The printed scales in setscales.f are different from the scsle in events.lhe

Dear Muhammad,

Did you plot those scale with the weight information include in your plot?
If you do not include the weight in your plot, then they are no reason that the two plot agrees.

In top of that , they are a couple of potential technical problem, like how do you handle the various iteration? (Do you plot only the last one?)

Cheers,

Olivier

On 20 May 2015, at 05:21, Muhammad <email address hidden> wrote:

> New question #267161 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/267161
>
> Dear experts,
> there something in Madgraph puzzling me and I do not have an explanation for it!
> We are generating tZ events at LO using dynamic scaling according to the following equation
>
> $\mu= 4 \times \sqrt {m_{b}^{2} + P_{T,b}^{2}$,
> thus the minimum value of the scale should be $4\times m_{b}$, and this occurs when the $P_{T,b}=0$
>
> I put some printout in the setscales.f file and I generated 10K events, I plotted their scale distribution and it peaks clearly at $4\times m_{b}~ 20$.
>
> If I plot the scale distribution from the Madgraph output text files events.lhe or unweighted_events.lhe then they peak at $m_{b}~5$.
>
> So a factor of 4 is missing somewhere! and we have no explanation for it, we are afraid that we are doing something wrong!
>
> What do you think?
>
> Cheers
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Muhammad,

Did you plot those scale with the weight information include in your plot?
If you do not include the weight in your plot, then they are no reason that the two plot agrees.

In top of that , they are a couple of potential technical problem, like how do you handle the various iteration? (Do you plot only the last one?)

Cheers,

Olivier

On 20 May 2015, at 05:21, Muhammad <email address hidden> wrote:

> New question #267161 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/267161
>
> Dear experts,
> there something in Madgraph puzzling me and I do not have an explanation for it!
> We are generating tZ events at LO using dynamic scaling according to the following equation
>
> $\mu= 4 \times \sqrt {m_{b}^{2} + P_{T,b}^{2}$,
> thus the minimum value of the scale should be $4\times m_{b}$, and this occurs when the $P_{T,b}=0$
>
> I put some printout in the setscales.f file and I generated 10K events, I plotted their scale distribution and it peaks clearly at $4\times m_{b}~ 20$.
>
> If I plot the scale distribution from the Madgraph output text files events.lhe or unweighted_events.lhe then they peak at $m_{b}~5$.
>
> So a factor of 4 is missing somewhere! and we have no explanation for it, we are afraid that we are doing something wrong!
>
> What do you think?
>
> Cheers
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
thanks for the reply. Actually I am a ware of the weights, since I am using LO simulation then all events in unweighted_events.lhe have weight=1
BUT

in setscales.f they have different weight, since I can not access the weights there I did not include it.

However,
I am not comparing the two plots, but I am looking the lower edge of each histogram to read the minimum values.
It is very clear that we have a sharp edge at 4\times m_{b} and m{b} before and after. [weighting events can not shift the distribution]

Hi,

> in setscales.f they have different weight, since I can not access the
> weights there I did not include it.

Well then it is very difficult to make any claim.

In any case, I would bet that you have missed something somewhere. And that’s likely to be in your way to make the plot.

What I would advise is to scan the events and check that the renormalisation scale is set as expected.

Cheers,

Olivier

On 20 May 2015, at 11:21, Muhammad <email address hidden> wrote:

> Question #267161 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/267161
>
> Status: Answered => Open
>
> Muhammad is still having a problem:
> Dear Olivier,
> thanks for the reply. Actually I am a ware of the weights, since I am using LO simulation then all events in unweighted_events.lhe have weight=1
> BUT
>
> in setscales.f they have different weight, since I can not access the
> weights there I did not include it.
>
> However,
> I am not comparing the two plots, but I am looking the lower edge of each histogram to read the minimum values.
> It is very clear that we have a sharp edge at 4\times m_{b} and m{b} before and after. [weighting events can not shift the distribution]
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.