MadGraph5_aMC@NLO

launch ends with non zero status

Asked by realshen

Dear MG authors,

I meet with a 'launch ends with non zero status' problem when I try to simulate the process 'p p > w+ w- w+ [QCD]' in loop_sm-ckm model.
When I try to set the wolfenstein block as
    1 2.254700e-01 # lamWS
    2 0.000000e-00 # AWS
    3 0.000000e-00 # rhoWS
    4 0.000000e-00 # eta,
it meets with error. But if the three values(2,3,4) are set to be none-zero, the calculation goes well.
How can I solve this problem?

The log is as belows:
launch
Traceback (most recent call last):
  File "/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/bin/internal/extended_cmd.py", line 879, in onecmd
    return self.onecmd_orig(line, **opt)
  File "/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/bin/internal/extended_cmd.py", line 872, in onecmd_orig
    return func(arg, **opt)
  File "/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/bin/internal/amcatnlo_run_interface.py", line 1192, in do_launch
    evt_file = self.run(mode, options)
  File "/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/bin/internal/amcatnlo_run_interface.py", line 1547, in run
    self.run_all(job_dict, [['2', 'F', '%d' % i]], status, split_jobs = split)
  File "/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/bin/internal/amcatnlo_run_interface.py", line 3189, in run_all
    self.wait_for_complete(run_type)
  File "/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/bin/internal/amcatnlo_run_interface.py", line 3153, in wait_for_complete
    self.cluster.wait(self.me_dir, update_status)
  File "/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/bin/internal/cluster.py", line 621, in wait
    raise Exception, msg
Exception: program /home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/SubProcesses/P0_usx_wpwmwp/ajob6 2 F 0 launch ends with non zero status: 1. Stop all computation
Value of current Options:
              text_editor : None
              web_browser : None
        cluster_temp_path : None
                  timeout : 60
            cluster_queue : madgraph
         madanalysis_path : None
                   lhapdf : /home/syb/lhapdf/bin/lhapdf-config
                 mg5_path : /home/syb/works/lab/MG5_aMC_v2_2_3
           cluster_memory : None
    cluster_status_update : (600, 30)
             cluster_time : None
               hepmc_path : None
             pythia8_path : None
                hwpp_path : None
   automatic_html_opening : False
       cluster_retry_wait : 300
             stdout_level : None
          pythia-pgs_path : None
                  td_path : None
             delphes_path : None
              thepeg_path : None
             cluster_type : condor
         cluster_nb_retry : 1
         fortran_compiler : None
                  nb_core : 8
              auto_update : 7
      exrootanalysis_path : None
               eps_viewer : None
             syscalc_path : None
             cpp_compiler : None
                 run_mode : 2

Thanks in advance!

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
Rikkert Frederix Edit question
Solved by:
realshen
Solved:
Last query:
Last reply:
Revision history for this message
marco zaro (marco-zaro) said : 		#1

Hi,
can you check and paste here the end (last 20 lines) of the log.txt file inside

home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/SubProcesses/P0_usx_wpwmwp/GF6

or

home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/SubProcesses/P0_usx_wpwmwp/all_G6

(depending if you ran event generation or fixed order)?

Thanks,

Marco

Revision history for this message
realshen (shenyb960) said : 		#2

Hi,

Thanks for your reply!

As a matter of fact, there doesn't not exist a GF6 or all_G6 directory inside
/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/Subprocesses/P0_usx_wpwmwp/

Revision history for this message
marco zaro (marco-zaro) said : 		#3

Hi,
so can you try to run
./ajob6 2 F 0
from inside
/home/syb/works/lab/MG5_aMC_v2_2_3/myprocess/ppwwwp-ckm/Subprocesses/P0_usx_wpwmwp/
it should create a directory (all_G* or GF*), inside which you should have a log.txt file.
Let me know,

Cheers,

Marco

Revision history for this message
realshen (shenyb960) said : 		#4

Hi,

sorry for the long delay.

I have both tried again the whole generation process and run ajob6 shell alone, there still do not exist any GF6 or all_G6 files. There are ajob1-16 and GF1,GF2,GF3,GF8,GF9,GF10,GF11,GF12,GF17,GF18,GF30,GF31 only.
I use madgraph 2.2.3 to generate the code. My input is as follows, I don't konw whether I made any mistakes.
import model loop_sm-ckm
generate p p > w+ w- w+ [QCD]
output ppwwwp-ckm
launch
3
1(set the wolfenstein block as
    1 2.254700e-01 # lamWS
    2 0.000000e-00 # AWS
    3 0.000000e-00 # rhoWS
    4 0.000000e-00 # eta)

Revision history for this message
Rikkert Frederix (frederix) said : 		#5

Hello,

The reason why the code is failing, is because the code is not very smart. By using your parameters, some of the Born diagrams that the code creates are exactly zero (in particular diagrams with internal top quark propagators). The code explicitly checks that all diagrams give a non-zero result and it crashes if this is not the case.

To prevent this from happening, you should massage the model restriction file such that the code only generates diagrams that give a non-zero contributions to the final result. This is rather simple:
In the models/loop_sm/ directory, copy the 'restrict_ckm.dat' file to, e.g., 'restrict_ckm2x2.dat', and update the wolfenstein block in this file according to your needs (in particular set the parameter labelled AWS to zero). Then, instead of importing the model 'loop_sm-ckm', load the model 'loop_sm-ckm2x2' before generating the process. Loading the model with this updated restriction, tells the code generation that the contributions with the top quark propagators are going to be always zero, and therefore these contributions are not generated. You'll also see that in the param_card for the new process, you won't be able to change the AWS parameter anymore.

Please let me know if this resolves your problem.

Best regards,
Rikkert

Revision history for this message
realshen (shenyb960) said : 		#6

Hi,

Thanks for your reply!

I generate the code as you said and the problem is missing.
Thank you again for your instruction!