MadGraph5_aMC@NLO

HOw MadGraph assigns the order (QED/QCD/...) automatically?

Asked by Olivier Mattelaer

(receive from email from Josh McFayden)

We were wondering if you could clarify a point for us regarding how different processes that contribute to the same final state are selected in madgraph.

For example when generating p p > W b b~ j j we do not see any diagrams containing t t~ even though both the t t~ and W+jets processes can have this same final state.

Given that no restrictions are applied to the numbers of QCD and QED couplings how does madgraph choose which diagrams to generate which to omit?

Thanks in advance,

Josh and Kerim.

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
Olivier Mattelaer (olivier-mattelaer) said : 		#1

Hi Josh, Kerim,

Let me first explain for the Standard Model / MSSM and after that I will explain how this is extended to any BSM theories.

So the rule for the SM/MSSM is in fact very simple, we take the order having the LOWEST QED order resulting to a non zero contribution. The motivation for such a rule (besides the fact that, in general, the QED part is sub-dominant) is for the matching procedure (for multijet production) where it's important that the multijet part is ONLY QCD.

In order to generalize the idea to have only the main contribution by default.
We use a functionality of the UFO assigning to each order a weight
(you can see those in MG5 by doing
display couplings_order)
mg5>display coupling_order
 QCD : weight = 1
 QED : weight = 2

Then we try to guess what is the minimal weight for a given process
(WEIGHT= NB_QCD *1 + NB_QED *2)
For example:
mg5>generate p p > W+ b b~ j j
INFO: Checking for minimal orders which gives processes.
INFO: Please specify coupling orders to bypass this step.
INFO: Trying coupling order WEIGHTED=6
Most of the time we guess it correctly, but sometimes we will need to increase the value of the weight by one or two in order to have some diagrams.

By using those weights, the model builder can specify an ordering between the couplings.
Of course such default is not always the best one (you can kill your signal in such case) but this is very useful in general.

Don't hesitate to reply to this if one point is not clear.

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#2

Passing this question in FAQ (This is the first time I do thqt, so I hope it will work)
FAQ #1942: “HOw MadGraph assigns the order (QED/QCD/...) automatically?”.