Polarization settings of VBS Same_sign WW process

Looks like your XS5 is the correct way to go.
Need to double check if I can make the other syntax the correct one, or if this is not possible for some reason.

Cheers,

Olivier

Dear Olivier,

Is there any news about this issue?

Thanks,
Jie

Hi Jie,

I'm not working for the physics department for the moment (I have only a part time job in the physics department).
So I will not investigate this more up to Monday. In any case this should not be a blocker for you, so I was not even expecting to report anything in this thread.

Cheers,

Olivier

Hi Jie,

So to sum up on this, since you seem to be confused here.
in 2.7.0 it seems that we do not always handle correctly the symmetry factor when we have identical particle decaying.
(which is your case)

So for the moment the contribution
add process p p > w-{T} w-{0} j j QED=4 QCD=0, w- > l- vl~ @ 4
is not multiply by a factor of 2 like it should (in principle it should not matter if you write w-{T} or w-{0}, it should be the same for MG5aMC)

In 2.7.1, we plan to have two mode one where the ordering matters (such that you can do
add process p p > w-{T} w-{0} j j QED=4 QCD=0, w- > l- vl~, w-> j j @ 4
where the w-{T} will always decay leptonically)
and a second mode which is unordered where
add process p p > w-{T} w-{0} j j QED=4 QCD=0, w- > l- vl~, w-> j j @ 4
will have both w decay in both mode.

In ordered mode, you will need to add w-{T} w-{0} and w-{0} w-{T}
 to get the full unpolarised result
while in the unordered mode you will need to include only one.

Thanks to have reported this to us.

Cheers,

Olivier

Hi,

Just to report on the progress with 2.7.1.
We are still validating/fixing issue but I wanted to share them with you such that you can comment and see if I miss some the bad behavior that you reported.

In the case of
p p > w+{0} w+{T} j j , w+ > e+ ve the cross-section will be twice bigger compare to 2.7.0 (and therefore you do not need to include the {T}{0} component here)

In the case of
p p > w+{0} w+{T} j j , w+ > e+ ve, w+ > mu+ vm.
The decay will be done in an "ordered way" i.e. this is equivalent to --non supported syntax-- p p > w+{0} w+{T} j j , w+{0} > e+ ve, w+{T} > mu+ vm

In the case of
p p > w+{0} w+{T} j j, w+ > l+ vl
The decay will be done for both w+ to both flavor

We are actually validating on a simpler process: p p > z z , z > e+ e-
and we have checked so far

1. that p p > z z , z > e+ e- [0.01056 +- 2.837e-05 pb]
is equivalent to the sum of
p p > z{0} z{T} , z > e+ e-
p p > z{T} z{T} , z > e+ e-
p p > z{0} z{0} , z > e+ e-
[0.01066792 +- 2.23143541338e-05 pb]

2. that p p > z z , z > e+ e-, z > mu+ mu- [0.02111 +- 5.673e-05 pb]
is equivalent to the sum of
p p > z{T} z{0} , z > e+ e- , z > mu+ mu-
p p > z{0} z{T} , z > e+ e- , z > mu+ mu-
p p > z{T} z{T} , z > e+ e-, z > mu+ mu-
p p > z{0} z{0} , z > e+ e-, z > mu+ mu-
[0.0212876 +- 4.47040156026e-05 pb]

3. that p p > z z , z > l+ l- [0.042149 +- 0.00012635]
is equivalent to the sum of
p p > z{T} z{0} , z > l+ l-
p p > z{T} z{T} , z > l+ l-
p p > z{0} z{0} , z > l+ l-
[0.04246 +- 0.0001064 pb]

Even if the cross-section are now fine, we still have issue due to some optimization of MG5aMC which are not valid anymore in presence of polarization of identical particles. I'm working on fixing that.

Thanks,

Olivier Mattelaer

Dear Olivier,

Thanks to let me know. I think the cross-sections are fine now. But we may also need to check the distribution of variables. Like the distributions in https://answers.launchpad.net/mg5amcnlo/+question/688752

And I also use mg270 generate four samples, their proc_card.dat:

define l- = e- mu- ta46 define l+ = e+ mu+ ta+
define vl˜ = ve˜ vm˜ vt˜
define vl = ve vm vt
generate p p > w+{X} w+{Y} j j QED=4 QCD=0, w+ > l+ vl
add process p p > w-{X} w-{Y} j j QED=4 QCD=0, w- > l- vl˜

Here, X and Y are replaced by all permutations of 0 (longitudinal helicity) and T (transverse helicity)

Thus, those four processes are
Process name X Y Cross section (pb).
LL 0 0 0.002023
TT T T 0.015951
LT 0 T 0.005220
TL T 0 0.005202
SSWW - - 0.028296
(SSWW is for inclusive process)

I found good agreement between 4 polarized samples and inlcusive sample:
http://jixiao.web.cern.ch/jixiao/polar/m_ptj1_LL.pdf
http://jixiao.web.cern.ch/jixiao/polar/m_ptj2_LL.pdf
http://jixiao.web.cern.ch/jixiao/polar/m_ptl1_LL.pdf
http://jixiao.web.cern.ch/jixiao/polar/m_ptl1_LL.pdf

Just for reference. I can do the same test once Madgraph5 2.7.1 is available.

Best regards,
Jie

Sooner or later the use of the take a look at technique command which you stated above, I am capable of discern out the issues in my model report. However, for other components of the version, the take a look at manner nevertheless indicates a problem. I will write to you with greater information about it. https://blendinggadgets.com/best-blender-to-crush-ice-and-make-smoothies/