MadGraph5_aMC@NLO

Formation of subprocesses

Asked by Jure Drobnak

hey,

i stumbled upon an issue that might not be too important, but is bothering me anyway.
first i run:
./bin/mg5
generate p p > j j
output dijet

looking in the dijet/SubProcesses/ folder i find only 5 sub folders
P0_gg_gg
P0_gg_qq
P0_gq_gq
P0_qq_gg
P0_gg_qq
which is a very nice and efficient grouping of diagrams from what i understand.

now my question is, how come i do not get the same nice grouping if take Feynrules program,
import the simple SM (sm.fr) included therein and export it as a SM_UFO then do:

./bin/mg5
import model SM_UFO
generate p p > j j
output dijet2

now dijet2 folder contains about 40 subfolders of subprocesses.
my question is why the nice grouping of the diagrams was not performed in the same way
as when using the default mg5 SM?
is the the UFO issue or is SM.fr written is such a way that this grouping is not possible?

thanks for the reply,
and all the best,
Jure

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
Last query:
Last reply:
Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said : 		#1

Hi Jure,

the SM_UFO is not the same model as the default MG5 model.
The differences are off course quite small (for example the parametrization of the CKM matrix)
but some of them are quite critical for the grouping.

The grouping of the SubProcesses works only if the particles have the same mass.
Since they are all free parameter in the SM_UFO (is it?), the merging is not possible.

In the model that we use by default (import model sm):
- the mass of the u/d/s are set to zero (restriction performed at FR level)
- the mass of the c is set to zero via a MG5 restriction card. (./models/sm/restrict_default.dat)
(so if you do import model sm-full, the c quark will have a non zero mass)

So since all the quark masses are the same (and here equal to zero), we can merge the resultant subprocesses

Cheers,

Olivier

Revision history for this message
Jure Drobnak (jure-drobnak) said : 		#2

Thanks Olivier Mattelaer, that solved my question.

Revision history for this message
Jure Drobnak (jure-drobnak) said : 		#3

Thanks a lot for the prompt reply!
Keep up the good work!