Matching/Merging with UNLOPS

Dear Soumyadip,

First of all, let me remind you that one cannot use the MG5_aMC interface to pythia8 to "automatically" shower the events. You have write your own pythia code that showers LHE events. There are 2 examples in the pythia8 example folder that deal with UNLOPS events: main88 and main89. Starting from those, and the information at http://home.thep.lu.se/Pythia/pythia82html/MatchingAndMerging.html it should be straight-forward.

Best regards,
Rikkert

Dear Rikkert,

Thanks, Rikkert for your reply.
So, please let me try to elaborate a bit on this. I am trying to set up the UNLOPS interface within CMS. In CMS it is two different steps to create the final output. One LHE step and then Showering step.

I have successfully completed the LHE step. But we don't have a default UNLOPS showering setup from Pythia8. I am trying to create that one. I am aware of the standalone Pythia8 setup. But I want to create a setup within CMS. So to create that setup I am trying to find the answers to the above-mentioned questions.

So, if you could give some info about that, it will be very helpful for me to work on it.

Thanks,
Soumyadip

Dear Soumyadip,

I'm not sure I understand what would be the difference between a standalone Pythia8 setup and a setup within CMS. I can only recommend about the standalone Pythia8 setup, since that's only what I know. And that part is rather well-explained on the webpage I linked as well as in the pythia8 main88 and main89 example files. I don't think I can help much more than that since I'm not at all familiar with the CMS framework.

best,
Rikkert

Dear Rikkert,

Ok, Thanks Rikkert for your suggestions. I will look into it and try to set it up within CMS.

Thanks Rikkert Frederix, that solved my question.

Hi Rikkert,

Sorry to open up the conversation again. I need some more information.
I want to produce a gridpack for the Drell-yan process for UNLOPS with icckw = 4 in the run card.
I am a bit confused with the process card settings.

[1]
generate p p > e+ e- [QCD] @0
add process p p > e+ e- j [QCD] @1
add process p p > e+ e- j j [QCD] @2

add process p p > e+ e- j [LOonly=QCD] @3
add process p p > e+ e- j j [LOonly=QCD] @4
add process p p > e+ e- j j j [LOonly=QCD] @5
*****************************************************

[2]
generate p p > e+ e- [QCD] @0
add process p p > e+ e- j [QCD] @1
add process p p > e+ e- j j [QCD] @2

add process p p > e+ e- j j j [LOonly=QCD] @3
******************************************************

Which setting is the correct one for producing the gridpack for the Drell-yan process.

Thanks,
Soumyadip

Dear Soumyadip,

As can be seen from the main89 in the examples directory from pythia8, it's [1].

Best,
Rikkert

Hi,

In main89 in the examples directory of pythia8, it expects 2 LHE files : one with NLO and one with LO.
But if I want to create a mixed gridpack of NLO and LO then it is also good to go with [1]?
Because in this case isn't there some double-counting with process ids 1 ,2 and 3 ,4 ?

Thanks,
Soumyadip

Thanks Rikkert Frederix, that solved my question.

Hi,

Sorry to open the thread again.
For UNLOPS we are using ickkw = 4 in the run card to set the nLO and npNLO attributes.

Recently, I came to know about this development branch which uses ickkw = 5 for UNLOPS [1].
Could you please let me know what is the primary difference between them?

[1]https://code.launchpad.net/~maddevelopers/mg5amcnlo/simple_unlops

Thanks,
Soumyadip

Thanks Rikkert Frederix, that solved my question.