Specifying decays of h3 for the loop induced process (g g > h3 -> z h2)

The only method available in this context is the spinmode=none of madspin.
Which does the decay without offshell effect (so only ok if the width of h2 is small enough) and without spin correlation (which should be ok if h2 is indeed a scalar)

Cheers,

Olivier

> On 19 Jul 2023, at 10:30, Zhiyuan <email address hidden> wrote:
>
> New question #707328 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/707328
>
> Dear MadGraph experts,
>
> I have a similar question to https://answers.launchpad.net/mg5amcnlo/+question/304480 but slightly different process:
>
> import model 2HDMtII_NLO
> generate p p > h3 > h2 z [QCD]
>
> Where the h3 is produced through a quark triangle.
>
> I would like to have the h2 decay to two SM higgs (h1), and decay the higgs and Z boson with Pythia, therefore I tried using: generate p p > h3 > h2 z , h2 > h1 h1[QCD].
> But this returns:
> "Error detected in sub-command generate p p > h3 > h2 z, h2 > h1 h1 [QCD]
> ...
> str : Decay processes cannot be pertubed"
>
> I wonder if there is a way to make it work?
>
> Best regards,
> Jordan
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Oliver,
Thanks a lot for the prompt answer.

In my madspind card I now have:

set ms_dir ./madspingrid

#initialization parameters
set max_weight_ps_point 400 # number of PS to estimate the maximum for each event
set max_running_process 1
# specify the decay for the final state particles
set spinmode none
decay h2 > h1 h1
decay z > l+ l-
# running the actual code
launch
#

Using this madspin card, I tested with and without the "h2 > h1 h1" line in the proc card. The former breaks the gridpack generation code with the same erro as posted. For the later case it 'works' (although with a warning CRITICAL: Branching ratio larger than one for 35 ).
I checked the particles' ID in the unweighted_events.lhe.gz file, and I only found the h3, h2, gluons and z, without finding the h1. I wonder if I have done it correctly making the h2 decay to h1?

Best,
Jordan

Hi,

This does not make sense... I guess that I do not understand correctly what you report.

Ff the error without "decay h2 > h1 h1"
is "Decay processes cannot be pertubed"
Then removing "decay h2 > h1 h1" or not should both lead to that same error.
(since it means that Madspin does not handle your syntax correctly.

Are you sure that you did not ask for "decay h2 > h1 h1 [QCD]" in that case the above error would make sense.

Obviously, if you do not ask for h2 to decay, then you will not find it in the lhef file.

But another point that I do not understand, if you do not ask h2 to decay, then madspin will not try to decay it, and will not generate events for such decay, in that case, it will not compute the partial-width associated to the process (since they are no process) and it will never realised that the partial width is larger than the total width written in the param_card.
(which is the reason for the warning "Branching ratio larger than one for 35").

So I'm really puzzled about what you actually run because I failed to make a consistent picture here.

Cheers,

Olivier

Hi Oliver,
Sorry for not being clear, let me illustrate: I did two tests, using the same madspin card as posted above. The only difference is the proc card:

1. import model 2HDMtII_NLO
define l+ = e+ mu+ ta+
define l- = e- mu- ta-
generate p p > h3 > h2 z [QCD]
output AZHToLLtt_mA1000_mH600 -nojpeg

2. import model 2HDMtII_NLO
define l+ = e+ mu+ ta+
define l- = e- mu- ta-
generate p p > h3 > h2 z , h2 > h1 h1 [QCD]
output AZHToLLtt_mA1000_mH600 -nojpeg

In case 1, the code runs through and produce lhe files with h2, h3, z and gluons, but no h1.
In case 2, the code crashes with the "Decay processes cannot be pertubed" error.

Please let me know if I can provide more info, and thanks for the patience!

Best,
Jordan

Addional info:
The customize card:

set param_card frblock 1 1.00e+00
set param_card frblock 2 9.9e-01
set param_card higgs 3 0.00e+00
set param_card mass 25 125.0
set param_card mass 35 600.0
set param_card mass 36 1000.0
set param_card mass 37 1000.0
set param_card width 35 18.00
set param_card width 36 95.00
set param_card width 37 95.00
set param_card higgs 1 -11.88587
set param_card higgs 2 11.04078

Hi,

Ok I understand now.
So indeed you need to use the syntax #1 for the proc_card.

Now you are using CMS framework and I have no experience in that framework. So keep that in mind if you observe miss match between what I say and what you observe. (for example, we do not have any file customize_card in our framework).

The typical workflow (outside of CMS) with MadSpin and gridpack should be
1) generate the gridpack without madspin
2) generate a sample for initializing madspin
3) run madspin on the initial sample.
(obviously CMS did some automation for Step1-3 for madspin and other type of post-processing.)

Now the gridpack is finished and ready for production:
4) run the gridpack to generate a sample (MadSpin should be automaticly run here --for recent enough version--)

Now most of those step produce an event file
Step 1, should not generate any (but sometimes does generate a unweighted_events.lhe.gz)
Step 2, should normally generates an events named events.lhe.gz
Step3, should normally generate an events named decay_events.lhe.gz
Now step 4 should produce a file named "events.lhe.gz"

So the fact that you discuss a file named unweighted_events.lhe.gz seems to indicate that you are not looking at the correct file.
Since that name seems to indicate an event file before the decay with madspin.
Can you check if you have other files are available and/or if you do have the madspin banner in the header of file that you are looking at?

Cheers,

Olivier

Hi Oliver,

Much appreciated for your detailed and clear answer. I do find in the work directory that there is a file named unweighted_events_decayed.lhe.gz, in which I found the decayed h1 :)

At this stage, I am not 100% sure if this gridpack with madspin decay can be used for CMS sample production (sine they use Pythia for decay and parton shower etc), I will check with my analysis group and come back to you.
Thanks again for your help!

Best,
Jordan