# Usage of "madspin mode" to decay of loop-induced process with MadSpin

Asked by Peter Meiring on 2018-11-06

A rather short question; I would like to generate pp>hh events with the decay chain included. I can do that with Madspin in mode "none":

define l = l+ l-
define v = vl vl~
generate p p > h h [QCD]
launch
edit madspin --after_line=banner import model heft ./models/heft/restrict_default.dat --bypass_check
decay h > b b~
decay h > w+ w- > l v l v --add
set nevents 10000

This Madspin mode works for the Higgs decay since it has spin 0, right? What consequences does this have however, for the W decays? Can I do the generation as shown above, or should I remove the W decay from the steps above, and let f.e. Pythia8 do those decays? And why can I not just set the Madspin mode to "ON"?

Cheers,

Peter Meiring

## Question information

Language:
English Edit question
Status:
Solved
For:
Assignee:
No assignee Edit question
Solved by:
Peter Meiring
Solved:
2018-11-06
Last query:
2018-11-06
2018-11-06
 Olivier Mattelaer (olivier-mattelaer) said on 2018-11-06: #1

Hi,

means
1) No spin correlation between production and decay
2) No off-shell effects (all particles are exactly on their pole mass.

The advantages compare to PY8 is that you can have very complex decay (and that within that decay you have full spin correlation/...). Like your four body of your example where the spin correlation between the lepton is very important.

Now in the case of the W, I would prefer to use PY8 indeed.

> And why can I not just set the Madspin mode to "ON"?

It is a limitation of MadLoop which is not able to create the matrix-element that we need in order to use the standard madspin method. If you know someone interested at coding that part into madloop this can obviously be done.

Cheers,

Olivier

> On 6 Nov 2018, at 10:03, Peter Meiring <email address hidden> wrote:
>
> New question #675957 on MadGraph5_aMC@NLO:
>
>
> A rather short question; I would like to generate pp>hh events with the decay chain included. I can do that with Madspin in mode "none":
>
> define l = l+ l-
> define v = vl vl~
> generate p p > h h [QCD]
> launch
> edit madspin --after_line=banner import model heft ./models/heft/restrict_default.dat --bypass_check
> decay h > b b~
> decay h > w+ w- > l v l v --add
> set nevents 10000
>
> This Madspin mode works for the Higgs decay since it has spin 0, right? What consequences does this have however, for the W decays? Can I do the generation as shown above, or should I remove the W decay from the steps above, and let f.e. Pythia8 do those decays? And why can I not just set the Madspin mode to "ON"?
>
> Cheers,
>
> Peter Meiring
>
> --

 Peter Meiring (pmeiring) said on 2018-11-06: #2

Hi Olivier,

Thank you very much, that helps a lot.
Now if I remove the W decays as follows:

define l = l+ l-
define v = vl vl~
generate p p > h h [QCD]
launch
edit madspin --after_line=banner import model heft ./models/heft/restrict_default.dat --bypass_check
decay h > b b~
decay h > w+ w- --add
set nevents 10000

It appears that there is no phase space. I guess this must be because the W's now are both on-shell, right? Sorry for bothering you with this, but how can I fix this (and why did this error not appear in the first method)?

Thanks again; Cheers,

Peter Meiring

 Olivier Mattelaer (olivier-mattelaer) said on 2018-11-06: #3

> It appears that there is no phase space. I guess this must be because
> the W's now are both on-shell, right? Sorry for bothering you with this,
> but how can I fix this (and why did this error not appear in the first
> method)?

Yes correct.

In the previous example
> decay h > w+ w- > l+ l- vl vl~ --add

You do not ask for W onshell, but just the W to be present in s-channel.
So they are no problem.

Cheers,

Olivier

> On 6 Nov 2018, at 10:43, Peter Meiring <email address hidden> wrote:
>
> Question #675957 on MadGraph5_aMC@NLO changed:
>
>
> Peter Meiring is still having a problem:
> Hi Olivier,
>
> Thank you very much, that helps a lot.
> Now if I remove the W decays as follows:
>
> define l = l+ l-
> define v = vl vl~
> generate p p > h h [QCD]
> launch
> edit madspin --after_line=banner import model heft ./models/heft/restrict_default.dat --bypass_check
> decay h > b b~
> decay h > w+ w- --add
> set nevents 10000
>
> It appears that there is no phase space. I guess this must be because
> the W's now are both on-shell, right? Sorry for bothering you with this,
> but how can I fix this (and why did this error not appear in the first
> method)?
>
> Thanks again; Cheers,
>
> Peter Meiring
>
> --

 Peter Meiring (pmeiring) said on 2018-11-06: #4

Thanks!

So would you suggest to still do:

decay h > w+ w- > l+ l- vl vl~ --add

But then let Pythia re-do the W decays?
Won't this give problems for the off-shell W, since it is not even in the event record?

Last question:
The madspin mode "none" does not allow for off-shell Higgs particles, so is there no way to include those cases?

Cheers,

Peter Meiring

 Olivier Mattelaer (olivier-mattelaer) said on 2018-11-06: #5

no,

> decay h > w+ w- > l+ l- vl vl~ --add

Doing this is perfect. and do not need Pythia8 for the decay of such W.

I thought that you were asking for the following process
generate g g > w+ w- [QCD]
where i would advised to use PY8 for such decay.

But here the decay of the W is done perfectly with the correct matrix-element (h to 4 lepton)
which include the full spin correlation between the four lepton.

> Last question:
> The madspin mode "none" does not allow for off-shell Higgs particles, so is there no way to include those cases?

Our code does not allow it. In this case I do not see the point. (the width of the Higgs is so small).
If you let Pythia8 do the decay (of the Higgs), it will perform a re-shuffling of the momenta.

Cheers,

Olivier

> On 6 Nov 2018, at 11:02, Peter Meiring <email address hidden> wrote:
>
> Question #675957 on MadGraph5_aMC@NLO changed:
>
>
> Peter Meiring is still having a problem:
> Thanks!
>
> So would you suggest to still do:
>
> decay h > w+ w- > l+ l- vl vl~ --add
>
> But then let Pythia re-do the W decays?
> Won't this give problems for the off-shell W, since it is not even in the event record?
>
> Last question:
> The madspin mode "none" does not allow for off-shell Higgs particles, so is there no way to include those cases?
>
> Cheers,
>
> Peter Meiring
>
> --