GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung

Home » PANDA » PandaRoot » Event Generators » [CLOSED] Specific event selection in Dpm generator
[CLOSED] Specific event selection in Dpm generator [message #13048] Thu, 09 February 2012 10:25 Go to next message
donghee is currently offline  donghee
Messages: 385
Registered: January 2009
Location: Germnay
first-grade participant
From: *dip.t-dialin.net
Hello everyone,

I'm trying to select and save a specific event type from DPM generator.

For instance, I would like to select and save only pi+pi- event from DPM generation.
In PndDpmDirect I can simply select those event after pid checking for generated event type.
Then I do not add any tracks into output file, if this condition is not satisfied. That mean only pi+pi- event have track information, and other event types are set to zero tracks.

After this modification in FairPrimaryGenerator, PndDpmDirect file, I want to save only pi+pi- event without any zero track event, which is marked by my decision, into my final output file. Now, I have a trouble to discarding zero track from my output file.


I'm wondering, are there some possible and simple way to select and save specific event.
One general way is of course handling of source code dpm_gen_.f.
I am not prefer this way.
Second convenient way for me is control setting or option at FairPrimaryGenerator, PndDpmDirect, and FairMCEventHeader after MC generation from DPM.

Thank you for your reading!
Donghee




[Updated on: Sun, 09 March 2014 13:32] by Moderator

Report message to a moderator

Re: Specific event selection in Dpm generator [message #13049 is a reply to message #13048] Thu, 09 February 2012 11:27 Go to previous messageGo to next message
donghee is currently offline  donghee
Messages: 385
Registered: January 2009
Location: Germnay
first-grade participant
From: *dip.t-dialin.net
Hello everyone,

I solve this problem with quite simple trick using while loop only in PndDpmDirect code.(posted below)
It might be useful that someone prepare a selector function inside PndDpmDirect in the future, if it is not exsiting.
Anyhow, thank you again for your reading...

Quote:


// ----- Public method ReadEvent --------------------------------------
Bool_t PndDpmDirect::ReadEvent(FairPrimaryGenerator* primGen) {

int npart, i;
Double_t fX, fY, fZ, radius;
double Px[1000],Py[1000],Pz[1000],E[1000],Pm[1000],Wh[1000];
int Id[1000];

float Generator=0.; // Format in which events are produced (0=pythia, 1=pluto)

Double_t weight = 1.0;
Int_t activeCnt=0;


Int_t NIDs=0;
Int_t event_type = 844;
Int_t particle_number = 4;
double PID_type_1 = 211;
double PID_type_2 = -211;

//4pi = 844
//2pi = 422

//////////////////////////////////////////////////////////////////////// ///////////////////////Donghee Kang
//Selector!!!

while(1){

bool selector = false;
// run generator
dpm_gen__(&Generator, &fSeed);

// Loop over all produced particles
npart = lujets_.n;

if(npart == particle_number){//////////////////////////////////////////Donghee Kang
NIDs = 0;
for (i= 0; i< npart; ++i) {
Id[i]=lujets_.k[i+1000];
if( (Id[i] == PID_type_1) || (Id[i] == PID_type_2) ){
NIDs += TMath::Abs(Id[i]);
} //Id[i] check!!!
}//loop
if(NIDs == event_type){
for (i= 0; i< npart; ++i) {
Id[i]=lujets_.k[i+1000];
Px[i]=lujets_.p[i];
Py[i]=lujets_.p[i+1000];
Pz[i]=lujets_.p[i+2000];
Pm[i]=lujets_.p[i+4000];
E[i]=lujets_.p[i+3000];
Wh[i]=1.0;

/* Check if fGasmode is set */
fX = 0.;
fY = 0.;
fZ = 0.;
if (fGasmode == 1) {

// define position of track start
// Random 2D point in a circle of radius r (simple beamprofile)
radius = gRandom->Gaus(0,fRsigma);
gRandom->Circle(fX, fY, radius);

// calculate fZ according to some (probability) density function of the gas
fZ=fDensityFunction->GetRandom();

}

// add track
//printf("- I -: new particle at: %f, %f, %f ...\n", fX, fY, fZ);
primGen->AddTrack(Id[i], Px[i], Py[i], Pz[i], fX, fY, fZ);
}//loop
selector= true;

}// //////////////////////////////////////////
else{
selector = false;
}
}//npart == 4
if(selector){
//cout<<"Yes I have 4 pi+"<<endl;
break;
}
}//End of while
return kTRUE; //////////////////////////////////////////



}





Re: Specific event selection in Dpm generator [message #15371 is a reply to message #13049] Wed, 28 August 2013 18:28 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
Hello all,

we ran now into the same problem as Donghee did more than a year ago (in the sense that we also wanted to filter the events which are produced by the dpm generator) while performing a background study. A solution was found pretty fast for our specific problem, but I believe that it would make sense to have a more general solution. Therefore, I implemented a filter for PndDpmDirect which can handle pretty general requests from the simulation macro in the form:

Dpm->AddFilterMinMax( 1, 5, 11, -211 ); // request at least 1 and at most 5 e- OR pi-
Dpm->AddFilterMinMax( 1, 5, -11, 211 ); // AND request at least 1 and at most 5 e+ OR pi+
Dpm->AddFilterMinMax( 3, 9999, 22 ); // AND request at least 3 and at most 9999 gamma


The standard behaviour is the same as before (i.e. no event filtering).

You can tell PndDpmDirect how often it should try to find a suitable event before giving up:
Dpm->SetFilterMaxTries(99999);


and it can tell you at the end of the simulation macro how many events dpm simulated in total to get the number of filtered events that you wanted as well as how many events reached the limit of tries without success:

if( UseDpm ){
cout << Dpm->GetNumberOfSimulatedEvents() << " events were simulated in dpm\n";
cout << Dpm->GetNumberOfFilterFailedEvents() << " unsuccessful attempts to find an event that suits your filters\n\n";
}


The code is currently still in testing and if you agree I would like to upload it to the trunk once it was tested to be reliable.

------------
EDIT:
Here was a paragraph about a restriction of the code which is obsolete now.

The restriction that you can add a filter for a specific pdg code ONLY ONCE has been removed in versions later than 2013-10-21:
------------

I am looking forward to reading your feedback.

Kind regards,
Martin

[Updated on: Wed, 09 April 2014 11:12]

Report message to a moderator

Re: Specific event selection in Dpm generator [message #15372 is a reply to message #15371] Thu, 29 August 2013 00:15 Go to previous messageGo to next message
StefanoSpataro is currently offline  StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino
first-grade participant

From: 2.235.190*
Hi Martin,
I think it would be good to have such code not only for DPM, but also for all the generators.
In this sense, instead of having it as an extension of DpmDirect, it would be good to implement the method inside FairPrimaryGenerator, if of course Mohammad and Florian agree. What do you think?
Re: Specific event selection in Dpm generator [message #15373 is a reply to message #15372] Thu, 29 August 2013 11:10 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
Hello Stefano,

sure, it would certainly be possible to integrate it in FairPrimaryGenerator if the experts agree.

I can either provide the code that I wrote for PndDpmDirect as an inspiration for the experts or I would need the advice of at least one of them to help me integrate it in FairPrimaryGenerator without breaking any functionality of PandaRoot or possibly other frameworks. Ideally, one could implement it so that no changes to all derived classes would be necessary.

-------
EDIT:
After talking to Klaus Götzen and Andreas Pitka I implemented the possibility for an arbitrary number of conditions per pdg code and also a maximum number of particles per event.

Also a filter just on particles' charges would be desirable.
In addition, a minimum and maximum (total/transversal/z) momentum should be implemented (both in the laboratory system as well as in arbitrary center of mass systems).
-------

Let's see what Mohammad and Florian think about it.

Kind regards,
Martin

[Updated on: Mon, 21 October 2013 16:40]

Report message to a moderator

Re: Specific event selection in Dpm generator [message #15374 is a reply to message #15373] Thu, 29 August 2013 12:20 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
I have just uploaded a first working and tested implementation to my development folder in the svn repository. You can find the files here:

PndDpmDirectWithFilter

------
EDIT:
Note that I have not yet implemented all the functionality which is described in the EDIT regions.
------

[Updated on: Mon, 21 October 2013 16:40]

Report message to a moderator

Re: Specific event selection in Dpm generator [message #15583 is a reply to message #15374] Fri, 18 October 2013 15:03 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
I have uploaded an updated version of the code. Now one can ask for a minimum number and a maximum number. I have adjusted the example code snippets in this forum, too.
Re: Specific event selection in Dpm generator [message #15588 is a reply to message #15583] Mon, 21 October 2013 16:34 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
The restriction of only one condition per pdg code has been removed. The filter can handle multiple conditions for the same PDG code now. I have just uploaded a debugged, compiling and running version to my development branch (see above).
The momentum and geometry cuts are still missing. Other than that the filter is fairly complete. Smile
Re: Specific event selection in Dpm generator [message #15589 is a reply to message #15588] Mon, 21 October 2013 16:56 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
You can find more information and a presentation concerning the filter here:

http://panda-wiki.gsi.de/cgi-bin/view/Computing/Minutes21October2013
Re: Specific event selection in Dpm generator [message #15806 is a reply to message #15589] Wed, 22 January 2014 18:41 Go to previous messageGo to next message
Anastasia Karavdina is currently offline  Anastasia Karavdina
Messages: 76
Registered: May 2010
Location: Mainz, Germany
continuous participant

From: *kph.uni-mainz.de
Hi Martin!
In your talk on the latest collaboration meeting, you mentioned as missing option filtering on kinematics (theta and phi angles). Are you working on it? Or maybe those are already implemented? Rolling Eyes

Cheers,
Anastasia.
Re: Specific event selection in Dpm generator [message #15807 is a reply to message #15806] Wed, 22 January 2014 18:59 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
Hello Anastasia,

the filter project is ongoing, a master student (as a side-project) and I (as a "hobby") are working part-time on it and that item is on the todo list. However, due to time pressure from the FTS TDR and a learning curve for the student the progress is not very fast. If you like to add such functionality in the meantime I am happy to supply the latest version of the code. Wink

Kind regards,
Martin
Re: Specific event selection in Dpm generator [message #15808 is a reply to message #15807] Wed, 22 January 2014 19:05 Go to previous messageGo to next message
Anastasia Karavdina is currently offline  Anastasia Karavdina
Messages: 76
Registered: May 2010
Location: Mainz, Germany
continuous participant

From: *kph.uni-mainz.de
Thank you for prompt reply!
We also have have TDR to write and I have no idea how implement this feature. But I'll think about it.
Is version in your development brunch the latest one?
Re: Specific event selection in Dpm generator [message #15809 is a reply to message #15808] Wed, 22 January 2014 19:08 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
Hello Anastasia,

yes, it is the latest version. Smile

If you have questions about the existing implementation I am happy to provide you with more information / add more comments to the source code.

Kind regards,
Martin
Re: Specific event selection in Dpm generator [message #15910 is a reply to message #15809] Fri, 28 February 2014 17:35 Go to previous messageGo to next message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *physik.uni-giessen.de
Katja Kleeberg and I have just ported the filter code to a class and modified FairPrimaryGenerator to use it. The code is still under development and has not been tested rigorously yet, but if someone would like to have the code, please let me know.
Re: Specific event selection in Dpm generator [message #15934 is a reply to message #15910] Sun, 09 March 2014 12:15 Go to previous message
MartinJGaluska is currently offline  MartinJGaluska
Messages: 203
Registered: March 2010
Location: Germany
first-grade participant
From: *pools.arcor-ip.net
I would recommend to close this topic. We have a new implementation of an event filtering framework independet of PndDpmDirect. I created a new topic and would suggest to continue the discussion there. Here is the link:
https://forum.gsi.de/index.php?t=msg&goto=15933&#msg_15933
Previous Topic: EvtGen workshop
Next Topic: No old EVtGen anymore
Goto Forum:
  


Current Time: Thu Mar 28 09:43:07 CET 2024

Total time taken to generate the page: 0.01365 seconds