PYTHIA Error in ProcessLevel::checkColours: incorrect colour assignment

Bug #1829266 reported by mehdirahmani on 2019-05-15
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
MadGraph5_aMC@NLO
Undecided
Unassigned

Bug Description

Hi Oliver,
/
Here is the bug report for the question #680841.

Attached is the lhe file of my events produced by MadGraph. Let me know if you need a description of my model and my model file. Here is the process:
generate p p > ZD, (ZD > SD SD~, SD > mu+ mu-, SD~ > mu+ mu-)

where ZD is a dark Z and SD is a dark scalar.

Thanks,

mehdirahmani (mehdirahmani) wrote :

Thanks for the file,

So indeed the input file has some color issue.
your quark/antiquark are in the correct 3/3bar representation.
This is likely related to your model.

While your modelname is inside our web database of model, I'm not able to generate your process and your param_card does not match the one expected for the online model.
So I can not check what is happening without your model. But the problem is likely within your model definition where the color structure of the u u~ g vertex is not correct. Can you check it?

In the standard model, when I'm doing "display interactions u u~ g" I have:
Interactions 44 : u~ u g has the following property:
{
    'id': 58,
    'particles': [-2,2,21],
    'color': [1 T(2,1,0)],
    'lorentz': ['FFV1'],
    'couplings': {(0, 0): 'GC_11'},
    'orders': {'QCD': 1},
    'loop_particles': None,
    'type': 'base',
    'perturbation_type': None
}

If you do not have the same in your model, this is likely to be the issue.
Also do you have warning about color structure when loading your model?

Cheers,

Olivier

mehdirahmani (mehdirahmani) wrote :

Thanks you for helping me out Oliver,

I am fairly new to this game so excuse the dumb questions that follows:

1. "your quark/antiquark are in the correct 3/3bar representation."

what do you exactly mean by this? Are you referring to SU(3) irreducible representation? Do I have control over quarks representations in the model file? How can I correct this (if it is wrong)

2. "While your model name is inside our web database of model, I'm not able to generate your process and your param_card does not match the one expected for the online model."

Yes I used the DMsimp_s_spin1 model, but I have made some changes to this model. My model file is attached this message.

3. " But the problem is likely within your model definition where the colour structure of the u u~ g vertex is not correct. Can you check it?"

I do not know to modify colour structures of u u~ g in model file. I thought it is being imported from SM model. How can I check that? Where to look for this definition?

4. When I import model sm and do "display interactions u u~ g" I get the same message as you do. but when I also import my model file ie. DMsimp_s_spin1_UFO, I get:
"No matching for any interactions"
Now what does that mean? What can be done to resolve this?

5. "Also do you have warning about colour structure when loading your model?"

loading my model to where? If you mean Mathematica, then the answer is no, I do not see any message related to colour. If you mean MadGraph, then the answer is also no. This is the only message I get when I import my model to Madgraph after I imported sm model to it:

INFO: Change particles name to pass to MG5 convention
Kept definitions of multiparticles l- / j / vl / l+ / p / vl~ unchanged
Defined multiparticle all = g a ve vm vt ve~ vm~ vt~ u c t d s b u~ c~ t~ d~ s~ b~ z w+ h fr sd gammad zd w- sd~ e- mu- ta- fd11 fd12 fd21 fd22 e+ mu+ ta+

Thank you,

Sorry but I do not have mathematica.
Could you attach the UFO file (this will also ensure that we do not use different version of FR leading to difference in the UFO file).

1. Sorry I mean that they are in the incorrect representation. FeynRules is quite bad in handling the difference between 3 and 3bar representation. If I'm correct this is define by which particle is define as particle/anti-particles. This leads to a lot of error in that sector. (Please double check with the latest FR manual since this might be outdated now)

4. This means that you do not have a u u~ g vertex in your model! Since we know that the definition of 3/3bar representation is unstable in FR, we typically use such type of interaction to correct the the 3/3bar representation coming from FR based on this type of interaction.
If you do not have coupling to gluon, then we are forced to trust FR.
So likely the easiest solution to your issue is to re-include that interaction.

Cheers,

Olivier

mehdirahmani (mehdirahmani) wrote :

Hi Oliver,

1. I have attached my zipped UFO folder. In this model the vertex "u u~ g" is included now:
L = ubar.Ga[mu].(gVu11 + gAu11 Ga[5]).u G[mu];

2. After u u~ g vertex is added, when I import sm model and my model in MadGraph and do "display interactions u u~ g", I get:

Interactions 3 : u~ u g has the following property:
{
    'id': 3,
    'particles': [-2,2,21],
    'color': [1 T(0,1)],
    'lorentz': ['FFV1', 'FFV2'],
    'couplings': {(0, 1): 'GC_5', (0, 0): 'GC_15'},
    'orders': {'DMV': 1},
    'loop_particles': None,
    'type': 'base',
    'perturbation_type': None
}

what does this mean now?

3. When I import my lhe file in and CMSSW and and run:

cmsDriver.py Configuration/GenProduction/python/Pythia8HadronizerFilter_13TeV_cfi.py --fileout file:out_sim.root --mc --eventcontent RAWSIM --datatier GEN-SIM --conditions 93X_mc2017_realistic_v3 --beamspot Realistic25ns13TeVEarly2017Collision --step GEN,SIM --geometry DB:Extended --era Run2_2017 --filetype LHE --filein file:events.lhe --python_filename SD_G.py --no_exec -n -1

and then cmsRun SD_G.py on it, I get the same error as before: "PYTHIA Error in ProcessLevel::checkColours: incorrect colour assignment". I have also attached a text file called errors.txt with PYTHIA output.

Thanks,

mehdirahmani (mehdirahmani) wrote :

Here is the PYTHIA output:

Hi,

> 'color': [1 T(0,1)],

means delta function. This is only possible if your particle 21 is colorless... Which does not help MG5aMC to
distinguish between 3 and 3 bar representation (and means your model is still wrong)

Cheers,

Olivier

> On 16 May 2019, at 19:37, mehdirahmani <email address hidden> wrote:
>
> Hi Oliver,
>
> 1. I have attached my zipped UFO folder. In this model the vertex "u u~ g" is included now:
> L = ubar.Ga[mu].(gVu11 + gAu11 Ga[5]).u G[mu];
>
> 2. After u u~ g vertex is added, when I import sm model and my model in
> MadGraph and do "display interactions u u~ g", I get:
>
> Interactions 3 : u~ u g has the following property:
> {
> 'id': 3,
> 'particles': [-2,2,21],
> 'color': [1 T(0,1)],
> 'lorentz': ['FFV1', 'FFV2'],
> 'couplings': {(0, 1): 'GC_5', (0, 0): 'GC_15'},
> 'orders': {'DMV': 1},
> 'loop_particles': None,
> 'type': 'base',
> 'perturbation_type': None
> }
>
> what does this mean now?
>
> 3. When I import my lhe file in and CMSSW and and run:
>
> cmsDriver.py
> Configuration/GenProduction/python/Pythia8HadronizerFilter_13TeV_cfi.py
> --fileout file:out_sim.root --mc --eventcontent RAWSIM --datatier GEN-
> SIM --conditions 93X_mc2017_realistic_v3 --beamspot
> Realistic25ns13TeVEarly2017Collision --step GEN,SIM --geometry
> DB:Extended --era Run2_2017 --filetype LHE --filein file:events.lhe
> --python_filename SD_G.py --no_exec -n -1
>
> and then cmsRun SD_G.py on it, I get the same error as before: "PYTHIA
> Error in ProcessLevel::checkColours: incorrect colour assignment". I
> have also attached a text file called errors.txt with PYTHIA output.
>
>
> Thanks,
>
>
> ** Attachment added: "DMsimp_s_spin1_UFO.tar.gz"
> https://bugs.launchpad.net/mg5amcnlo/+bug/1829266/+attachment/5264350/+files/DMsimp_s_spin1_UFO.tar.gz
>
> --
> You received this bug notification because you are subscribed to
> MadGraph5_aMC@NLO.
> https://bugs.launchpad.net/bugs/1829266
>
> Title:
> PYTHIA Error in ProcessLevel::checkColours: incorrect colour
> assignment
>
> Status in MadGraph5_aMC@NLO:
> New
>
> Bug description:
> Hi Oliver,
> /
> Here is the bug report for the question #680841.
>
> Attached is the lhe file of my events produced by MadGraph. Let me know if you need a description of my model and my model file. Here is the process:
> generate p p > ZD, (ZD > SD SD~, SD > mu+ mu-, SD~ > mu+ mu-)
>
> where ZD is a dark Z and SD is a dark scalar.
>
> Thanks,
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/mg5amcnlo/+bug/1829266/+subscriptions

mehdirahmani (mehdirahmani) wrote :

Thank you for your response.

Particle 21 is Gluon which is defined in sm model as follows:

  V[4] == {
    ClassName -> G,
    SelfConjugate -> True,
    Indices -> {Index[Gluon]},
    Mass -> 0,
    Width -> 0,
    ParticleName -> "g",
    PDG -> 21,
    PropagatorLabel -> "G",
    PropagatorType -> C,
    PropagatorArrow -> None,
    FullName -> "G"
  },

When I do "display particles g" in MG5_aMC I get:
{
    'name': 'g',
    'antiname': 'g',
    'spin': 3,
    'color': 8,
    'charge': 0.00,
    'mass': 'ZERO',
    'width': 'ZERO',
    'pdg_code': 21,
    'line': 'curly',
    'propagator': 0,
    'is_part': True,
    'self_antipart': True,
    'type': '',
    'counterterm': {}
},

Do I have to modify anything? Do you see any problems with these definitions?

The definition of the gluon at the UFO level correctly set the color representation to 8.
So that part is correct. Now I will not be able to help you on how to modify the FR model to have a consistent model.

Cheers,

Olivier

On 16 May 2019, at 21:13, mehdirahmani <<email address hidden><mailto:<email address hidden>>> wrote:

color': 8,

Changed in mg5amcnlo:
status: New → Invalid
mehdirahmani (mehdirahmani) wrote :

Hi again Oliver,

So I included an appropriate lagrangian for uplike uplike g vertex, now when I do "display interactions u u~ g" in MadGraph, I get:
{
    'id': 3,
    'particles': [-2,2,21],
    'color': [1 T(2,1,0)],
    'lorentz': ['FFV1'],
    'couplings': {(0, 0): 'GC_1'},
    'orders': {'QCD': 1},
    'loop_particles': None,
    'type': 'base',
    'perturbation_type': None
}

Is this correct now?

Then I went on and produced 10 events and tried to simulate cms on them. Now I think it works for 7 out of 10 events but it still gives me errors about 3 events: 3 Error in ProcessLevel::checkColours: incorrect colour assignment
but the final cross section after filter is not zero anymore. I have attached the output of cmsRun named errors_1.
Do you have further suggestions to solve this?

Hi,

Did you have your gluon coupling to ANY quark?
This error means that you still have issue in your model.

Cheers,

Olivier

> On 20 May 2019, at 20:50, mehdirahmani <email address hidden> wrote:
>
> Hi again Oliver,
>
> So I included an appropriate lagrangian for uplike uplike g vertex, now when I do "display interactions u u~ g" in MadGraph, I get:
> {
> 'id': 3,
> 'particles': [-2,2,21],
> 'color': [1 T(2,1,0)],
> 'lorentz': ['FFV1'],
> 'couplings': {(0, 0): 'GC_1'},
> 'orders': {'QCD': 1},
> 'loop_particles': None,
> 'type': 'base',
> 'perturbation_type': None
> }
>
> Is this correct now?
>
>
> Then I went on and produced 10 events and tried to simulate cms on them. Now I think it works for 7 out of 10 events but it still gives me errors about 3 events: 3 Error in ProcessLevel::checkColours: incorrect colour assignment
> but the final cross section after filter is not zero anymore. I have attached the output of cmsRun named errors_1.
> Do you have further suggestions to solve this?
>
> ** Attachment added: "errors_1.txt"
> https://bugs.launchpad.net/mg5amcnlo/+bug/1829266/+attachment/5265121/+files/errors_1.txt
>
> --
> You received this bug notification because you are subscribed to
> MadGraph5_aMC@NLO.
> https://bugs.launchpad.net/bugs/1829266
>
> Title:
> PYTHIA Error in ProcessLevel::checkColours: incorrect colour
> assignment
>
> Status in MadGraph5_aMC@NLO:
> Invalid
>
> Bug description:
> Hi Oliver,
> /
> Here is the bug report for the question #680841.
>
> Attached is the lhe file of my events produced by MadGraph. Let me know if you need a description of my model and my model file. Here is the process:
> generate p p > ZD, (ZD > SD SD~, SD > mu+ mu-, SD~ > mu+ mu-)
>
> where ZD is a dark Z and SD is a dark scalar.
>
> Thanks,
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/mg5amcnlo/+bug/1829266/+subscriptions

mehdirahmani (mehdirahmani) wrote :

I added gluon coupling to all uplike quarks:
LuqG := gs Ga[mu, s, r] T[a, i, j] uqbar[s, f, i].uq[r, f, j] G[mu, a];

>>This error means that you still have issue in your model.
but doesn't the "display u u~ g" command output mean that the model is fine now, or at least addressed your concern about "difference between 3 and 3bar representation in FR" and color structure?

Thanks,

But clearly your error means that some 3/3bar information are not yet correct (or it might be something else obviously)

> LuqG := gs Ga[mu, s, r] T[a, i, j] uqbar[s, f, i].uq[r, f, j] G[mu, a];

Do. no know what this line means actually.
But I would guess that it does not include the c c~ g interaction (neither the d d~ one, ....)
I guess you need to check that ALL particle in 3/3bar state does couple to the gluon.

Cheers,

Olivier

> On 20 May 2019, at 22:27, mehdirahmani <email address hidden> wrote:
>
> I added gluon coupling to all uplike quarks:
> LuqG := gs Ga[mu, s, r] T[a, i, j] uqbar[s, f, i].uq[r, f, j] G[mu, a];
>
>>> This error means that you still have issue in your model.
> but doesn't the "display u u~ g" command output mean that the model is fine now, or at least addressed your concern about "difference between 3 and 3bar representation in FR" and color structure?
>
> Thanks,
>
> --
> You received this bug notification because you are subscribed to
> MadGraph5_aMC@NLO.
> https://bugs.launchpad.net/bugs/1829266
>
> Title:
> PYTHIA Error in ProcessLevel::checkColours: incorrect colour
> assignment
>
> Status in MadGraph5_aMC@NLO:
> Invalid
>
> Bug description:
> Hi Oliver,
> /
> Here is the bug report for the question #680841.
>
> Attached is the lhe file of my events produced by MadGraph. Let me know if you need a description of my model and my model file. Here is the process:
> generate p p > ZD, (ZD > SD SD~, SD > mu+ mu-, SD~ > mu+ mu-)
>
> where ZD is a dark Z and SD is a dark scalar.
>
> Thanks,
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/mg5amcnlo/+bug/1829266/+subscriptions

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers