A value of the decay width and selection criteria

Hi,

1) I do not agree with you that sigma_all ~ GX^-1
This would be True in the Narrow-width approximation which is not your case here.

2) I do not know which cut you use. but it is certainly possible to have a dependency in the width.
simply because you allow less/more energetic events those one can be more boosted/...)

Cheers,

Olivier

> On 12 Jul 2019, at 09:23, Maksym Ovchynnikov <email address hidden> wrote:
>
> New question #681983 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/681983
>
> I have a process
>
> p p > X+Y, X > mu+ mu-
>
> Here, Y is schematically some SM particle, while X is the new particle with some mass and decay width.
> They are chosen as mX = 1 GeV, GX = 1 GeV.
>
> The choice of $GX$ obviously changes the cross section of the above process. However, I need to calculate the fraction of events eta satisfying some cuts on the final products mu+ mu-. Naively, I expect that the value of the fraction does not depend on the value of GX, since it cancels as a simple pre-factor:
>
> eta = sigma|_cuts/sigma_all, where sigma|_cuts ~ GX^-1, sigma_all ~ GX^-1
>
> However, I am not sure whether the MadGraph interprets this in the similar way. Could you please tell me?
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thank you. Could you please tell me what the syntax should I use for calculating the cross section in the narrow width approximation, and what value of GX should be in order to satisfy it in MadGraph? For mX = 1 GeV, what should be the value of GX?

Hi,

What the hell....

what value of GX should be in order to satisfy it in MadGraph?

MadGraph is "happy" with whatever bullshit you provide. Do not try to put the culprint on a code.

So let's talk about physics:

1) The width is NOT a free parameter of your model (in general), you should compute that value for your benchmark and not scan on it. (For some model with free-parameter for invisible decay, you can only have a minimal value off-course).

2) Narrow width approximation is a physical hyppothesis. In general this requires that the width should be at most a couple of percent of the mass But this is NOT an enough condition to have such approximation valid.
The real condition is that the matrix-element should be flat within N times the width. So this forbids any particles within that invariant mass gap, but also any threshold within such gap. (and I probably miss a lot of effects that invalidates NWA)

So for me the correct question is
1) What is the "correct" value of your width?
2) For such value are you in Narrow-width approximation?

 Could you please tell me what the syntax should I use for
calculating the cross section in the narrow width approximation,

Your syntax:
p p > X Y, X > mu+ mu-
assumes that NWA is valid but does not use NWA for the computation.
If you want to use the NWA for the value of the cross-section, then you need to use MadSpin.
Now MadSpin will have issue as soon as NWA is not valid (and/or when your width is not the LO one)
(issue will be wrong cross-section --obviously--, slow code and potentially crash)

If you ask madspin to not have spin-correlation and not have off-shell effects, then MadSpin can give you some result even if NWA is not valid. (Now the result are as good as your hyppothesis...)
This can be done via:
generate p p > X Y
output
launch
madspin=none
decay X > mu+ mu-

The final cross-section will be using NWA (even if it is not valid here).
More exactly it will take the cross-section for p p > X Y
and multiply it by the width of the decay process (X > mu+ mu-)
and divide by the total width (as written in your param_card).
Consequently if your provide a total width smaller than the --partial-- decay width to muon, your
full cross-section will be bigger than your production one (which is as good as your input parameter, i.e. bad)

Cheers,

Olivier

On 12 Jul 2019, at 09:43, Maksym Ovchynnikov <<email address hidden><mailto:<email address hidden>>> wrote:

Question #681983 on MadGraph5_aMC@NLO changed:
https://answers.launchpad.net/mg5amcnlo/+question/681983

   Status: Answered => Open

Maksym Ovchynnikov is still having a problem:
Thank you. Could you please tell me what the syntax should I use for
calculating the cross section in the narrow width approximation, and
what value of GX should be in order to satisfy it in MadGraph? For mX =
1 GeV, what should be the value of GX?

--
You received this question notification because you are an answer
contact for MadGraph5_aMC@NLO.

Dear Olivier,

thank you for the answer!

Below I just want to clarify the underlying reason of my question.

In my model the decay width is always much smaller than its mass. I just wanted to know whether MadGraph's predictions for the amount of selected events depend on the value of the decay width. Initially I just did not know to insert the precise value of the decay width for each mass and thought that the syntax

p p > X+Y, X > mu+ mu-

forces MadGraph to calculate the cross section using the narrow width approximation. And it does not matter if the cross section of X production times Br(X-> mu mu) is larger than the cross section of X production itself, since I was interested in the fraction of events, not in the value of the cross section.

This question was expired because it remained in the 'Open' state without activity for the last 15 days.