Merging default configuration

Hi Guillermo,

For your first issue, could you please compare with the merging plots automatically generated by madgraph? They should be similar to those returned by madanalysis 5. Could you pelase confirm this? Concerning your second concern, if the DJR plots are not smooth, this simply indicates that the choice of the merging parameters may be not appropriate. You may need to regenerate your backgrounds. Finally, the layout of the DJR plots can be changed by editing the layout/merging_plots.py file.

Regards,

Benjamin

On 14 Oct 2014, at 10:21 , Guillermo <email address hidden> wrote:

> New question #255702 on MadAnalysis 5:
> https://answers.launchpad.net/madanalysis5/+question/255702
>
> Dear ma5 team,
>
> I am simulating in madgraph the next two processes, which are background for an analysis I'm carrying out. (SM background and the model I am using is "sm-no_masses " )
>
> 1. generate p p > t t~, (t > b w+ , w+ > all all), (t~ > b~ w-, w- > all all) @0
> add process p p > tt~ j , (t > b w+ , w+ > all all), (t~ > b~ w-, w- > all all) @1
> add process p p > tt~ j j , (t > b w+ , w+ > all all), (t~ > b~ w-, w- > all all) @2
> add process p p > tt~ j j j , (t > b w+ , w+ > all all), (t~ > b~ w-, w- > all all) @3
>
> 3. generate p p > b b~ / a z h @0
> add process p p > b b~ j / a z h @1
> add process p p > b b~ j j / a z h @2
> add process p p > b b~ j j j / a z h @3
>
> The matching squeme is MLM, and I have chosen a xqcut = 20. Then after pyhia hadronisation the samples are imported to madanalysis 5_v1.1.11_patch1b. The first thing that I check is whether the differential jet rate (DJR) distribution is smooth.
>
> I proceed with the next in madanalysis
>
> ./bin/ma5 -H
> import ../samples/qcd.hep.gz as ttbar
> set main.merging.check = true
> set main.merging.njets = 5
> submit submit merging_test
>
> And at the end, I found that the distribution is not smooth (I am simulating 50k events at the Madgraph level and only 11k remains after pythia ). and in the plots cross_section vs log_10(DJR1) for both backgrounds the distribution begins at log_10(DJR1) = 1.2 approximately. However in the public talks, I saw that the distributions begins at log_10(DJR1) = 0.
>
> The same happens for cross_section vs log_10(DJRN) for n= 2,3,4 ...
>
> I have tried to modified the file called merging_configuration.py ( inside madanalysis/configuration) however I did not find any preference for the use any of the available clustering algorithms.
>
> Can someone help me with this issue? How to obtain a smooth diferential jet rate distribution? and why my plots begin around log_10(DJR1) = 1.2
>
> Thanks a lot
>
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hello,

 I'm have the same problem than you. I've done the same procedure but for different canals : ttbar + jets and Z + 4 jets in order to compare the results with the plots given in the talk of Benjamin last year ( https://madanalysis.irmp.ucl.ac.be/raw-attachment/wiki/Talks/fuks_nsusy.pdf ).

 The DJR plots from MadAnalysis and thus given by Madgraph in the end of the simulation are not the same at all. The ones from Madgraph are smoothy but the one from MadAnalysis are not smoothy and there is a big overlap between the contributions. I think the problem come from MadAnalysis.

 Do you have an idea where does the problem come from ?

 Regards,

 Simon

Hi Simon,

In the current version of MadAnalysis 5, the DJR routines are broken. We are working on a fix and let you know asap.

Cheers,

Benj

Hi Simon,

Could you please try with v.1.1.12beta and tell me whether the problem is fixed?

Cheers,

Benjamin