Poles not cancel when running aMC@NLO

Asked by Fengwangdong Zhang on 2013-11-07

Dear experts,

I generated two processes and outputed them into a directory by use of aMC@NLO. However, when I generated events, it always failed. The error information is as follows:

{
Traceback (most recent call last):
  File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/extended_cmd.py", line 825, in onecmd
    return self.onecmd_orig(line, **opt)
  File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/extended_cmd.py", line 818, in onecmd_orig
    return func(arg, **opt)
  File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 1057, in do_launch
    self.compile(mode, options)
  File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 2360, in compile
      self.check_tests(test, this_dir)
  File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 2372, in check_tests
    return self.parse_check_poles_log(pjoin(dir, '%s.log' % test))
  File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 2407, in parse_check_poles_log
    raise aMCatNLOError('Poles do not cancel, run cannot continue')
aMCatNLOError: Poles do not cancel, run cannot continue
}

Could you help me with this problem? Thanks.

Best regards,
Feng

Question information

Language:
English Edit question
Status:
Open
For:
MadGraph5_aMC@NLO Edit question
Assignee:
marco zaro Edit question
Last query:
2013-11-09
Last reply:
2013-11-08

Hi Feng,

We will need information about the processes to see what the problem is.
Did you look on the amc@nlo website to see if the process that you try is in the list of
process which can't be deal with amc@nlo?

Cheers,

Olivier

On Nov 7, 2013, at 8:51 AM, Fengwangdong Zhang <email address hidden> wrote:

> New question #238819 on MadGraph5:
> https://answers.launchpad.net/madgraph5/+question/238819
>
> Dear experts,
>
> I generated two processes and outputed them into a directory by use of aMC@NLO. However, when I generated events, it always failed. The error information is as follows:
>
> {
> Traceback (most recent call last):
> File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/extended_cmd.py", line 825, in onecmd
> return self.onecmd_orig(line, **opt)
> File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/extended_cmd.py", line 818, in onecmd_orig
> return func(arg, **opt)
> File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 1057, in do_launch
> self.compile(mode, options)
> File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 2360, in compile
> self.check_tests(test, this_dir)
> File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 2372, in check_tests
> return self.parse_check_poles_log(pjoin(dir, '%s.log' % test))
> File "/ClusterDisks/HDN14/WorkSpace/cmsMC/farm_MCNLOv2_beta3/EWKWAJJ_MC_SG/bin/internal/amcatnlo_run_interface.py", line 2407, in parse_check_poles_log
> raise aMCatNLOError('Poles do not cancel, run cannot continue')
> aMCatNLOError: Poles do not cancel, run cannot continue
> }
>
>
> Could you help me with this problem? Thanks.
>
> Best regards,
> Feng
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Fengwangdong Zhang (zhangfwd) said : #2

Hi Olivier,

I generated two processes:

1) p p > w+ a j j QCD=0 [QCD]
2) p p > w- a j j QCD=0 [QCD]

I didn't find the list of process on the website. Could you take a look or send to me the link?

Thanks.

Best regards,
Feng

marco zaro (marco-zaro) said : #3

Dear Feng,
I will try to reproduce your error on my computer.
In the meanwhile, can you tell me which version of aMC@NLO are you using and attach the proc_card, param_card and run_card?
(or send them to me by email)?
Thanks a lot,

Best!

Marco

Fengwangdong Zhang (zhangfwd) said : #4

Dear Marco,

All right. I have sent a email to you. Thanks.

Best,
Feng

Fengwangdong Zhang (zhangfwd) said : #5

Dear Marco,

Sure. I use the version as 2.0.0.beta3. Please find the cards in attachment.
Thanks very much for help.

Best,
Feng

At 2013-11-08 16:56:14,"marco zaro" <email address hidden> wrote:
>Your question #238819 on MadGraph5 changed:
>https://answers.launchpad.net/madgraph5/+question/238819
>
> Status: Open => Needs information
>
>marco zaro requested more information:
>Dear Feng,
>I will try to reproduce your error on my computer.
>In the meanwhile, can you tell me which version of aMC@NLO are you using and attach the proc_card, param_card and run_card?
>(or send them to me by email)?
>Thanks a lot,
>
>Best!
>
>Marco
>
>--
>To answer this request for more information, you can either reply to
>this email or enter your reply at the following page:
>https://answers.launchpad.net/madgraph5/+question/238819
>
>You received this question notification because you asked the question.

marco zaro (marco-zaro) said : #6

Hi Feng,
I did not receive the email with the cards.
I have tried the process in the beta4 (still in development) branch and the compilation went trough fine, with all tests passing.
I kept all settings to their default.
Here you find what i did

set group_subprocesses Auto
set ignore_six_quark_processes False
set loop_optimized_output True
set gauge unitary
set complex_mass_scheme False
import model sm
define p = g u c d s u~ c~ d~ s~
define j = g u c d s u~ c~ d~ s~
define l+ = e+ mu+
define l- = e- mu-
define vl = ve vm vt
define vl~ = ve~ vm~ vt~
import model loop_sm
generate p p > w- a j j [QCD]
output

Let me know,
Cheers,

Marco

On 08 Nov 2013, at 10:31, Fengwangdong Zhang <email address hidden> wrote:

> Question #238819 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/238819
>
> Status: Needs information => Open
>
> Fengwangdong Zhang gave more information on the question:
> Dear Marco,
>
> All right. I have sent a email to you. Thanks.
>
> Best,
> Feng
>
> --
> You received this question notification because you are the assignee for
> this question.

Fengwangdong Zhang (zhangfwd) said : #7

Hi marco,

Well, it may be because the email has some problems. I used the version as 2.0.0.beta3.

The information of my proc_card_mg5 is as follows:
{
set group_subprocesses Auto
set ignore_six_quark_processes False
set loop_optimized_output True
set gauge unitary
set complex_mass_scheme False
import model sm
define p = g u c d s u~ c~ d~ s~
define j = g u c d s u~ c~ d~ s~
define l+ = e+ mu+ ta+
define l- = e- mu- ta-
define vl = ve vm vt
define vl~ = ve~ vm~ vt~
import model loop_sm
generate p p > w+ a j j QCD =0 [QCD]
add process p p > w- a j j QCD=0 [QCD]
output EWKWAJJ_MC_SG
}

I modified some parameters in run_card:
{
jetalgo = -1
ptj = 20
etaj = 7
ptl =10
etal= 2.5
drll = 0.3
mll = 0
maxjetflavor = 4
}

I didn't change other parameters. Do you know where the problem located?

Thanks.

Best,
Feng

marco zaro (marco-zaro) said : #8

Dear Feng,
I see now the origin of the problem: you generate the process asking for QCD=0.
The problem is that there are many loop diagrams which have electroweak bosons (w,z,gamma) as loop particles.
These loops are currently discarded (in higgs production via vbf these contributions are either zero because of color or very small), because such weak boson in the loop would increase too much the rank of the numerator.
You would have been warned when the process is generated.
Since you don’t have some loops, poles do not cancel any longer.
If you are confident that discarded loops are small/negligible, you can comment out line 2471 (in 2.0.0beta4) of bin/internal/amcatnlo_run_interface.py
The line is
                tests.append('check_poles’)

ffurthermore, you should either disable the poles check in SubProcesses/BinothLHA.f, or set the tolerance of the poles check to a negative value for in FKSparams.dat
Let me know if this helps you.

BTW, why do you set QCD=0?

Best!
Marco

> Question #238819 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/238819
>
> Status: Answered => Open
>
> Fengwangdong Zhang is still having a problem:
> Hi marco,
>
> Well, it may be because the email has some problems. I used the version
> as 2.0.0.beta3.
>
> The information of my proc_card_mg5 is as follows:
> {
> set group_subprocesses Auto
> set ignore_six_quark_processes False
> set loop_optimized_output True
> set gauge unitary
> set complex_mass_scheme False
> import model sm
> define p = g u c d s u~ c~ d~ s~
> define j = g u c d s u~ c~ d~ s~
> define l+ = e+ mu+ ta+
> define l- = e- mu- ta-
> define vl = ve vm vt
> define vl~ = ve~ vm~ vt~
> import model loop_sm
> generate p p > w+ a j j QCD =0 [QCD]
> add process p p > w- a j j QCD=0 [QCD]
> output EWKWAJJ_MC_SG
> }
>
> I modified some parameters in run_card:
> {
> jetalgo = -1
> ptj = 20
> etaj = 7
> ptl =10
> etal= 2.5
> drll = 0.3
> mll = 0
> maxjetflavor = 4
> }
>
> I didn't change other parameters. Do you know where the problem located?
>
> Thanks.
>
> Best,
> Feng
>
> --
> You received this question notification because you are the assignee for
> this question.

Fengwangdong Zhang (zhangfwd) said : #9

Dear Marco,

Ah, that's it! Because the wajj process is the signal of our analysis, I thought the QCD background should be set to zero when generating events. I'm sorry that I was not quite familiar with the principle.

Yes, when I generated the process, it always warned me the information, but I didn't take it into account.

Now I just ask that if we don't set QCD =0, then the QCD background is inevitably mixed into the signal process, right?

Thanks very much!

Best,
Feng

Can you help with this problem?

Provide an answer of your own, or ask Fengwangdong Zhang for more information if necessary.

To post a message you must log in.