Calculating decay width produces segmentation fault

Hi Sebastian,

Indeed I face the same error on my laptop.
Thanks for reporting this buggy behavior.

Could you test the following fix:
modify the file Source/dsample.f
around the line 154:
            call graph_store
         endif
 99 enddo
c
c All done
c

add the following lines:
      tmean = 0d0
      trmean = 0d0
      tsigma = 0d0
      tdem = 0d0

Note that for 1>2, the easiest is to compute the width via an UFO extension.
If you create your UFO model with the development version of FR.
Then you will have the analytical expression of all 1->2 width.

In order to have the width, you can set them in the param_card on "Auto"
and MG will use the UFO information in order to have the width associated to your spectrum.

Cheers,

Olivier

On Jul 30, 2013, at 8:21 AM, Sebastian T. <email address hidden> wrote:

> New question #233253 on MadGraph5:
> https://answers.launchpad.net/madgraph5/+question/233253
>
> Hi,
>
> I produced and UFO model for an simplified SUSY model we are working on. Since there is no theoretical prediction for the decay width of the new particles I wanted to calculate the width with madgraph. Lets assume the particle is called X then I proceed to calculate the width with
>
> ./bin/mg5
> import model Simplified_Model
> generate X > all all
> output
> launch
>
> However for some decays I get an error message like.
> /ajob1: line 4: 22360 Segmentation fault ../madevent > $k < input_app.txt
> and no decay width can be calculated even though event production seems to have no problem with the decay.
> First I thought it has to do with my UFO model, however, I get the same error message if I try to calculate the W partial decay width into light quarks the standard model via
> generate w+ > u d~
>
> Is there something wrong with they way I calculate the widths? If i try to calculate the width into electron+neutrino it works without error message and I get a value.
> I am not sure what the $k in the error message is, in the W case I guess it is in SubProcesses/P0_wp_udx/G1/run1_app.log, which is an empty file.
>
> I am using Madgraph 1.5.11 with python 2.6.4.
>
>
> Best
> Sebastian
>
>
>
>
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Hi,

thanks for the fast answer, I could fix the problem by removing the matching in the run_card.
It fails if I have 1 = ickkw
but runs without any problem if I set it to 0. So I will keep it this way(I like it more than changing the code itself) and try the development version of feynrules soon. I guess you mean the beta 1.8?

Best
Sebastian