Problem decaying tops with Madspin

Asked by Afiq Aize

Hi,

I encountered a strange problem where suddenly MadSpin refuses to decay the tops. The same thing I did in the past worked without problem, but I'm stuck at figuring out what went wrong this time. Here's the error message:

Command "launch" interrupted with error:
KeyError : ((2, 21), (-6, 2, 6))
Please report this bug to developers

           More information is found in 'MS_debug'.

           Please attach this file to your report.

Any help on this is greatly appreciated. The MS_debug file content below [1].

Best regards,
Afiq

[1]
set seed 849483662

decay t > w+ b, (w+ > l+ vl)
decay t~ > w- b~, (w- > l- vl~)
decay w+ > l+ vl
decay w- > l- vl~
import output/smtt-nlo-02/Events/run_01/events.lhe.gz
launch
Traceback (most recent call last):
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/madgraph/interface/extended_cmd.py", line 908, in onecmd
    return self.onecmd_orig(line, **opt)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/madgraph/interface/extended_cmd.py", line 897, in onecmd_orig
    return func(arg, **opt)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/interface_madspin.py", line 537, in do_launch
    generate_all.run()
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/decay.py", line 2091, in run
    self.get_max_weight_from_event(decay_mapping)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/decay.py", line 3089, in get_max_weight_from_event
    production_tag, event_map = self.load_event()
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/decay.py", line 3226, in load_event
    P_order = self.all_ME[production_tag]['tag2order'][production_tag]
KeyError: ((2, 21), (-6, 2, 6))
Value of current Options:
                 spinmode : madspin
             onlyhelicity : False
                     seed : 849483662
      max_weight_ps_point : 400
                 nb_sigma : 6.31936672574
                 run_card : None
   Nevents_for_max_weigth : 221
                   BW_cut : 15.0
              use_old_dir : False
               max_weight : -1
                 curr_dir : /nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/output/smtt-nlo-02/Events/run_01/../..
                   ms_dir : None
      max_running_process : 100

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
Last query:
Last reply:
Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said :
#1

Can you try with the 2.4.0.beta version?

This sounds a bug resolved that version.

Cheers,

Olivier
> On May 9, 2016, at 20:52, Afiq Aize <email address hidden> wrote:
>
> New question #293513 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/293513
>
> Hi,
>
> I encountered a strange problem where suddenly MadSpin refuses to decay the tops. The same thing I did in the past worked without problem, but I'm stuck at figuring out what went wrong this time. Here's the error message:
>
> Command "launch" interrupted with error:
> KeyError : ((2, 21), (-6, 2, 6))
> Please report this bug to developers
>
> More information is found in 'MS_debug'.
>
> Please attach this file to your report.
>
> Any help on this is greatly appreciated. The MS_debug file content below [1].
>
> Best regards,
> Afiq
>
> [1]
> set seed 849483662
>
>
> decay t > w+ b, (w+ > l+ vl)
> decay t~ > w- b~, (w- > l- vl~)
> decay w+ > l+ vl
> decay w- > l- vl~
> import output/smtt-nlo-02/Events/run_01/events.lhe.gz
> launch
> Traceback (most recent call last):
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/madgraph/interface/extended_cmd.py", line 908, in onecmd
> return self.onecmd_orig(line, **opt)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/madgraph/interface/extended_cmd.py", line 897, in onecmd_orig
> return func(arg, **opt)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/interface_madspin.py", line 537, in do_launch
> generate_all.run()
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/decay.py", line 2091, in run
> self.get_max_weight_from_event(decay_mapping)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/decay.py", line 3089, in get_max_weight_from_event
> production_tag, event_map = self.load_event()
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/MadSpin/decay.py", line 3226, in load_event
> P_order = self.all_ME[production_tag]['tag2order'][production_tag]
> KeyError: ((2, 21), (-6, 2, 6))
> Value of current Options:
> spinmode : madspin
> onlyhelicity : False
> seed : 849483662
> max_weight_ps_point : 400
> nb_sigma : 6.31936672574
> run_card : None
> Nevents_for_max_weigth : 221
> BW_cut : 15.0
> use_old_dir : False
> max_weight : -1
> curr_dir : /nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc233/output/smtt-nlo-02/Events/run_01/../..
> ms_dir : None
> max_running_process : 100
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Afiq Aize (watashinokonoka) said :
#2

Hi,

Now I end up with a probably the same bug:

Command "launch" interrupted with error:
KeyError : ((1, 21), (-6, 1, 6))
Please report this bug to developers

           More information is found in 'MS_debug'.

           Please attach this file to your report.

File content here [1].

Btw I got a strange error when changing some options such as giving a pythia, fastjet, lhapdf paths into the mg5_configuration.txt [2].

Any idea what could be the issue?

Best regards,
Afiq

[1]
set seed 870034820

decay t > w+ b, (w+ > l+ vl)
decay t~ > w- b~, (w- > l- vl~)
decay w+ > l+ vl
decay w- > l- vl~
import output/smtt-nlo-02/Events/run_01/events.lhe.gz
launch
Traceback (most recent call last):
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/extended_cmd.py", line 1009, in onecmd
    return self.onecmd_orig(line, **opt)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/extended_cmd.py", line 964, in onecmd_orig
    return func(arg, **opt)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/interface_madspin.py", line 566, in do_launch
    generate_all.run()
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/decay.py", line 2102, in run
    self.get_max_weight_from_event(decay_mapping)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/decay.py", line 3120, in get_max_weight_from_event
    production_tag, event_map = self.load_event()
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/decay.py", line 3259, in load_event
    P_order = self.all_ME[production_tag]['tag2order'][production_tag]
KeyError: ((1, 21), (-6, 1, 6))
Value of current Options:
                 spinmode : madspin
             onlyhelicity : False
                     seed : 870034820
      max_weight_ps_point : 400
                 nb_sigma : 4.5
                 run_card : None
   Nevents_for_max_weigth : 75
                   BW_cut : 15.0
              use_old_dir : False
               max_weight : -1
                 curr_dir : /nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/output/smtt-nlo-02/Events/run_01/../..
                   ms_dir : None
      max_running_process : 100

[2]
load MG5 configuration from input/mg5_configuration.txt
set fastjet to /nfs/dust/cms/user/afiqaize/cms/mg5HC/fastjet313/bin/fastjet-config
set lhapdf to /nfs/dust/cms/user/afiqaize/cms/mg5HC/lhapdf616/bin/lhapdf-config
Traceback (most recent call last):
  File "./bin/mg5_aMC", line 153, in <module>
    cmd_line = interface.MasterCmd(mgme_dir = options.mgme_dir)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/master_interface.py", line 585, in __init__
    self.cmd.__init__(self, *args, **opt)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/madgraph_interface.py", line 2777, in __init__
    self.set_configuration()
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/master_interface.py", line 560, in set_configuration
    return self.cmd.set_configuration(self, *args, **opts)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/madgraph_interface.py", line 6192, in set_configuration
    return self.set_configuration(config_path, final)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/master_interface.py", line 560, in set_configuration
    return self.cmd.set_configuration(self, *args, **opts)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/madgraph_interface.py", line 6321, in set_configuration
    warnings = misc.mg5amc_py8_interface_consistency_warning(self.options)
  File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/various/misc.py", line 66, in mg5amc_py8_interface_consistency_warning
    if not options['mg5amc_py8_interface_path']:
KeyError: 'mg5amc_py8_interface_path'

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said :
#3

Hi,

For [2] thanks for reporting it, I have fixed it and this should be solved when 2.4.0 will be released.

For [1] .
Could you indicating me the various command that you run to face such problem?
or better indicate me the command reported in the header of the event file that you want to decay.

I have tried the following
generate p p > t t~ [QCD]
output
launch
madspin=ON
[then set the madspin card to the one that you specify]
and did not face any problem.

Just for information, it looks like MadSpin miss to generate one of the matrix element needed for the computation.
(more exactly it miss to generate u g > t t~ u). This is typically related to a definition problem of the core process.

Cheers,

Olivier

> On May 9, 2016, at 22:31, Afiq Aize <email address hidden> wrote:
>
> Question #293513 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/293513
>
> Status: Answered => Open
>
> Afiq Aize is still having a problem:
> Hi,
>
> Now I end up with a probably the same bug:
>
> Command "launch" interrupted with error:
> KeyError : ((1, 21), (-6, 1, 6))
> Please report this bug to developers
>
> More information is found in 'MS_debug'.
>
> Please attach this file to your report.
>
> File content here [1].
>
> Btw I got a strange error when changing some options such as giving a
> pythia, fastjet, lhapdf paths into the mg5_configuration.txt [2].
>
> Any idea what could be the issue?
>
> Best regards,
> Afiq
>
> [1]
> set seed 870034820
>
>
> decay t > w+ b, (w+ > l+ vl)
> decay t~ > w- b~, (w- > l- vl~)
> decay w+ > l+ vl
> decay w- > l- vl~
> import output/smtt-nlo-02/Events/run_01/events.lhe.gz
> launch
> Traceback (most recent call last):
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/extended_cmd.py", line 1009, in onecmd
> return self.onecmd_orig(line, **opt)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/extended_cmd.py", line 964, in onecmd_orig
> return func(arg, **opt)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/interface_madspin.py", line 566, in do_launch
> generate_all.run()
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/decay.py", line 2102, in run
> self.get_max_weight_from_event(decay_mapping)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/decay.py", line 3120, in get_max_weight_from_event
> production_tag, event_map = self.load_event()
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/MadSpin/decay.py", line 3259, in load_event
> P_order = self.all_ME[production_tag]['tag2order'][production_tag]
> KeyError: ((1, 21), (-6, 1, 6))
> Value of current Options:
> spinmode : madspin
> onlyhelicity : False
> seed : 870034820
> max_weight_ps_point : 400
> nb_sigma : 4.5
> run_card : None
> Nevents_for_max_weigth : 75
> BW_cut : 15.0
> use_old_dir : False
> max_weight : -1
> curr_dir : /nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/output/smtt-nlo-02/Events/run_01/../..
> ms_dir : None
> max_running_process : 100
>
> [2]
> load MG5 configuration from input/mg5_configuration.txt
> set fastjet to /nfs/dust/cms/user/afiqaize/cms/mg5HC/fastjet313/bin/fastjet-config
> set lhapdf to /nfs/dust/cms/user/afiqaize/cms/mg5HC/lhapdf616/bin/lhapdf-config
> Traceback (most recent call last):
> File "./bin/mg5_aMC", line 153, in <module>
> cmd_line = interface.MasterCmd(mgme_dir = options.mgme_dir)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/master_interface.py", line 585, in __init__
> self.cmd.__init__(self, *args, **opt)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/madgraph_interface.py", line 2777, in __init__
> self.set_configuration()
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/master_interface.py", line 560, in set_configuration
> return self.cmd.set_configuration(self, *args, **opts)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/madgraph_interface.py", line 6192, in set_configuration
> return self.set_configuration(config_path, final)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/master_interface.py", line 560, in set_configuration
> return self.cmd.set_configuration(self, *args, **opts)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/interface/madgraph_interface.py", line 6321, in set_configuration
> warnings = misc.mg5amc_py8_interface_consistency_warning(self.options)
> File "/nfs/dust/cms/user/afiqaize/cms/mg5HC/mg5amc240b/madgraph/various/misc.py", line 66, in mg5amc_py8_interface_consistency_warning
> if not options['mg5amc_py8_interface_path']:
> KeyError: 'mg5amc_py8_interface_path'
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Afiq Aize (watashinokonoka) said :
#4

Hi,

It's exactly the same as your example, but I had instead of p p this:

generate g g > t t~ [QCD]

Because I don't want the qq diagrams.

Best regards,
Afiq

Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said :
#5

Hi,

So this is the problem @NLO you can not (in general) ask for g g in the initial state.

Indeed starting from g g and add one QCD emission and you arrive one
u g > u t t~
but this set of diagram can also came from the QCD emission (of a gluon) from
u u~ > t t~

Therefore you are not allowed to only generate g g > t t~.
You should have a huge warning associated to this.
I'm actually surprise that our test at run time do not fail for such process.
Since I doubt that such definition is IR safe.

Cheers,

Olivier

Revision history for this message
Afiq Aize (watashinokonoka) said :
#6

Hi,

Ok, so either I go to LO or do p p @ NLO. Indeed there was a warning but due to my ignorance of its severity I just ignored it. Sorry about that.

I guess the problem is solved. Thanks!

Best regards,
Afiq

Revision history for this message
Afiq Aize (watashinokonoka) said :
#7

Thanks Olivier Mattelaer, that solved my question.