Dear ROOTers,
I ask for a suggestion to solve a (maybe) silly problem in an
application of mine.
My problem is the following: I have a lot of files containing an event,
which all reconstructed tracks.
What I have to do is selecting the tracks which are useful for my
analysis and collect them into some kind of strudture, according to
their PID (specifically, 4 arrays: pi+, pi-, K+ and K-). This is needed
because I could repeat the successive analysis on them more than once,
so this will accelerate the computing operations.
What I tried to do is creating a TTree object whose branches are 4
TTrees, each one containing all interesting data (momentum, vertex, PID)
of each particle (one TTree for each kind of particle). Each of these
branches was a TTree with many branches of "primitive" data types (6
Double_t and 4 Int_t), which I created with a structure.
Unfortunately, when I try to read these data, I often encounter some
kinds of errors, or some memory leaks, and I thought that this could be
due to using a TTree as a branch object.
So, I would ask if it is completely wrong to do something like that (namely, using TTrees as branches of a higher-level TTree), or if I must do something at the initialization level, in order to work properly with this kind of structure.
Thanks a lot and best regards,
Alberto Received on Mon Dec 19 2005 - 18:21:56 MET
This archive was generated by hypermail 2.2.0 : Tue Jan 02 2007 - 14:45:14 MET