reweight fails to find new param_card

[Michal Szleper]

I was trying to use the reweight command with MG5_aMC_v2_2_3, either by running directly

./bin/madevent reweight runs_smp1 -f

or in two steps, and I would always get this:

[...]
Command "reweight runs_smp1" interrupted with error:
InvalidCmd : original card and new card are identical
Please report this bug on https://bugs.launchpad.net/madgraph5

even though I had specified a valid param_card in my reweight_card.dat and the cards were not identical.

I had a rough look at the code and found that in reweight_interface.py, lines 258-260 could be replaced by a single line

files.cp(pjoin(self.me_dir, 'Cards', 'param_card.dat'), pjoin(self.me_dir, 'rw_me','Cards', 'param_card.dat'))

and then it seems to me everything was working.

[Olivier Mattelaer (olivier-mattelaer)]

Hi,

Thanks a lot,

Note that the version that you are using is more than a year old.
Concerning the re-weighting module various improvement have been done in the last 12 month:
1) in version 2.3.2. (August 2015) Full re-writting and use f2py for the code allowing faster and more stable code (and allow to have different model/…)
2) in version 2.4.0 (May 2016) The code was deeply modified to allow NLO re-weighting
3) in the next version 2.5.0 (September 2016) The code will be running on multi-core and not on a single core.
Finally the paper associated to the re-weighting (both LO and NLO) is now released: arXiv:1607.00763

Funny enough even if the code is radically different the bug is still present in the latest version but only if you have a relative path to the card.
Since this is due to a line/functionality which was not existing in 2.2.3.
So thanks to have allow to spot this.

Cheers,

Olivier

 status fixcommited

> On Aug 24, 2016, at 21:13, Michal Szleper <email address hidden> wrote:
>
> Public bug reported:
>
> I was trying to use the reweight command with MG5_aMC_v2_2_3, either by
> running directly
>
> ./bin/madevent reweight runs_smp1 -f
>
> or in two steps, and I would always get this:
>
> [...]
> Command "reweight runs_smp1" interrupted with error:
> InvalidCmd : original card and new card are identical
> Please report this bug on https://bugs.launchpad.net/madgraph5
>
> even though I had specified a valid param_card in my reweight_card.dat
> and the cards were not identical.
>
> I had a rough look at the code and found that in reweight_interface.py,
> lines 258-260 could be replaced by a single line
>
> files.cp(pjoin(self.me_dir, 'Cards', 'param_card.dat'),
> pjoin(self.me_dir, 'rw_me','Cards', 'param_card.dat'))
>
> and then it seems to me everything was working.
>
> ** Affects: mg5amcnlo
> Importance: Undecided
> Status: New
>
> --
> You received this bug notification because you are subscribed to
> MadGraph5_aMC@NLO.
> https://bugs.launchpad.net/bugs/1616621
>
> Title:
> reweight fails to find new param_card
>
> Status in MadGraph5_aMC@NLO:
> New
>
> Bug description:
> I was trying to use the reweight command with MG5_aMC_v2_2_3, either
> by running directly
>
> ./bin/madevent reweight runs_smp1 -f
>
> or in two steps, and I would always get this:
>
> [...]
> Command "reweight runs_smp1" interrupted with error:
> InvalidCmd : original card and new card are identical
> Please report this bug on https://bugs.launchpad.net/madgraph5
>
> even though I had specified a valid param_card in my reweight_card.dat
> and the cards were not identical.
>
> I had a rough look at the code and found that in
> reweight_interface.py, lines 258-260 could be replaced by a single
> line
>
> files.cp(pjoin(self.me_dir, 'Cards', 'param_card.dat'),
> pjoin(self.me_dir, 'rw_me','Cards', 'param_card.dat'))
>
> and then it seems to me everything was working.
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/mg5amcnlo/+bug/1616621/+subscriptions