GSI Forum: Event Generators » EvtGen Lumi Theta-Distribution

Home » PANDA » PandaRoot » Event Generators » EvtGen Lumi Theta-Distribution

Show: Today's Messages :: Polls :: Message Navigator

EvtGen Lumi Theta-Distribution [message #10204]

Tue, 16 February 2010 11:50

Mathias Michel
Messages: 21
Registered: May 2009
Location: Mainz

occasional visitor

From: *kph.uni-mainz.de

Hi all,

when generating with EvtGen for the luminosity monitor, i'm especially interested in forward angles below 10mrad theta. The generated particles in this region tend to specific angles rather then make a smooth distribution i would expect.

As an example i plotted theta and the z-component of the momentum for pbarpsystem -> pi+ pi- at 15GeV beam. Theta is in 0.1mrad bins and Pz in 0.2MeV bins.

The precision of the EvtGen-output is much higher then the binning, so is there some way to 'tune' EvtGen for the forward direction or am i maybe doing something wrong?

Thanks
Mathias

Attachment: pi+pi-_15GeV.pdf
(Size: 15.42KB, Downloaded 553 times)

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10205 is a reply to message #10204]

Tue, 16 February 2010 12:02

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *to.infn.it

Hi,
are you plotting data from MCTrack or from the luminosity detector? Is it possible that it is related to some segmentation of the lumi? Or maybe it is effect of the grid size of the field maps, have you tried without magnetic field?
In theory the precision of EvtGen should be enough to give you a smooth distribution, in theory.

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10206 is a reply to message #10205]

Tue, 16 February 2010 12:36

Mathias Michel
Messages: 21
Registered: May 2009
Location: Mainz

occasional visitor

From: *kph.uni-mainz.de

thanks for the fast answer,

this is MC-data, i have a macro which reads the output-ascii-file of EvtGen the same way it's done in pandaroot.

I tried also to read the file with a c++ stream and calculated theta in different ways to check if i loose information somewhere, but that's why i also plotted the z of momentum which is directly from the EvtGen file.

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10207 is a reply to message #10206]

Tue, 16 February 2010 12:55

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *to.infn.it

You could try to modify the file:

pgenerators/EvtGen/EvtBenBase/EvtStdHep.cc

and set the precision from 8 to something higher (and then recompile EvtGen), just to see if it is only a problem of rounding or maybe something deeper.

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10208 is a reply to message #10207]

Tue, 16 February 2010 13:44

Mathias Michel
Messages: 21
Registered: May 2009
Location: Mainz

occasional visitor

From: *kph.uni-mainz.de

I tried precision 12 and get the same results.

I also looked at some events, the numbers have more digis behind but the first 8 digis stay the same.

Attachment: pi+pi-_15GeV_12.pdf
(Size: 15.42KB, Downloaded 507 times)

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10213 is a reply to message #10208]

Wed, 17 February 2010 10:20

Mathias Michel
Messages: 21
Registered: May 2009
Location: Mainz

occasional visitor

From: *kph.uni-mainz.de

Attached are the EvtGen-sourcefile and the macro i use to analyse the output.

The structure is also visible in generated decay's with pandaEvtGen. I just added a filter to have more statistics and faster simulation in forward direction.

Attachment: myEvtGen.cc
(Size: 4.76KB, Downloaded 463 times)
Attachment: anaVert.C
(Size: 3.74KB, Downloaded 481 times)

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10216 is a reply to message #10204]

Wed, 17 February 2010 11:20

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *to.infn.it

Let's start from the beginning.

How have you created your evtgen file? With the standard simpleEvtGen.cc, or with the old pandaEvtGen?
Could you please write exactly what have you run, how many events, which energy and so on?
In order to reproduce your problem.

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10218 is a reply to message #10216]

Wed, 17 February 2010 14:08

Mathias Michel
Messages: 21
Registered: May 2009
Location: Mainz

occasional visitor

From: *kph.uni-mainz.de

I use the simpleEvtGen, to reproduce do:

# ./simpleEvtGen pbarpSystem pp_Pi+Pi-.dec 1000000 15
# root -l anaVert.C

files are attached. Thanks for checking!

Attachment: anaVert.C
(Size: 3.75KB, Downloaded 512 times)
Attachment: pp_Pi+Pi-.dec
(Size: 0.06KB, Downloaded 502 times)

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10249 is a reply to message #10218]

Mon, 22 February 2010 08:53

Mathias Michel
Messages: 21
Registered: May 2009
Location: Mainz

occasional visitor

From: *kph.uni-mainz.de

if somebody is able to reproduce or not please tell me, in case i'm just imagine that there is something strange or using EvtGen wrong Smile

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10251 is a reply to message #10249]

Mon, 22 February 2010 12:26

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *to.infn.it

Hi,
I have followed what you have done, and I have almost the same plot as you, if I analyse the file with your macro.

I have run fast_sim to check the output. It seems each pz has a "spacing" of exactly 0.0005 GeV/c.
I have checked the output.evt file, but there all the numbers are not so spaced. For this reason, I think the problem stays not in EvtGen, but in our reading the ASCII file. Maybe there is some conversion/rounding which is cutting our momentum with a resolution of 500 keV -> even the angles suffer from this rounding.
At the moment I cannot find the source of such a problem, maybe somebody else could comment about.

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10268 is a reply to message #10251]

Tue, 23 February 2010 12:37

Mathias Michel
Messages: 21
Registered: May 2009
Location: Mainz

occasional visitor

From: *kph.uni-mainz.de

Thanks for the affirmation!
I tried already different way's of reading the file, so i am out of idea's here.

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #10312 is a reply to message #10268]

Thu, 25 February 2010 18:56

StefanoSpataro
Messages: 2736
Registered: June 2005
Location: Torino

first-grade participant

From: *to.infn.it

Hi,
I have done some tests.
The problems are coming from EvtGen. I have run 100k events of J/Psi, and I have seen that in the output files there are 14 events with a track with exactly Pz = 15.46610609!!!! (but different px/py).

This means that it does not depend on pandaroot but only in evtgen. It sounds weird also the fact that only in Pz this problem appears, while Px and Py have smooth distributions.

I could think that the problem is coming from the random generator, or... no idea!!

I would like to ask to people doing analysis if they have found this kind of problem in other channels (maybe a problem of the PHSP model?). You have only to take your primary tracks, plot Pz in a very narrow window and to check the distribution.

And, if somebody is using a random generator different from the EvtGen one, this check could be also helpful.

Waiting for some feedback...

Ste

Report message to a moderator

Re: EvtGen Lumi Theta-Distribution [message #13924 is a reply to message #10204]

Tue, 04 September 2012 11:00

Stefan Pflueger
Messages: 99
Registered: February 2012

continuous participant

From: *kph.uni-mainz.de

Hi,

I'm observing similar things, at least for larger beam momenta (in my case 8.9GeV). The t spectrum looks quite nasty, as it has periodic oscillation/spikes. Are there any news on this matter?

Ups. Just realized that this post was talking about EvtGen. I was referring to DPM. I will open a new thread.. https://forum.gsi.de/index.php?t=tree&th=3668&start=0&rid=17 90&S=52628bf8929831bf7aea080256a9ef18

Best regards,

Stefan

[Updated on: Mon, 17 September 2012 17:32]

Report message to a moderator

Previous Topic:	DPM theta cut off
Next Topic:	EvtGen update?

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Mon Jun 01 05:55:37 CEST 2026

Total time taken to generate the page: 0.00839 seconds