only for pdfwgt=True? Do you think pdfwgt=false is still ok for 2.4.0?

(3) BTW: as far as I understand, 2.2.2 is not affected by the bug, right?

Best,
Qiang

(1) the difference should technically be nlo. So it is fine to not include it as long as your djr are smooth enough

(2) yes this is a bug only when that parameter is true. The generation with false was working as expected

(3) I would say so

Cheers

Olivier

Get Outlook for iOS<https://aka.ms/o0ukef>
________________________________
From: <email address hidden> <email address hidden> on behalf of Qiang <email address hidden>
Sent: Saturday, July 28, 2018 6:09:35 AM
To: Olivier Mattelaer
Subject: Re: [Question #670715]: Algorithm when pdfwgt == True

Question #670715 on MadGraph5_aMC@NLO changed:
https://answers.launchpad.net/mg5amcnlo/+question/670715

Qiang posted a new comment:
Hi Olivier,

Thanks for the answers! I have some further questions:

(1)
> If you set it to False, then it will change the behaviour and yes it will be taken into account.

So the default in 261 for pdfwgt is "True", do you think "False" is
also a good option?

(2)
And I am now a bit confused about the following in UpdateNotes of 2.6.1:
> OM: Fix a bug in matching/merging forbiding the pdf reweighting for some processes (since 2.4.0)

Is this a bug only for pdfwgt=True? Do you think pdfwgt=false is still
ok for 2.4.0?

(3) BTW: as far as I understand, 2.2.2 is not affected by the bug,
right?

Best,
Qiang

--
You received this question notification because you are an answer
contact for MadGraph5_aMC@NLO.