Re: [ROOT] TNtuple

From: Rene Brun (Rene.Brun@cern.ch)
Date: Fri May 12 2000 - 09:29:18 MEST


Hi Christian,
You do not say how your data were written (Fortran, C, C++ program?)
To read binary data from C++, you can use the iostream library
or directly the open,read, etc function. See for example
  man 2 open
or look at an example in the Root class TFile.
  http://root.cern.ch/root/html/src/TFile.cxx.html#TFile:SysOpen

If you want to store your data in Root format, you should create
a TTree object with one branch corresponding to your ADCs and one branch
corresponding to your wired chamber coordinates.
I cannot give more info at this point without knowing more on your
data structure. Fixed number of ADCs, coordinates, etc.

For creating a Tree, I suggest you look at the page recently announced
by Suzanne Panacek at URL:
http://www-pat.fnal.gov/root/Persistence/Persistence.htm
see also
http://www-pat.fnal.gov/root/

Rene Brun


On Thu, 11 May 2000, Christian D. HAEBERLI wrote:

> Dear Rooters
> 
> You helped me already a lot of times and I need your help once more.
> Thank you in advance.
> 
> This week I made some test measurements in T9 at CERN. I read out 50 ADC
> channels and  2 wired chambers (4 plains). I made runs of about 20'000
> evemts. 
> 
> Now I have to following data structure in binary files:
> data is structured in spills of about 400 events. First there all ADC
> values of one spill, last second are the values (coordinates)of the
> wired chambers of one spill.
> 
> Now, I have to:
> - read in binary data
> - match the events together 
> - make plots (wired chamber coordinates against ADC values)
> 
> My questions about this:
> -What's is the fastest way to read in binary data? The flags, the adc
> values, the wired chambers coordinates are all u_shorts in the DAQ
> software. So I have to read words of 2 bytes.
> - I can fill a TNtuple with  the ADC values and an other NTuple with the
> wired chambers coordinates. How can I match these NTuples together? 
> - Can I do plots out of two NTuples?
> - Or is TNtuple the wrong way to solve my problem?
> 
> 
> 
> Thank you very much for your help!!
> 
> Christian  
> 
> 
> 
> -- 
> Christian Haeberli                                          
> University of Bern
> Laboratory for High Energy Physics
> Sidlerstr. 5                                                    
> CH-3012 Bern
> 



This archive was generated by hypermail 2b29 : Tue Jan 02 2001 - 11:50:25 MET