Error while decaying through DECAY package

Hi,

We do not provide support anymore for the DECAY package.

Cheers,

Olivier

On 08 Jun 2015, at 15:56, Ramkrishna <email address hidden> wrote:

> New question #267900 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/267900
>
> Dear experts,
>
> I generated events using p p > w+ w- j j QCD=0, with the latest version of MadGraph MG5_aMC_v2_2_3 and then decayed it though the DECAY package from older version of MadGraph MadGraph5_v1_5_14 then I am getting the error [1].
>
> Please suggest how could I solve this.
>
> Also, When I tried to generate and decay events from the same version (MadGraph5_v1_5_14) of madgraph then it decayed it successfully.
>
> with regards,
> Ram
>
> [1]
> Wait....Writing Out Decayed Event file
>
> ----------------------------------------------
> Output events information
> ----------------------------------------------
>
> Output Event file : Output.lhe
> Number of Events : 50
> Integrated weight (pb) : 0.14861884999999994
> Max wgt : -0.99702762300000003
> Average wgt : 2.97237699999999870E-003
>
> At line 472 of file decay.f (unit = 16, file = 'Test.lhe')
> Fortran runtime error: Integer overflow while reading item 6
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,

Actually, I am working on the polarised ww scattering and as you know (also suggest by you at https://answers.launchpad.net/mg5amcnlo/+question/257782) we have to use the DECAY package to get event by event polarization information.

I have two w's and while decaying the second w I am facing the following error:

At line 472 of file decay.f (unit = 16, file = 'aux1.lhe')
Fortran runtime error: Integer overflow while reading item 6

This may be a typical fortran error and I do not have any experience in fortran so please suggest some solution.

I will highly appreciate your help.

Please find the attached decay.f file at link https://drive.google.com/file/d/0BxBy-RZ-_Z2ERXlyWjg3MGxiQ0U/view or you can see the patch of code below**:

**
 464 C Write out compulsory init info
 465 write(luni,'(a)') '<init>'
 466
 467 rewind(lunr)
 468 do while(index(buff,'</init') .eq. 0)
 469 read(lunr,'(a132)',err=99) buff
 470 if (index(buff,'<init') .eq. 0) cycle
 471 read(lunr,*) (idbmup(i),i=1,2),(ebmup(i),i=1,2),(pdfgup(i),i=1,2),
 472 $ (pdfsup(i),i=1,2),idwtup,nprup
 473 write(luni,90) (idbmup(i),i=1,2),(ebmup(i),i=1,2),(pdfgup(i),i=1,2),
 474 $ (pdfsup(i),i=1,2),idwtup,nprup
 475 do i=1,sizeievent
 476 read(lunr,*) xsecup,xerrup,xmaxup,lprup
 477 write(luni,91) sum(eventnr(lprup)),xerrup*sum(eventnr(lprup))/xsecup,
 478 $ maxwgt(eventnr(lprup)),lprup
 479 enddo
 480 enddo
 481 write(luni,'(a)') '</init>'

with regards,
Ram

Sounds a formatting problem.
You will need to add a couple of write statement to see at which line it fails to read the information and why.

One other suggestion is to put in the run_card “lhe_version=2”. This might turn off the formatting issue.

Cheers,

Olivier
On 09 Jun 2015, at 14:11, Ramkrishna <email address hidden> wrote:

> Question #267900 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/267900
>
> Status: Answered => Open
>
> Ramkrishna is still having a problem:
> Dear Olivier,
>
> Actually, I am working on the polarised ww scattering and as you know
> (also suggest by you at
> https://answers.launchpad.net/mg5amcnlo/+question/257782) we have to use
> the DECAY package to get event by event polarization information.
>
> I have two w's and while decaying the second w I am facing the following
> error:
>
> At line 472 of file decay.f (unit = 16, file = 'aux1.lhe')
> Fortran runtime error: Integer overflow while reading item 6
>
> This may be a typical fortran error and I do not have any experience in
> fortran so please suggest some solution.
>
> I will highly appreciate your help.
>
> Please find the attached decay.f file at link
> https://drive.google.com/file/d/0BxBy-RZ-_Z2ERXlyWjg3MGxiQ0U/view or you
> can see the patch of code below**:
>
> **
> 464 C Write out compulsory init info
> 465 write(luni,'(a)') '<init>'
> 466
> 467 rewind(lunr)
> 468 do while(index(buff,'</init') .eq. 0)
> 469 read(lunr,'(a132)',err=99) buff
> 470 if (index(buff,'<init') .eq. 0) cycle
> 471 read(lunr,*) (idbmup(i),i=1,2),(ebmup(i),i=1,2),(pdfgup(i),i=1,2),
> 472 $ (pdfsup(i),i=1,2),idwtup,nprup
> 473 write(luni,90) (idbmup(i),i=1,2),(ebmup(i),i=1,2),(pdfgup(i),i=1,2),
> 474 $ (pdfsup(i),i=1,2),idwtup,nprup
> 475 do i=1,sizeievent
> 476 read(lunr,*) xsecup,xerrup,xmaxup,lprup
> 477 write(luni,91) sum(eventnr(lprup)),xerrup*sum(eventnr(lprup))/xsecup,
> 478 $ maxwgt(eventnr(lprup)),lprup
> 479 enddo
> 480 enddo
> 481 write(luni,'(a)') '</init>'
>
>
> with regards,
> Ram
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,

I tried to generate events by changing "lhe_version=2". But I am getting the same error.

If possible, could you please try to have a look if I give you the test lhe file with say 50 events. Since, as I mentioned I am not familiar with fortran as well as I may not be able to understand the output of that.

Sorry but I really do not have enough time in a day to start to support code that I’m not an author of it
and who are not officially part of the distribution anymore.

Cheers,

Olivier
On 09 Jun 2015, at 15:16, Ramkrishna <email address hidden> wrote:

> Question #267900 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/267900
>
> Status: Answered => Open
>
> Ramkrishna is still having a problem:
> Dear Olivier,
>
> I tried to generate events by changing "lhe_version=2". But I am getting
> the same error.
>
> If possible, could you please try to have a look if I give you the test
> lhe file with say 50 events. Since, as I mentioned I am not familiar
> with fortran as well as I may not be able to understand the output of
> that.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,

This is just for information if someone else have this kind of problem you can suggest him/her the solution.

Actually the problem is with the format of <init> info thing in lhe file.

In older version looks like this:

<init>
     2212 2212 0.65000000000E+04 0.65000000000E+04 0 0 10042 10042 3 1
  0.74749182763E+00 0.77595808899E-03 0.14951800000E-04 0
</init>

where the numbers 10042 are of 5 characters long. While in newer version it is of 6 character long, which looks like

 <init>
      2212 2212 0.65000000000E+04 0.65000000000E+04 0 0 247000 247000 3 1
   0.73953136200E+00 0.89907684295E-02 0.13932600000E-01 0
 </init>

So, with old decay.f after processing this <init> patch merges the last three numbers, which was the source of problem. And it looks like this:

<init>
  2212 2212 0.65000000000E+04 0.65000000000E+04 0 0247000247000 3 1
  0.29760030000E-02 0.36180418023E-04 0.29760030000E-02 0
</init>

So, To fix it I just changed line number 483 from

90 FORMAT(2i6,2e19.11,2i2,2i6,i2,i3)

to

90 FORMAT(2i6,2e19.11,2i2,2i7,i2,i3)

I am able to solve this with the help of my CMS colleagues. Thanks to them.

with regards,
Ram

Excellent,

thanks,

Olivier
On 10 Jun 2015, at 14:56, Ramkrishna <email address hidden> wrote:

> Question #267900 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/267900
>
> Status: Answered => Solved
>
> Ramkrishna confirmed that the question is solved:
> Dear Olivier,
>
> This is just for information if someone else have this kind of problem
> you can suggest him/her the solution.
>
> Actually the problem is with the format of <init> info thing in lhe
> file.
>
> In older version looks like this:
>
> <init>
> 2212 2212 0.65000000000E+04 0.65000000000E+04 0 0 10042 10042 3 1
> 0.74749182763E+00 0.77595808899E-03 0.14951800000E-04 0
> </init>
>
> where the numbers 10042 are of 5 characters long. While in newer version
> it is of 6 character long, which looks like
>
> <init>
> 2212 2212 0.65000000000E+04 0.65000000000E+04 0 0 247000 247000 3 1
> 0.73953136200E+00 0.89907684295E-02 0.13932600000E-01 0
> </init>
>
> So, with old decay.f after processing this <init> patch merges the last
> three numbers, which was the source of problem. And it looks like this:
>
> <init>
> 2212 2212 0.65000000000E+04 0.65000000000E+04 0 0247000247000 3 1
> 0.29760030000E-02 0.36180418023E-04 0.29760030000E-02 0
> </init>
>
> So, To fix it I just changed line number 483 from
>
> 90 FORMAT(2i6,2e19.11,2i2,2i6,i2,i3)
>
> to
>
> 90 FORMAT(2i6,2e19.11,2i2,2i7,i2,i3)
>
> I am able to solve this with the help of my CMS colleagues. Thanks to
> them.
>
> with regards,
> Ram
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.