Using fastjet on non-showered event files

Dear Jonathan,

The best way to deal with your issue is to implement the b-matching procedure by yourself, at the level of the analysis. You could decide to tag as a b-jet any jet that lies at a distance deltaR smaller than some threshold from a b-quark. Of course, this requires to use the code in its expert mode. This should be not too complicated by starting from the working directory generated by MA5. I hope this helps.

Regards,

Benjamin

> On 22 Jan 2021, at 11:55 , Jonathan Hermann <email address hidden> wrote:
>
> New question #695141 on MadAnalysis 5:
> https://answers.launchpad.net/madanalysis5/+question/695141
>
> Dear MadAnalysis5 team.
>
> I am trying to use the fastjet implementation in MadAnalysis5 in order to cluster the LHE file generated by MadGraph without running a Parton Shower.
>
> The process I am considering is p p > t t~ xd xd~ / a z w+ w- [QCD] with the DMsimp model and I am decaying the top-pair using MadSpin (both semileptonically). Now, I would like to use fastjet to determine whether the extra light jet that sometimes appears is a resolved jet or if it should be recombined with one of the b-jets.
>
> In principle this seems to work, however the problem that I encounter is that in the LHE file created by MadAnalysis the b-jets are not identified as b-jets but as light jets. As far as understood, this is because MadAnalysis only identifies a jet as a b-jet if it can be matched to a B-hadron. Howerver, since I am not running a Parton Shower there are obviously no B-hadrons.
> I do not want to use a Parton Shower here since I am also using other programs for which a matching to Partons Showers is not possible.
>
> Do you know if there is a work-around for this or whether I have just missed something that would solve this problem?
> Of course, I could compare the original LHE file with the clustered one but I would like to avoid that if possible.
>
> My input into MadAnalysis is the following:
>
> set main.fastsim.package = fastjet
> set main.fastsim.algorithm = antikt
> set main.fastsim.radius = 0.5
> set main.fastsim.ptmin = 0
> import /scratch/work/hermann/Eventfiles/DM_MadSpin_NLO/Events/run_01_decayed_1/events.lhe as DM
> set main.outputfile = clustered_events.lhe
> define invisible = invisible 52 -52
> set main.fastsim.exclusive_id = false
> submit
>
>
> Thank you in advance,
> Jonathan
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Dear Benjamin,

first of all, thanks for the response.
Unfortunately, I have never used the expert mode. I went through a couple of tutorials but it is still unclear to me how this should work.
How I can access both the b-quarks and the reconstructed jets in one analysis? Because as far as I understood, I can either access the particles if I do not run the clustering or the reconstructed jets if I run the clustering.

Also, I encountered a problem with the clustered LHEF. For some reason the final state positron appears twice in the file, once as a positron and once as a light jet. However, this does not happen for the final state muon. Do you have any idea why this happens?

Best regards,
Jonathan

Dear Jonathan

> Unfortunately, I have never used the expert mode. I went through a couple of tutorials but it is still unclear to me how this should work.
> How I can access both the b-quarks and the reconstructed jets in one analysis? Because as far as I understood, I can either access the particles if I do not run the clustering or the reconstructed jets if I run the clustering.

The normal mode is designed to take two different event types LHE and HEPMC where LHE is parton level events and HEPMC showered events. Hence fastjet is not available for parton level events through normal mode. Thats why Benjamin was suggesting using the expert mode. There you have access to everything but for what you are trying to do, I suggest using the fastjet directly in your analysis instead of using the ma5 modules for clustering. Since what you are doing is highly irregular, ma5 is not designed to ease your workflow there. To do this you will need to add fastjet libraries to the Makefile by yourself and then it will be available for usage through the analysis. You will have access to all the particles within your LHE file and then you can cluster them as you like.

> Also, I encountered a problem with the clustered LHEF. For some reason
> the final state positron appears twice in the file, once as a positron
> and once as a light jet. However, this does not happen for the final
> state muon. Do you have any idea why this happens?

I believe this is related to your generator. But are you sure that is the case? Their momentum might be the same it's possible.

Hope it helps
Cheers

—
Jack Y. Araz

> On Jan 26, 2021, at 10:36 AM, Jonathan Hermann <email address hidden> wrote:
>
> Question #695141 on MadAnalysis 5 changed:
> https://answers.launchpad.net/madanalysis5/+question/695141
>
> Status: Answered => Open
>
> Jonathan Hermann is still having a problem:
> Dear Benjamin,
>
> first of all, thanks for the response.
> Unfortunately, I have never used the expert mode. I went through a couple of tutorials but it is still unclear to me how this should work.
> How I can access both the b-quarks and the reconstructed jets in one analysis? Because as far as I understood, I can either access the particles if I do not run the clustering or the reconstructed jets if I run the clustering.
>
> Also, I encountered a problem with the clustered LHEF. For some reason
> the final state positron appears twice in the file, once as a positron
> and once as a light jet. However, this does not happen for the final
> state muon. Do you have any idea why this happens?
>
> Best regards,
> Jonathan
>
> --
> You received this question notification because you are an answer
> contact for MadAnalysis 5.

Dear Jack,

thank you for your help!
I actually ended up using fastjet as a standalone program and incorporated it into my own LHE file parser which now it seems to work the way I wanted it to.

Best regards,
Jonathan