Re: [ROOT] test/MainEvent example - problems with Draw afterwards

From: Rene Brun (Rene.Brun@cern.ch)
Date: Fri Jun 13 2003 - 18:06:03 MEST


Hi Torsten,

Yes, things are clear. In case of split 1, you read the full TClonesArray
in memory at each event, the array is never destroyed.

In case of split=2, the TClonesArray is used as intended.
Objects of next event will overwrite objects of previous event.

Rene Brun

Torsten Harenberg wrote:
> 
> On Thu, 12 Jun 2003, Rene Brun wrote:
> 
> > Hi Torsten,
> >
> > I looked at your file and I understand why you get the problem.
> > You have specified split=1 in your class Event. You should specify
> > at least split=2 if you want to see each member of your TClonesArray
> > in a separate branch.
> > If you do T.Print(), you will see:
> >
> > *Br   47 :fgenTracks :
> > *Entries :   186059 : Total  Size=  557009404 bytes  File Size  =
> > 36544499
> > *Baskets :     8739 : Basket Size=      64000 bytes  Compression=  15.24
> >
> > Any traversal of this branch will load 577 Megabytes in memory !
> >
> 
> Thanks Rene,
> 
> for your quick answer: that worked and solved my problem. :)
> 
> Just for curiosity: is it clear to you why - in case of split=1 - the
> memory for every Draw() is not cleared after the plot has finished, while
> now -with split=2- I can Draw()  thousands of times without any increase
> in memory consumption?
> 
> Thanks again,
> 
>     Torsten
> 
> <><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
> <>                                                              <>
> <> Torsten Harenberg         harenberg@physik.uni-wuppertal.de  <>
> <> Bergische Universitaet                                       <>
> <> FB 8 - Physik             Tel.: +49 (0)202 439-2740          <>
> <> Gaussstr. 20              Fax : +49 (0)202 439-2811          <>
> <> 42097 Wuppertal                                              <>
> <>                                                              <>
> <><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>



This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:12 MET