Isolating single processes from the standalone (cpp) output of MG5

Hello Jack,

It sounds like you are on just the right track - to select certain diagrams, you set the amp[i]'s of the other diagrams to zero, and to select certain helicities, you skip the unwanted combinations from the helicities[i][j] array. Note however that selecting certain diagrams might not be physically meaningful, since this in general doesn't respect gauge invariance. Selecting certain helicities shouldn't be a problem though; just be careful with the meaning of helicities +1 and -1 for initial and final state particles, and for particles/antiparticles.

All the best,
Johan

Thanks Johan, I'm not too concerned about gauge invariance at the moment, just at making sure various helicity combinations match up with my calculations!
Do you know the convention for helicities for incoming/outgoing and particle/antiparticle? Since I am in the case of massless quarks (So I treat the v(p_positron) as a u(positron) I have just been calling everything left handed -1 and everything right handed +1.

I am also a bit concerned about what the GC_XX are in MadGraph. I have been working under the assumption that the couplings were parameterised into vector and axial components: (c_V + c_A * \gamma^5) but I realise now that madgraph my instead work with (c_L * P_L + c_R * P_R), do you know the conventions used for this?

The reason I think this is that my result agrees only when the Z emitting quark and the leptons are negative (the helicity of the other quark doesnt matter), and when I include anything with positive helicity my result deviates.
Cheers
Jack

Hi Jack,

> I am also a bit concerned about what the GC_XX are in MadGraph. I have been working under the assumption that the couplings
> were parameterised into vector and axial components: (c_V + c_A * \gamma^5) but I realise now that madgraph my instead work
> with (c_L * P_L + c_R * P_R), do you know the conventions used for this?

In fact, MG didn't assume any precise form for the coupling, i.e. the definition of the GC_XX depends of the model.
Our Standard Model comes from FeynRules, where the splitting in two component is done in a automatic (and therefore not really human friendly way)

For exemple:
for z e+ e-: [GC_50 (1-\gamma_5)/2 + GC_59 (3+\gamma_5)/2 ] * \gamma^ \mu
But z u u~ didn't use the same splitting:
[GC_51 * (1-\gamma_5)/2 + GC_58 * (5+3\gamma_5)/2)]

In order to know this:
import model sm
display interactions z e+ e-
this returns:
{
    'id': 40,
    'particles': [-11,11,23],
    'color': [1 ],
    'lorentz': ['FFV2', 'FFV4'],
    'couplings': {(0, 1): 'GC_59', (0, 0): 'GC_50'},
    'orders': {'QED': 1},
    'loop_particles': None,
    'type': 'base',
    'perturbation_type': None
}
so you can see that the lorentz structure is
GC_50 * FFV2 + GC_59 *FFV4
then you can have the definition of FFV2 and FFV4 by doing:
display lorentz FFV2
display lorentz FFV4
which returns (for FFV2):
structure : Gamma(3,2,-1)*ProjM(-1,1)
name : FFV2
spins : [2, 2, 3]
(for FFV4)
structure : Gamma(3,2,-1)*ProjM(-1,1) + 2*Gamma(3,2,-1)*ProjP(-1,1)
name : FFV4
spins : [2, 2, 3]

On this point this is in fact more easy to customize the output of the sm_v4 model where all coupling are splitted in Right / Left coupling.

Concerning the helicity, we use the HELAS convention:
http://ccdb5fs.kek.jp/cgi-bin/img_index?199124011

Cheers,

Olivier

Thank Olivier, thats cleared that up!