madanalysis missing energy distribution

Hi

There is not particle defined with a PDG code 36 in the lhco format. If you check your event file, you will notice that all MET-related entries are associated with the PDG code 12. This is why you do not observe any difference in the distributions.

Regards,

Benjamin

On 07 Mar 2016, at 12:27 , Najimuddin Khan <email address hidden> wrote:

> New question #288011 on MadAnalysis 5:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> Hi,
>
> I have generated .lhco file for some BSM model using MG5_NLO . After importing .lhco file to madanalysis , I simply tried to draw some distributions. But I am having problem with MET plot.
>
> In the new model there is a new invisible particle ( pdg code say X).The process that I am considering has X as it's main source for MET. So, I use the following command for adding this particle in 'invisible' class :
>
> add invisible 36 -16 -14 -12 12 14 16
>
> I get some distribution . Then I remove 36 from definition of 'invisible' and redraw the plot. The plot is unchanged!!
>
> For further checking I issue the command :
> add invisible Y
>
> where Y is pdg of a particle which is not at all included in my model. So , I would expect no MET distribution. But instead of that I am getting the same plot!
>
> Could you please suggest what is wrong? Is there anything wrong in my definition?
>
> Regards,
> Najimuddin
>
>
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi,

Thanks a lot for the prompt reply. Should I add pdg code 36 in lhco file so that it could be treated as invisible particle?
Could u please specify where to add the pdg?
One more thing. I thought that If I use the command "define invisible" in normal mode it will automatically add 36 to invisible class.Is that not true?

Hi,

No, the LHCO format is fixed and cannot be changed. The event file that you are analyzing does not contain (in principle) any "36" particles. Those guys are included in the MET (PDG code 12), provided you have setup your detector simulator properly. Otherwise, I do not know what could have happened with them, but there is good chances that they are counted as jets.

Cheers,

Benjamin

On 07 Mar 2016, at 18:07 , Najimuddin Khan <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> Status: Answered => Open
>
> Najimuddin Khan is still having a problem:
> Hi,
>
> Thanks a lot for the prompt reply. Should I add pdg code 36 in lhco file so that it could be treated as invisible particle?
> Could u please specify where to add the pdg?
> One more thing. I thought that If I use the command "define invisible" in normal mode it will automatically add 36 to invisible class.Is that not true?
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,

So if I want the new particle to contribute to MET I should add it in MET class at madgraph level. (I am running delphes inside madgraph ). Is that correct?

And what is purpose of the "define invisible" command then ?

Thanks and regards,
Najimuddin

Hi Najimuddin

I think that you should add it at the delphes level, and probably at the pythia level as well.

Cheers,

Benjamin

On 08 Mar 2016, at 04:47 , Najimuddin Khan <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> Najimuddin Khan posted a new comment:
> Hi Benjamin,
>
> So if I want the new particle to contribute to MET I should add it in
> MET class at madgraph level. (I am running delphes inside madgraph ). Is
> that correct?
>
> And what is purpose of the "define invisible" command then ?
>
> Thanks and regards,
> Najimuddin
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                              I have a similar question. Suppose I am analyzing .hep files produced for a process which contains Dark matter Candidates. I am calling "Delphes" in MA5 normal mode through some commands in the script i.e.

set main.fastsim.package = delphes
set main.fastsim.detector = atlas

Then, where should I add the invisible particle ID ( in my case it is 35) ? Shall I add the PDG particle ID in the script itself ? Will it work ?

Regards,
Shiba

Hi Shiba,

You need to edit the delphes card, and add the properties of the DM particle into it. Please have a look how it is done for the neutralino (look for '1000022') and copy these lines, but with the pdg code fo your particle.

Cheers,

Benj

On 09 Mar 2016, at 06:22 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu posted a new comment:
> Hi Benjamin,
> I have a similar question. Suppose I am analyzing .hep files produced for a process which contains Dark matter Candidates. I am calling "Delphes" in MA5 normal mode through some commands in the script i.e.
>
> set main.fastsim.package = delphes
> set main.fastsim.detector = atlas
>
>
> Then, where should I add the invisible particle ID ( in my case it is
> 35) ? Shall I add the PDG particle ID in the script itself ? Will
> it work ?
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                           Thanks for the reply. Only I have to add the following line in the "Energy Fraction Block" in the Delphes Card i.e.

add EnergyFraction {35} {0.0 0.0}

Will it solve the job ? Do I need to edit anything else ?

regards,
Shiba

Hi,

This should work, provided the shower code is not erasing the "35" PDG code of the final state particle. You verify that you have such particles in the final state, please import the stdhep or hepmc fil in ma5 and try
  define dm = 35 -35
  plot N(35) [logY]
and check that your plot is not empty.

Cheers,

Benjamin

On 09 Mar 2016, at 08:17 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu posted a new comment:
> Hi Benjamin,
> Thanks for the reply. Only I have to add the following line in the "Energy Fraction Block" in the Delphes Card i.e.
>
> add EnergyFraction {35} {0.0 0.0}
>
>
> Will it solve the job ? Do I need to edit anything else ?
>
>
>
> regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                            I think it is plot N(dm) [logY] . To count, the dm particles, I am using the following script.

set main.fastsim.package = delphes
set main.fastsim.detector = atlas
import sample/2jplusMET/IDM/2jeth2h2_IDM_mHp150.hep.gz as sig
set sig.type = signal
set sig.backcolor = red
set sig.xsection = 0.2046
set main.stacking_method = superimpose
set main.lumi = 1000
define dm = 35 -35
select N(j)=2
plot N(dm)[logY]
plot MET
plot THT
submit

I am getting a non-empty plot. I am not sure whether the correct count plot is coming or not. In my case, the process contains two dm particles. But the plot is going up to "1" in the X-axis. Shall I send the plot to your e-mail so that you can tell where I am doing the mistake ?

NOTE : NOW I am not editing the Delphes Card here.

Regards,
Shiba

Hi,

You are right. It is indeed plot(dm). However, please split the detector simulation and the plotting in two steps.
  1. generation of the root files
   2. analysis of the root files.
And you must edit the delphes card.

Cheers,

Benjamin

On 09 Mar 2016, at 10:33 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu requested more information:
> Hi Benjamin,
> I think it is plot N(dm) [logY] . To count, the dm particles, I am using the following script.
>
> set main.fastsim.package = delphes
> set main.fastsim.detector = atlas
> import sample/2jplusMET/IDM/2jeth2h2_IDM_mHp150.hep.gz as sig
> set sig.type = signal
> set sig.backcolor = red
> set sig.xsection = 0.2046
> set main.stacking_method = superimpose
> set main.lumi = 1000
> define dm = 35 -35
> select N(j)=2
> plot N(dm)[logY]
> plot MET
> plot THT
> submit
>
> I am getting a non-empty plot. I am not sure whether the correct count plot is coming or not. In my case, the process contains two dm particles. But the plot is going up to "1" in the X-axis. Shall I send the plot to your e-mail so that you can tell where I am doing the mistake ?
>
> NOTE : NOW I am not editing the Delphes Card here.
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                               Sorry. I did not get your point. The steps you are suggesting to get the correct "MET" plot. right ??

Please Clarify your comment.

Regards,
Shiba

Hi,

What I am saying is that first, you need to take care of the detector simulations. Start from the StdHep or HepMC files and get to the delphes root files with the detector simulation executed. Next, you can analyze these root files and plot whater you want from them, in particular the MET plot.

Regards,

Benjamin

On 09 Mar 2016, at 10:47 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu posted a new comment:
> Hi Benjamin,
> Sorry. I did not get your point. The steps you are suggesting to get the correct "MET" plot. right ??
>
> Please Clarify your comment.
>
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                            Thanks for the reply. First I apologize for lengthening this thread. Another confusion in MET distribution. As you suggested, first I generated the root file.

1. In the first case, during generation of root file, I edited the delphes card (i.e. added the line "addEnergyFraction{35}{0.0 0.0} " in appropriate place). Then I have taken that root file and generated the histogram for "MET".

2. In the second case, without editing the Delphes Card, I have generated the root file . Then I have taken that root file and generated the histogram for "MET".

CONCLUSION : It is surprising that how in both cases, the "MET" plot is coming exactly same.

Am I doing anything wrong ??

Regards,
Shiba

This is indeed suspicious. Have you tested the presence of "35" particles at the level of the hep file?

Cheers,

Benjamin

On 09 Mar 2016, at 11:47 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu posted a new comment:
> Hi Benjamin,
> Thanks for the reply. First I apologize for lengthening this thread. Another confusion in MET distribution. As you suggested, first I generated the root file.
>
> 1. In the first case, during generation of root file, I edited the
> delphes card (i.e. added the line "addEnergyFraction{35}{0.0 0.0} " in
> appropriate place). Then I have taken that root file and generated the
> histogram for "MET".
>
>
> 2. In the second case, without editing the Delphes Card, I have generated the root file . Then I have taken that root file and generated the histogram for "MET".
>
>
> CONCLUSION : It is surprising that how in both cases, the "MET" plot is coming exactly same.
>
>
> Am I doing anything wrong ??
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                                 as earlier you suggested I checked the plot N(dm). I am getting a non-empty plot.

For the above two cases, that I was mentioning , no where I have included "35" in the invisible list(except in one case only edited the delphes card). Is this the problem ?

Regards,
Shiba

Hi,

You should include it in the invisible list for the hep case.

Cheers,

Benj

On 09 Mar 2016, at 12:07 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu posted a new comment:
> Hi Benjamin,
> as earlier you suggested I checked the plot N(dm). I am getting a non-empty plot.
>
> For the above two cases, that I was mentioning , no where I have
> included "35" in the invisible list(except in one case only edited the
> delphes card). Is this the problem ?
>
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                                  Are you saying to include the DM particle in the invisible list while generating the root file from .hep file ?
If yes, I am following the below script.

SCRIPT
===========

set main.fastsim.package = delphes
set main.fastsim.detector = atlas
import sample/2jplusMET/IDM/2jeth2h2_IDM_mHp150.hep.gz as sig
define invisible = 35 12 -12 14 -14 16 -16
submit

Regards,
Shiba

No. That's not correct. This is unrelated. You need to add it in the delphes card. As already said.

Cheers,

Benj

On 09 Mar 2016, at 12:52 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu posted a new comment:
> Hi Benjamin,
> Are you saying to include the DM particle in the invisible list while generating the root file from .hep file ?
> If yes, I am following the below script.
>
> SCRIPT
> ===========
>
> set main.fastsim.package = delphes
> set main.fastsim.detector = atlas
> import sample/2jplusMET/IDM/2jeth2h2_IDM_mHp150.hep.gz as sig
> define invisible = 35 12 -12 14 -14 16 -16
> submit
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi Benjamin,
                            During generation of root file, I edited the delphes card (i.e. added the line "addEnergyFraction{35}{0.0 0.0} " in appropriate place). Then I got a distribution for "MET".

 As earlier said, without editing the Delphes Card, I have generated the root file . Then I have taken that root file and generated the histogram for "MET".

But I got the two MET plots are same. surprisingly!!!

>>> You need to add it in the delphes card. As already said

are you talking any other thing to add for the invisible particle ?

Regards,
Shiba

Hi,

Sorry I was confused (too many e-mails in a short amount of time). To me, this means something wrong is done with the 35 particles at the shower/hadronization level. But you checked that the 35 particles were there in the hep file, didn't you. Therefore, I have no clue but there is definitely something weird somewhere.

Maybe should you ask the delphes authors directly (and please let me know)?

Cheers,

Benj

On 09 Mar 2016, at 13:57 , shibasipu <email address hidden> wrote:

> Question #288011 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/288011
>
> shibasipu posted a new comment:
> Hi Benjamin,
> During generation of root file, I edited the delphes card (i.e. added the line "addEnergyFraction{35}{0.0 0.0} " in appropriate place). Then I got a distribution for "MET".
>
> As earlier said, without editing the Delphes Card, I have generated the
> root file . Then I have taken that root file and generated the histogram
> for "MET".
>
> But I got the two MET plots are same. surprisingly!!!
>
>>>> You need to add it in the delphes card. As already said
>
> are you talking any other thing to add for the invisible particle ?
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Hi,
I was following the question. I am having exactly the similar problem. I asked the Delphes authors. They suggested to add the line "addEnergyFraction{x}{0.0 0.0}" in the delphes card.

I tried it for two different cases : adding the part mentioned above in delphes card and not adding the part.

The process that I am considering has "X" as its main source for MET. So, one would expect a softer plot for MET fot the second case (where "X" is not explicitly declared as "invisible" ) . But I am getting exactly same plot for both the cases !!!! This is totally unexpected.

Please help me to unsderstand what is going wrong?