HerwigPP with aMC@NLO

Heres the relevant lines from the bug report.

Traceback (most recent call last):
  File "/Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/madgraph/interface/extended_cmd.py", line 872, in onecmd
    return self.onecmd_orig(line, **opt)
  File "/Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/madgraph/interface/extended_cmd.py", line 865, in onecmd_orig
    return func(arg, **opt)
  File "/Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/madgraph/interface/amcatnlo_run_interface.py", line 1199, in do_launch
    self.run_mcatnlo(evt_file)
  File "/Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/madgraph/interface/amcatnlo_run_interface.py", line 2328, in run_mcatnlo
    files.mv(showerfile_dict[shower], hep_file)
  File "/Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/madgraph/iolibs/files.py", line 173, in mv
    shutil.move(path1, path2)
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/shutil.py", line 301, in move
    copy2(src, real_dst)
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/shutil.py", line 130, in copy2
    copyfile(src, dst)
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/shutil.py", line 82, in copyfile
    with open(src, 'rb') as fsrc:
IOError: [Errno 2] No such file or directory: '/Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/ppwwnlo/MCatNLO/RUN_HERWIGPP_1/MCATNLO_HERWIGPP.hepmc'
Value of current Options:
              text_editor : None
              web_browser : None
        cluster_temp_path : None
       group_subprocesses : Auto
                  timeout : 60
            cluster_queue : None
         madanalysis_path : None
                   lhapdf : ../lhapdf-5.9.1/bin/lhapdf-config
ignore_six_quark_processes : False
             stdout_level : 20
    loop_optimized_output : True
    cluster_status_update : (600, 30)
               hepmc_path : /usr/local/include/HepMC
             pythia8_path : /Users/harikrishnan/Downloads/pythia8185
                hwpp_path : /usr/local/include/Herwig++
   automatic_html_opening : False
       cluster_retry_wait : 300
      output_dependencies : external
          pythia-pgs_path : /Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/pythia-pgs
                 mg5_path : /Users/harikrishnan/Downloads/MG5_aMC_v2_1_1
                  td_path : None
             delphes_path : /Users/harikrishnan/Downloads/MG5_aMC_v2_1_1/Delphes
              thepeg_path : /usr/local/include/ThePEG
             cluster_type : condor
      exrootanalysis_path : None
         fortran_compiler : None
                      OLP : MadLoop
              auto_update : 7
         cluster_nb_retry : 1
               eps_viewer : None
             syscalc_path : None
                  fastjet : fastjet-config
                    gauge : unitary
      complex_mass_scheme : False
                  nb_core : 4
                 run_mode : 2

Hi Harikrishnan,
have you installed HepMC as well? HepMC has to be installed first,
before ThePEG and ThePEG has to be compiled with the flag
--with-hepmc=<path to hepmc> (you should look in the ThePEG
documentation for more detail).
This way, Herwig++ can output HepMC n-tuples during the run.
Cheers.
Paolo

Hello Paolo,
yes! i have it installed. in fact i am now able to shower the amc generated events separately thru standalone herwig++ and i successfully create .hepmc files so am suspecting theres something else that goes wrong when amc@nlo interacts with herwig++
Harikrishnan

Hi Harikrishnan,
good that you have HepMC on. Are you’re now saying that when you
do ./bin/shower run_** it goes through, while it fails if you do
./bin/generate_events (or launch) with the shower on?
Thanks.
Cheers.
Paolo

On 26 Jun 2014, at 06:46, Harikrishnan Ramani <email address hidden> wrote:

> Question #250762 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/250762
>
> Status: Answered => Open
>
> Harikrishnan Ramani is still having a problem:
> Hello Paolo,
> yes! i have it installed. in fact i am now able to shower the amc generated events separately thru standalone herwig++ and i successfully create .hepmc files so am suspecting theres something else that goes wrong when amc@nlo interacts with herwig++
> Harikrishnan
>
> --
> You received this question notification because you are a direct
> subscriber of the question.

Hi Paolo,
By standalone herwig I mean, I take the lhe events and supply it to herwig++ outside the structure of amc@nlo. essentially i treat amc@nlo as an event generator and shower it with my own herwig++ without amc@nlo being involved at all.
Sincerely
Harikrishnan

Hi Harikrishnan,
ok thanks for the explanation.
I see you are using 2.1.1. Could you just try with 2.1.2_beta2?
And also, could you try with 2.1.1 but with another process
(like p p > w+, without and with MadSpin)?
I'll anyway try to reproduce your error in 2.1.1.
Cheers.
Paolo

Hi Harikrishnan,
I could reproduce your HW++ error, it is due to the b mass passed to the Monte Carlo. In order for it to work
you should use a non-zero value for this mass (like 4.95, or so), while the default until recently was a zero value.

As a minimal solution for the samples you have already generated, you could go inside the header of the event
file, inside the block '<MonteCarloMasses>' and replace '5 0.000000E+00' with '5 0.495000E+01'.
If you plan to do further runs with HW++ for this same process, you can edit SubProcesses/MCmasses_HERWIGPP.inc
and replace 'mcmass(5)=0.d0' with 'mcmass(5)=4.95d0', so that the new samples of this process will automatically
inherit this change.
Moreover, you can also do the same in Template/NLO/SubProcesses/MCmasses_HERWIGPP.inc, so this modification
will be propagated to all new processes you'll generate in the future.

This b-mass fix will become public in the official release 2.1.2, to appear soon.

Cheers.
Paolo

Hi Paolo and Harikrishnan,

Do you know the command of installing Herwig7/Herwig++ in MadGraph5_aMC@NLO?
just do:
install herwig7
install herwig++
in the interactive command line session doesn't seem to work.
(or would they be installed automatically along with "install hepmc"?)

Many thanks.

best,
Shu

Dear all,
sorry for my late reply.

There is no such installation command for herwig++/7.
One needs to install it (together with ThePEG and HepMC)
externally, and specify the relevant paths in the file
input/mg5_configuration.txt

Best,
Paolo

> On 17 Jul 2017, at 17:03, Shu Li <email address hidden> wrote:
>
> Question #250762 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/250762
>
> Shu Li posted a new comment:
> Hi Paolo and Harikrishnan,
>
> Do you know the command of installing Herwig7/Herwig++ in MadGraph5_aMC@NLO?
> just do:
> install herwig7
> install herwig++
> in the interactive command line session doesn't seem to work.
> (or would they be installed automatically along with "install hepmc"?)
>
> Many thanks.
>
> best,
> Shu
>
> --
> You received this question notification because you are subscribed to
> the question.

Thanks a lot, Paolo.
Would you have the website or url with a recommended H7/H++ version that you would suggest us to download and use to interface MG5?
Many thanks.

best,
Shu

Dear Shu,

in https://herwig.hepforge.org <https://herwig.hepforge.org/> you can find the newest HW7 releases,
as well as the old ones. In principle I would recommend the latest,
7.1.1, but I suspect there could be some minor changes we still have
to implement in our interface in order to be fully compatible with that.

I recently (i.e. two or three months ago) used HW 7.0.3 with MG5_aMC,
that one worked fine for me.

Cheers.
Paolo

> On 24 Jul 2017, at 14:47, Shu Li <email address hidden> wrote:
>
> Question #250762 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/250762
>
> Shu Li posted a new comment:
> Thanks a lot, Paolo.
> Would you have the website or url with a recommended H7/H++ version that you would suggest us to download and use to interface MG5?
> Many thanks.
>
> best,
> Shu
>
> --
> You received this question notification because you are subscribed to
> the question.