MadGraph5_aMC@NLO

Process taking too long to shower

Asked by Sourav

Dear experts,
I am generating the following process at NLO :
define top = t t~
define bot = b b~
define w = w+ w-
define pb = p b b~
generate pb pb > top w [QCD] @1
add process pb pb > top w pb [QCD] @2
add process pb pb > top bot w [QCD] @3

I am using the model sm with no_b_mass restriction and using pythia8 for showering. For generating 1000 events, it is taking more than 8 hours (stopped the program after that) even with 20 cores. I also tried with just one process at a time, and still it is taking too long. Can some experts please help me to resolve the issue.

Question information

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

Hi,

At NLO, it is important to distinguish clearly four flavor computation (that is done with loop_sm model)
and five flavor computation (with loop_sm-no_b_mass).
using
define pb = p b b~
is a bad idea at NLO, since it mixes 4 and 5 flavour.

Now, the following process
generate p p > top w [QCD] @1
is problematic in five flavor since it include at the real level "p p > t t~" resonances.
So if you want to do that computation in a five flavor framework, you need to substract such resonance (which is possible via the MadSTR framework).

Additionally, the following process
add process p p > top bot w [QCD] @3
does only make sense in four flavor since you need to have the b massive to be allowed to single out in a NLO computation. So I'm really confuse about the type of computation you want to do.

Finally, the sum of the three processes does not make any sense:
process "3" is included within "2" why do you try to double count them.
For the double counting between "1" and "2", I guess that you do have FxFX (or UNLOPS?) activated? But again here I would be worry about your mix 4/5 flavor computation.

So given the above point, I guess that you are asking a question that does not really make sense at NLO accuracy and that you are unlucky that our code did not detect it via a missmatch of pole/...

Cheers,

Olivier

Can you help with this problem?

Provide an answer of your own, or ask Sourav for more information if necessary.

To post a message you must log in.