top polarisation with Madgraph/Madspin

Hi,

MadSpin do not take into account the polarisation written in the lhe file for making the decay (since MadSpin want to make a full spin correlation decay).
The problem is that spin information associate to an intermediate particle is not physical and that you loose the full spin correlation effect by projecting it.

A correct method would be to implement a custom propagator in MG5 which apply this projection of both on shell and of-sheel propagator.

In practise, a lot of people pass to not full spin-correlation method for this method.
You can use the DECAY package which were shipped with MG5 version before version 2.0.0 (which is not officially support anymore) if
you want to have a method which keeps most of the spin correlation and which is based on the information written on the lhe file.

Cheers,

Olivier

On 05 May 2015, at 22:51, javier <email address hidden> wrote:

> Question #266533 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/266533
>
> Description changed to:
> Hello,
> I have a problem trying to reproduce the distribution of cos(theta) for the CMS top polarization measurement in this note:
> http://cds.cern.ch/record/1460102
>
> Now, to do this I generate two ttbar samples using Madgraph, one where
> both tops are set to have positive helicity and the other with negative
> helicity. This is done using 'generate p p > t t~' and then following
> the instructions here
>
> https://answers.launchpad.net/mg5amcnlo/+faq/2243
>
> Therefore, the four helicity configurations (for the ++ case) considered
> are:
>
> DATA (NHEL(I, 1),I=1,4) /-1,-1, 1, 1/
> DATA (NHEL(I, 2),I=1,4) /-1, 1, 1, 1/
> DATA (NHEL(I, 3),I=1,4) / 1,-1, 1, 1/
> DATA (NHEL(I, 4),I=1,4) / 1, 1, 1, 1/
> DATA IDEN/256/
>
> I checked that both t and t~ have the correct helicity in the last
> column of the lhe file. The tops are decayed using Madspin in the t->b e
> nu channel. The events are showered using Pythia8 and the angle between
> the direction of the lepton in the top rest frame and the top momentum
> in the ttbar rest frame is analyzed. I get the following result:
>
> http://postimg.org/image/a9hg6kmnn/
>
> I would expect the two distributions to be quite different one to the
> other, but they are almost the same. Is there something obviously wrong
> in my procedure? Do you have an idea of what can be failing?
>
> Thank you very much,
> Javier.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,
Thank you for your quick answer.
Could you, please, be more specific in the way to implement a custom propagator in MG5? Is this explained somewhere else?
I tried using the DECAY package in MG5 v 1.5.14 to decay events generated using 2.2.3 and it gives an error...
Of course letting Pythia do the decay does not solve the problem either.

Thank you. Best,
Javier.

Dear Javier,

The reference for custom propagator is that paper: arXiv:1308.1668

Cheers,

Olivier

Hi Olivier,
Thanks. Using the DECAY package in Madgraph 1.5.14 fixes the problem.

Cheers,
Javier.

Hi Olivier,
I read that for version 2.3 there will be the option 'set spinmode [bridge|none|madspin]'.
Just for my info, will any of these spin modes take the spin information from the lhe file and use it in the decay?

Best,
Javier.

Hi Javier,

Actually only set spin mode none is supported in 2.3
The mode “bridge” was intended to keep some spin information at the same level as “DECAY” does, but the implementation was not the one that I expected.
So this mode is going to be properly implemented at a later date and this is mode is currently not allowed to be used

Cheers,

Olivier

On 06 May 2015, at 23:56, javier <email address hidden> wrote:

> Question #266533 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/266533
>
> javier posted a new comment:
> Hi Olivier,
> I read that for version 2.3 there will be the option 'set spinmode [bridge|none|madspin]'.
> Just for my info, will any of these spin modes take the spin information from the lhe file and use it in the decay?
>
> Best,
> Javier.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

I wanted to add a question here.

If I were to implement the propagators for massive bosons, would the following be something similar to what you'd do? I don't think they are quite right but this is the best I could think of...

denominator = "P('mu', id) * P('mu', id) - Mass(id) * Mass(id) + complex(0,1) * Mass(id) * Width(id)"
# transverse
VT = Propagator(name = "VT",
                numerator = "complex(0,1) * (-1 * Metric(l1, l2))",
                denominator = denominator
               )

# longitudinal
VL = Propagator(name = "VL",
                numerator = "complex(0,1) * Metric(l1,'mu')* P('mu', id) * P(l2, id) / Mass(id)**2 ",
                denominator = denominator
               )

Hi Phillip,

I do not know, I would have check the literature to check how it was done in the past.
But I see the point of your decomposition and this doesn’t look awful, but I have no guarantee on the rest.

Cheers,

Olivier

On 02 Oct 2015, at 00:51, Philip Chang <email address hidden> wrote:

> Question #266533 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/266533
>
> Philip Chang posted a new comment:
> Hi Olivier,
>
> I wanted to add a question here.
>
> If I were to implement the propagators for massive bosons, would the
> following be something similar to what you'd do? I don't think they are
> quite right but this is the best I could think of...
>
> denominator = "P('mu', id) * P('mu', id) - Mass(id) * Mass(id) + complex(0,1) * Mass(id) * Width(id)"
> # transverse
> VT = Propagator(name = "VT",
> numerator = "complex(0,1) * (-1 * Metric(l1, l2))",
> denominator = denominator
> )
>
> # longitudinal
> VL = Propagator(name = "VL",
> numerator = "complex(0,1) * Metric(l1,'mu')* P('mu', id) * P(l2, id) / Mass(id)**2 ",
> denominator = denominator
> )
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.