Re: [ROOT] How to write TGraphErrors Object to a rootfile?

From: Rene Brun (Rene.Brun@cern.ch)
Date: Thu Feb 22 2001 - 18:05:33 MET


Hi Thomas,
I am afraid that you do not provide enough details on your procedure.
Anyhow, let me make a guess!

It is my impression that you do the following:
 - file->cd();
 - create some histograms (say TH1 *h1, *h2, etc)
 - create some graphs (say TGraph *gr1, *gr2, etc)
 - file->Write()

At this point, the file will contain all your histograms, but not the graphs.
The reason is that objects like TH1 and TTrees are automatically added to a list
of objects associated with the current directory. file->Write() will then write
all these objects to the file.
To save the TGraph objects to the file, you must have an explicit gr->Write()
for each graph. Note that, by default, TGraph are unnamed objects.
In the above example with gr1 and gr2, you can do:
 - gr1->SetName("gr1"); gr1->Write();
   gr2->SetName("gr2"); gr2->Write();
or
 - gr1->Write("gr1"); 
   gr2->Write("gr2");

Rene Brun


Thomas Feser wrote:
> 
>   Hi all,
> 
>   I'm generating a rootfile containing several histograms (in root-batch
> mode), which works fine. I simply generate them after a "file->cd();"
> command, after which I'm able to read them from the file again.
>   However, the same does not work with TGraph objects. Do I have to give a
> ->Draw() command first? I don't really have an idea.
>   _
>   |om
> 
> ------------------------------------------------------------------------
> -     Thomas Feser                                                     -
> -                                                                      -
> -     Institut fuer Physik                phone: +49-6131-39-25165     -
> -     ETAP/AMANDA                           fax: +49-6131-39-25169     -
> -     Staudingerweg 7 / 04-426                                         -
> -     D-55099 Universitaet Mainz                                       -
> -                                                                      -
> -     Thomas.Feser@uni-mainz.de                                        -
> -     http://iphcip1.physik.uni-mainz.de/~tfeser                       -
> ------------------------------------------------------------------------



This archive was generated by hypermail 2b29 : Tue Jan 01 2002 - 17:50:37 MET