sherpa is hosted by Hepforge, IPPP Durham
close Warning: Can't synchronize with repository "(default)" (/hepforge/svn/sherpa does not appear to be a Subversion repository.). Look in the Trac log for more information.

Opened 5 years ago

Closed 4 years ago

#304 closed defect (wontfix)

Sherpa crashes with "Missing colour partner"

Reported by: Owned by:
Priority: major Milestone: rel-2.2.0
Component: Unknown Version: 1.4.5
Keywords: Cc:


Dear Sherpa authors,

I tried generating events with the Run.dat file attached in Sherpa and when events start being generated Sherpa crashes with the following error. I use Sherpa-1.4.5.

I can see a few errors like this before the crash:

METS_Scale_Setter::SetScales?(): Failed to determine \mu. METS_Scale_Setter::SetScales?(): Failed to determine \mu. Error in Ahadic::Hadronize: blob seem to be fishy.

... (dump of particle information)

ERROR in Ahadic::SanityCheck? :

Momentum/particle-blob number violation for 0 remaining clusters (norm2 = 1e+10). Protoparticles = 0/ parts = 64 vs. 64 : (28.87,19.9,-0.9361,20.23)

Blob [32]( 7, Cluster Formation , 39 -> 64 @ (0,0,0,0) Incoming particles :

... (more particle data)

Do you have any idea for why this happens? Could you let me know how to fix it?

Cheers, Danilo

Sherpa: CS_Shower::TranslateAmplitude? throws fatal error:

Missing colour partner

Exception_Handler::GenerateStackTrace?(..): Generating stack trace {

0x7f8237312f20 in 'ATOOLS::Terminate()' (Exception_Handler.C:215) 0x402b71 in 'main'


Attachments (1)

Run.dat (1.3 KB) - added by 5 years ago.
Run card.

Download all attachments as: .zip

Change History (10)

Changed 5 years ago by

Attachment: Run.dat added

Run card.

comment:1 Changed 5 years ago by Frank Siegert

Hi Danilo,

I don't quite understand what you are trying to achieve with the given run card. You are generating electroweak gammagammajj production with effective Higgs couplings to gluons and photons, and are doing this including electroweak Sudakovs but without any additional EW emissions in the matrix elements. Have you already been discussing this setup with somebody from the Sherpa team or can you let us know what the purpose of this is?

Cheers, Frank

comment:2 Changed 5 years ago by

Hi Frank,

the purpose of this run card is to generate any possible EW (only) production of jjaa. It is a background for an analysis I'm working on. I would like the H->aa production to be included in this background and the H->qq (where q != b, since the b is massive and I used the 93 container for "j") as well.

I have not discussed this with anyone in the Sherpa team, so if you think that this run card has been incorrectly prepared, please let me know and we can discuss. I'm no expert and I assume I probably made a mistake in the run card preparation.

Why do you claim that there would not be any additional EW emissions in the matrix element? My objective here was to restrict the diagrams so that only EW interactions are used, since I have another run card in which the final state is reached with no QED interactions (Order_EW = 0, and everything else is the same as this run card).

Could you let me know where I made a mistake in this run card?

Cheers, Danilo

comment:3 Changed 5 years ago by

Hello Frank,

I would like to add to the previous comment that, using "ME_QED Off" and/or setting "CSS_EW_MODE 0" (in any combination) does not change this error. Since I have no leptons in the final state, following the Sherpa manual, I did not see the need to set ME_QED to Off (*). Could you let me know if this is needed to avoid double counting? Even if it is, it does not seem to solve the problem.

I followed a similar setup than the one in this process.

The differences are that I add one extra photon for the core process, the collision energy (and the particles) are different, I use no enhancement function and I do enable the fragmentation and hadron decays. The HEFT is also used in my setup, but I don't see how any of those changes would cause a crash due to a "Missing colour partner" when generating events.

I tried a few different combinations for the run card and none work, but in essence my objective is simply to generate electroweak jjaa production and include the Higgs effective coupling to quarks, gluons and photons.

Thanks a lot for your help, Danilo

(*) From the Sherpa Manual: "By default, only QCD splitting functions are enabled in the shower. If you also want to allow for photon splittings, you can enable them by using ‘CSS_EW_MODE=1’. Note, that if you have leptons in your matrix-element final state, they are by default treated by a soft photon resummation as explained in QED Corrections. To avoid double counting, this has to be disabled as explained in that section."

comment:4 Changed 5 years ago by

Hello Frank,

I wonder if you managed to figure out what is wrong, since I got no reply from you since a while. It seems that this is not the only situation in which this happens.

If I try to generate p p > j b a a + 1 jet (with CKKW and HEFT), I also get this problem. It also happens with bbja+1 jet. This does not happen with only 1 or 2 EWK orders. Changing the CSS_EW_MODE doesn't seem to help much.

Cheers, Danilo

comment:5 Changed 5 years ago by

Dear Frank,

did you manage to take a look at this? I updated my run card to use ME_QED=Off, HEFT but extra cuts at the core process, with no difference. After the cross section integration, the "Missing colour partner" error on the shower still happens.

Many thanks in advance, Danilo

comment:6 in reply to:  2 Changed 5 years ago by Frank Siegert

Hi Danilo,

I have finally found the time to look at this in a bit more detail. I would suggest to start with a very simple setup, even without any additional jets in the ME at first for testing, because otherwise it will get relatively expensive. Furthermore, you have to use a more recent Sherpa version to be able to generate such a 2->4 core process with ME+PS merging, so I suggest to start right away with Sherpa 2.1.1.

the purpose of this run card is to generate any possible EW (only) production of jjaa. It is a background for an analysis I'm working on.

If it wasn't for your next sentence (cf. below), the run card would look fairly simple, something like this:

  BEAM_1 = 2212; BEAM_ENERGY_1 = 4000;
  BEAM_2 = 2212; BEAM_ENERGY_2 = 4000;

  CORE_SCALE VAR{Abs2(p[2]+p[3]+p[4]+p[5])}

  Process 93 93 -> 22 22 93 93 93{0}
  Order_EW 4;
  CKKW sqr(30/E_CMS)
  End process;

  NJetFinder 2 30 0 0.3 -1 3.0 10.0;
  PT 22 30.0 E_CMS
  DeltaR 22 93 0.3 1000.0;

The three extra lines in the run section make sure that we stop any clustering at the 2->4 core process and thus allow for defining a factorisation/renormalisation scale based on the 2->4 momenta. Here I've just for simplicity used \mu=m_{jjyy}, but you can define any scale you want. I have also included explicit photon isolation cuts at parton level, which you can adapt according to your needs.

I would like the H->aa production to be included in this background and the H->qq (where q != b, since the b is massive and I used the 93 container for "j") as well.

So, now coming to this request of including H->yy in the pp->jjyy production. Since there is no H->yy vertex in the SM this can only be induced through top or W loops and thus such diagrams will carry an additional alpha_s2 or alpha2 with it. Due to this they could be separated from the other diagrams (o(alpha4 * alpha_s0)) in a gauge invariant way and one could generate this as a separate sample. Whether you generate it separately or not, there are a few more issues to take into consideration:

  • You'll have to enable the effective vertices by using the SM+EHC model, which only works with the Amegic matrix element generator.
  • Do you also want to allow gg->H couplings in the same way?
  • You'll have to think about which o(alpha)/o(alpha_s) all this implies. For example, if simply done like in your run card, this also includes processes like gg->yygg where the Higgs couples to gluons (o(alpha1 * alpha_s2)) and photons (o(alpha3)) through the effective coupling. Is that what you want?

As for H->qq couplings, by default the Yukawa couplings are disabled for massless particles, so you would have to enable them explicitly, cf. I'm not sure how much sense this makes for the light partons, and you would also have to consider that this will include Higgs production from light quarks.

I hope this helps to get you started.

Cheers, Frank

comment:7 Changed 5 years ago by

Hi Frank,

thanks a lot. I finally managed to use this run card and shower events correctly (for pp > jjaa+1jet CKKW etc.) if I add ME_QED=Off, but this error happens in the same way for p p > bbja + 1 jet CKKW etc. (Order_EW=3, but same setup otherwise). I wonder if the jjaa QED=4 setup would be ok in that scenario.

I tried now running the p p > bbja setup (Order_EW=3) using no merging at first, removing the CSS EW shower and ME_QED = Off, but with the three extra lines you mentioned.

I would prefer to have the Higgs effective coupling in Amegic, though, including gg->H from the HEFT using the triangle fermion loop (O(alpha * alpha_S2)) and photons (O(alpha3)), but I'm not quite sure how that is done in Sherpa. Does Order_EW count the orders in alpha_EM including all that is in HEFT, or does it count orders in HEFT vertices as one? I'm sorry, but I'm not quite sure how this works and it is possible my Order_EW is not correct. Is that what you meant?

Many thanks, Danilo

comment:8 Changed 5 years ago by

Hi Frank,

if I may add another question to the previous set: what does the EXCLUSIVE_CLUSTER_MODE line does? I looked all around the manuals and could not find a reference to it. Is this the line that would correct the colour connections in the showering? Does it have other effects?

Also, why exactly the 1.4.5 version would not be able to run the 2->4 core process in jjaa? (I already switched to 2.1.0, but it would be nice to know what is missing)

Cheers, Danilo

comment:9 Changed 4 years ago by Stefan Hoeche

Resolution: wontfix
Status: newclosed

Closing ticket due to inactivity and because it relates to an old version. Please reopen if needed.

Modify Ticket

Change Properties
as closed The owner will remain

Add Comment

E-mail address and name can be saved in the Preferences.

Note: See TracTickets for help on using tickets.