No continius distribution for binned process

Hi,

Such behaviour are expected to happen at a point or another. Indeed this means that some of the channel of integration returns zero since they fail to find the point passing the (strong) cut that you applied.
After a given number of try the code assumes that the contribution is zero. Clearly in your case your cut are too strong to allow our code to find any point passing your cut in a reasonable amount of try.

If you want to increase this number this is the line 1230 of Source/dsample.f
      parameter (max_events=500000) !Maximum # events before get non_zero

Cheers,

Olivier

On 29 Jan 2015, at 10:46, Klaas Padeken <email address hidden> wrote:

> New question #261378 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/261378
>
> Hi,
> I have observed a strange behaviour at high pt (~TeV)
> For the sm pocess:
> generate p p > l vl
> If one produces multiple samples binned in ptl for high values of pt the normalisation is off by one orders of magnitude.
> The plot that shows this is here:
> https://mddp.rz.rwth-aachen.de/home/link/browser?l=0000000008l00058ZZo_NSna2Rd60R%3F%5EelaPgOShD%5E%40
> And the process with all the configurations is here:
> https://mddp.rz.rwth-aachen.de/home/link/browser?l=0000000008l00059%60eeIc%5DABZ%3D%3CN3%5Ee4B%3B%5D%5C6__8fM8
>
> This is also the case for an UFO model with a loaded sm.
> The pt bin above 2TeV gives the error:
> Survey return zero cross section.
>
> Best,
> Klaas
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.