generating more than 50k events

Dear Goutam,

The problem with more than 50k event at pythia level is that the output file will be larger than 4Gb which means that this is a file which can not be handle in a 32bit computer/32 bit program.
If you use the latest version of madgraph and of pythia-pgs, everything should be define such that you should be able to write those files. But to avoid problem with those large files, I would always advise to avoid such large output file and make your analysis on a series of smaller file.

Cheers,

Olivier

On 13 Jan 2015, at 07:56, Goutam Das <email address hidden> wrote:

> New question #260597 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/260597
>
> Dear Madgraph Team,
> I need to have data points in the high invariant mass region for a process in LO merging. But with pythia running I can only generate upto 50K events (as indicated in run_card), this becomes impossible.
> Is the any way to generate more merged events (say 1 lac )?
> I have looked on this question (https://answers.launchpad.net/mg5amcnlo/+question/247589), but it was not helpful since I tried to generate 1lac merged events , everything went fine but it generates tag_1_pythia_events.hep file instead of tag_1_pythia_events.hep.gz and when I try to use this file for further analysis in pythia it fails and says the file is corrupted.
>
> Also I want to ask you if I can use multirun option for this purpose.
>
> Best Regards,
> Goutam
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                           Thanks for your quick response. The problem I am facing is that pythia is not running on 1lac merged events. I am using the latest MG and pythia-pgs and my machine is 64bit.
What I really need is to have data set in the higher invariant mass region as well. Doing analysis on several smaller files is not helpful in that sense. Any suggestion on how to get data points in the high invariant mass region?
Thanks for helping.

Best regards,
Goutam

Dear Goutam,

> What I really need is to have data set in the higher invariant mass region as well. Doing analysis on several smaller files is not helpful in that sense.

I do not understand why this is not helpful, it should.

> Any suggestion on how to get data points in the high invariant mass region?

I have in my todo list a way to bias the event generation to allow to have more point in part of the phase-space.
But this is not yet implemented (not even started actually).

Cheers,

Olivier

On 13 Jan 2015, at 13:21, Goutam Das <email address hidden> wrote:

> Question #260597 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/260597
>
> Status: Answered => Open
>
> Goutam Das is still having a problem:
> Dear Olivier,
> Thanks for your quick response. The problem I am facing is that pythia is not running on 1lac merged events. I am using the latest MG and pythia-pgs and my machine is 64bit.
> What I really need is to have data set in the higher invariant mass region as well. Doing analysis on several smaller files is not helpful in that sense. Any suggestion on how to get data points in the high invariant mass region?
> Thanks for helping.
>
> Best regards,
> Goutam
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                      Just to follow up, if I use multi_run and generate several 50k merged events and combine those at the level of analysis I should get data points in the higher invariant mass region as well. Is that right?

Also this point was not clear to me :
> If you use the latest version of madgraph and of pythia-pgs, everything should be define such that you should be able to write those files.
Is it possible to change in the code for 64-bit machine so that pythia can handle more than 50k events? If so, can you please tell me a simple way to do this?
Thanks for helping.

Many Regards,
Goutam

Dear Goutam,

> Just to follow up, if I use multi_run and generate several 50k merged events and combine those at the level of analysis I should get data points in the higher invariant mass region as well. Is that right?

Yes

> Also this point was not clear to me :
>> If you use the latest version of madgraph and of pythia-pgs, everything should be define such that you should be able to write those files.
> Is it possible to change in the code for 64-bit machine so that pythia can handle more than 50k events? If so, can you please tell me a simple way to do this?
> Thanks for helping.

In principle the latest version of pythia-pgs is able to handle such file. so I would advise to rerun “install pythia-pgs” to ensure that you have the latest version.

Cheers,

Olivier

On 13 Jan 2015, at 16:26, Goutam Das <email address hidden> wrote:

> Question #260597 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/260597
>
> Status: Answered => Open
>
> Goutam Das is still having a problem:
> Dear Olivier,
> Just to follow up, if I use multi_run and generate several 50k merged events and combine those at the level of analysis I should get data points in the higher invariant mass region as well. Is that right?
>
> Also this point was not clear to me :
>> If you use the latest version of madgraph and of pythia-pgs, everything should be define such that you should be able to write those files.
> Is it possible to change in the code for 64-bit machine so that pythia can handle more than 50k events? If so, can you please tell me a simple way to do this?
> Thanks for helping.
>
> Many Regards,
> Goutam
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                           I have installed the latest pythia-pgs on latest MG. I gave a run for 1lac events and pythia ran sucessfully. Here is the result:

     Cross-section : 14.78 +- 0.01366 pb
     Nb of events : 100000
     Matched Cross-section : 11.71 +- 0.02183 pb
     Nb of events after Matching : 79239

But in the Events/run_** folder I am getting the "tag_1_pythia_events.hep" file instead of "tag_1_pythia_events.hep.gz". And when I try to analyze it says corrupted. Can you help me how to fix this. I think in the "madevent_interface.py", misc.gzip is creating the problem. Is it possible to have the .hep file instead of hep.gz when pythia running?

Thanks for helping,

Best Regards,
Goutam

Dear Olivier,
                      Thanks for your suggestion. My latest gzip version is 1.6 and its not working. So I went for your second suggestion i.e. forbid misc.zip . Just one more quesion, if I do not want to run PGS/DELPHES further how much events/how large hep file can I generate? Is there a limit on that.

Best Regards,
Goutam

Hi,

The limit is that the size of the file should be writable in 64 bit. I have no idea how much event this means but clearly a very huge number (more than a billion).
You will face other problem before that (in particular madgraph will fail to generate such huge number of events/…)

Cheers,

Olivier

On 15 Jan 2015, at 02:47, Goutam Das <email address hidden> wrote:

> Question #260597 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/260597
>
> Goutam Das posted a new comment:
> Dear Olivier,
> Thanks for your suggestion. My latest gzip version is 1.6 and its not working. So I went for your second suggestion i.e. forbid misc.zip . Just one more quesion, if I do not want to run PGS/DELPHES further how much events/how large hep file can I generate? Is there a limit on that.
>
> Best Regards,
> Goutam
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                            Thank you so much for your help.

Best Regards,
Goutam

Thanks Olivier Mattelaer, that solved my question.