Running MadGraph on an NNPDF error set

Hi Kayla,
thanks for your message.
so
- if you are using LO madgraph, you need to first install syscalc (type "install SysCalc" in the MG command line), then edit the run_card and add
NNPDF30_lo_as_0130 1 = sys_pdf
-if you are using the NLO mode, just set
True = reweight_PDF
and
263001 = PDF_set_min
263100 = PDF_set_max

That should work (at least in a MG5_aMC standalone installation, I am not sure what happens inside ATHENA).

Let us know...

Bye,

Marco

Hi Marco,

Thank you for your answer. I'm not quite sure if this is what I want. If I understand correctly, SysCalc will only provide the weights for each event. However I want to actually generate events according to the error PDF (the first one in the set for 263000). Is this possible with MadGraph?

Thanks,
Kayla

Hi again Marco,

Here is some clarification for what I'm trying to do.

We have already been using the PDF reweighting method to study our signal acceptance systematics. This is for a mono-Z dark matter study. However we are seeing some problems with samples at larger mediator/dark matter masses, in some corners of x1, x2 space. Therefore we would like to generate samples using the error PDFs themselves in order to study these problems more thoroughly.

Cheers,
Kayla

Dear Kayla,
thanks for the explaination, now I see what you need to do.
So, setting e.g. lhaid = 263001 works with lhapdf5, not with lhapdf6. The reason is that the pdfsets.index has changed, while in lhapdf5 it contains all id's (also for error sets), in v6 it only contains the id for the central one. If you can edit the pdfsets.index of your installation, i suggest you to add this line
263001 NNPDF30_lo_as_0130 2
otherwise, we need to find another solution...

Let me know....
Cheers,

Marco

Hi Marco,

I'm currently getting some help on how to do this with ATHENA specifically, and it is following your advice for editing the pdfsets.index file.

I have a question about the syntax used in this file; in pdfsets.index I see the line '263000 NNPDF30_lo_as_0130 2'.

This is very similar to the line you give me to add, except for the 263001. The line that you suggest, is that actually pointing to the error PDF set? It is very similar to the line already there for 263000 and it looks like they both point to the same PDF (?). Also what does the '2' mean?

Hi,
On 20 Feb 2016, at 01:08, Kayla McLean <email address hidden> wrote:

> Question #285750 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/285750
>
> Kayla McLean posted a new comment:
> Hi Marco,
>
> I'm currently getting some help on how to do this with ATHENA
> specifically, and it is following your advice for editing the
> pdfsets.index file.
>
> I have a question about the syntax used in this file; in pdfsets.index I
> see the line '263000 NNPDF30_lo_as_0130 2'.
>
> This is very similar to the line you give me to add, except for the
> 263001. The line that you suggest, is that actually pointing to the
> error PDF set? It is very similar to the line already there for 263000
> and it looks like they both point to the same PDF (?). Also what does
> the '2' mean?
>
yes, but the error PDF set will be of the same set of the central one, so it should be correct this way.
For the 2 i don’t know, it may be the number of loop for the running of alphas…
Cheers,

Marco

> --
> You received this question notification because you are subscribed to
> the question.