2HDM and pythia8 unphysical colour flow

Could you copy paste here the events that fails the color-flow information inside pythia8?

Thanks,

Olivier

Hi Olivier,

you can find a file with the event content in https://www.dropbox.com/s/jilv27aulwebbzu/event1.txt?dl=0
I don't know if any events would run in pythia successfully, because the first one fails.

Cheers,
  Alberto

Hi all,

thanks for the replies. I will try with version 2.3 and let you know whether it works.

Cheers,
  Alberto

Hi all,

I tried the new version and it seems to work fine. I generated a process similar to the one in the example (the only difference being that instead of decaying to anything I was simulating 2 leptonic and 2 hadronic decays of the Ws) and passed to Pythia. All events in the sample were successfully showered.

Thanks a lot.

Cheers,
   Alberto

Thanks Olivier Mattelaer, that solved my question.

That's good news,

thanks,

Olivier

Hi Olivier and Pierre,

could this change in MadSpin be included as a patch in version 2.2.2? For
personal production I'm using 2.3.0 now, but for centralized production
within my collaboration I should restrict to 2.2.2.

Cheers,
   Alberto

Hi,

I have send you the patch file which upgrade MadSpin from the 2.2.2 version to the one present in 2.3

Cheers,

Olivier

On 09 May 2015, at 11:36, Alberto Gascón <email address hidden> wrote:

> Question #266488 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/266488
>
> Alberto Gascón posted a new comment:
> Hi Olivier and Pierre,
>
> could this change in MadSpin be included as a patch in version 2.2.2? For
> personal production I'm using 2.3.0 now, but for centralized production
> within my collaboration I should restrict to 2.2.2.
>
> Cheers,
> Alberto
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thanks Alberto,

I get the model but you forget to send me the benchmark point and which definition of particle you use for "nuh".

Actually for "nuh = nu4 nu5" and with the default benchmark point, I can reproduce the "error message" which I fixed by the patch shown below.
However this error happens only if you matrix-element evaluation is exactly zero, which tends to indicate that I'm using a stupid benchmark point (I actually have a zero cross-section).

Cheers and thanks,

Olivier

=== modified file 'madgraph/iolibs/template_files/matrix_madevent_group_v4.inc'
--- old/madgraph/iolibs/template_files/matrix_madevent_group_v4.inc 2014-11-24 14:15:44 +0000
+++ new/madgraph/iolibs/template_files/matrix_madevent_group_v4.inc 2015-05-10 04:49:09 +0000
@@ -176,7 +176,7 @@

           WRITE(HEL_BUFF,'(20i5)')(NHEL(II,I),II=1,NEXTERNAL)
     ENDIF
- IF (ISUM_HEL .NE. 1.or.(HEL_PICKED.eq.-1)) THEN
+ IF (ANS.ne.0d0.and.(ISUM_HEL .NE. 1.or.HEL_PICKED.eq.-1)) THEN
     CALL RANMAR(R)
     SUMHEL=0d0
     DO I=1,NCOMB

Hi Olivier,

after including the patch in version 2.2.2, when I try to open madspin standalone in the command line I get the following error message:

    import tests.parallel_tests.test_aloha as test_aloha
ImportError: No module named test_aloha

I see that before the patch the import was:

import tests.unit_tests.various.test_aloha as test_aloha

Is there something missing on my configuration? Or should the library imported have a different name?

Cheers,
   Alberto

Hi,

Do you have bazaar (bzr in the shell)

If yes, then the easiest is to do:

bzr branch lp:~maddevelopers/mg5amcnlo/2.3.1_fix_addmothers

and you will have the exact same setup than mine.

Cheers,

Olivier

On 12 May 2015, at 16:06, Alberto Gascón <email address hidden> wrote:

> Question #266488 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/266488
>
> Alberto Gascón posted a new comment:
> Hi Olivier,
>
> after including the patch in version 2.2.2, when I try to open madspin
> standalone in the command line I get the following error message:
>
> import tests.parallel_tests.test_aloha as test_aloha
> ImportError: No module named test_aloha
>
> I see that before the patch the import was:
>
> import tests.unit_tests.various.test_aloha as test_aloha
>
> Is there something missing on my configuration? Or should the library
> imported have a different name?
>
> Cheers,
> Alberto
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

the version I get this way is equivalent to 2.2.2 + patches ?

Alberto

Hi,

No this correspond more to 2.3.0 + a lot of patches. Or 2.3.1 minus a lot of patches but + the patches concerning epsilon_ijk

The final 2.3.1 should be the merge of the current version named 2.3.1 and the epsilon_ijk patch present in that version.

Cheers,

Olivier

On 12 May 2015, at 17:32, Alberto Gascón <email address hidden> wrote:

> Question #266488 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/266488
>
> Alberto Gascón posted a new comment:
> Hi Olivier,
>
> the version I get this way is equivalent to 2.2.2 + patches ?
>
> Alberto
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

sorry but, the versions you're referring to are MadGraph or MadSpin versions?
I'm really limited to stick to MadGraph version 2.2.2 for the time being.

Cheers,
   Alberto

Dear Alberto,

Ok sorry, I mixed two conversation.
I was confused by this sentences:

"""
    import tests.parallel_tests.test_aloha as test_aloha
ImportError: No module named test_aloha

I see that before the patch the import was:

import tests.unit_tests.various.test_aloha as test_aloha
"""

The patch that i send in this thread only change one line in one file: madgraph/iolibs/template_files/matrix_madevent_group_v4.inc
So I have no idea how you have modified the import of the test_aloha module which indeed change path between 2.2.x and 2.3.x

Cheers,

Olivier

Hi Olivier,

that change was in the patch you sent by email, updating MadSpin to 2.3.1.
In message #11, when I said I had included the patch, I meant this one,
not the one in message #10

Cheers,
   Alberto

Ah ok now I understand,

MadSpin is a part of the MG package suite and therefore they share the same version number.
The patch that i send to you was upgrading MadSpin to 2.3.1 without touching to the rest of the code.
Which is obviously not an ideal situation. As you notice this actually seems to not work.
Obviously, You can patch the patch to not use the new import but the old one, but you will still have problem that MadSpin use other files from the MG side, and that it is difficult to test if this is ok or not. (I guess it should be as long as you do not use the new options of MadSpin linked to 2.3)

> I'm really limited to stick to MadGraph version 2.2.2 for the time being.

This is a weird statement to me, since if it is really true, then I can not do anything for you.
If the word "patch" allows to apply modification to the code, then I can create a patch which upgrades the full code from 2.2.2 to 2.3.1.

Cheers,

Olivier