[ROOT] [TROOT] Making TChain and TTree friends

From: Caius Howcroft (howcroft@hep.phy.cam.ac.uk)
Date: Tue Apr 22 2003 - 15:33:40 MEST


Hi,


I have a chain (ch_reco) of 4 files that I run over with some MakeClass 
code to preoduce another tree of calculated variables (calctree).  I then want 
to make the original chain and the output tree.  So I do:

 ch_reco->AddFriend("calctree", "newtree.root") 
 
however when I try draw

ch_reco->Draw("newvar:oldvar")

thinigs look a little odd, it seem to be that then ch_reco gets to the end 
of the first file and instead of just carrying on calctree starts again 
from the first entry. Like so:

(to demonstrate this I copied the run/event number into the new tree then 
compaired with the old)

root [3] ch_reco->Scan("evthdr.run:new.run:evthdr.snarl:new.snarl", "")                                              
************************************************************
*    Row   * evthdr.ru *   new.run * evthdr.sn * new.snarl *
************************************************************
*        0 *     14218 *     14218 *        10 *        10 *
*        1 *     14218 *     14218 *        57 *        57 *
* 

ch_reco->Scan("evthdr.run:new.run:evthdr.snarl:new.snarl", 
"evthdr.run!=new.run || evthdr.snarl!=new.snarl")

************************************************************
*    Row   * evthdr.ru *   new.run * evthdr.sn * new.snarl *
************************************************************
*     5070 *     14220 *     14218 *         3 *        10 *
*     5071 *     14220 *     14218 *        10 *        57 *
* 


When I try to make ch_reco a friend of the new tree, root just segv's

I'm sure that people want to calculate new variables from an existing 
tree then plot the new and old  all the time, so there must be an easy 
way of doing this.

What am I doing wrong?


Caius
________________________________________
-           CAIUS HOWCROFT             -
-  +44 (0)1223 764 128      UK(office) -
========================================



This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:11 MET