This problem was solved by Matthieu,
He was mixing an old version of rootcint with the new ROOT libs.
Rene Brun
Matthieu Guillo wrote:
>
> Hi Rene
>
> I tried version 3.02/03 and at first I thought my problem was fixed.
> Namely the method "Draw" gave me the answers I expected. However, the
> following line fails and return a NULL pointer:
>
> TBranch* brInfo = tree->GetBranch("info");
>
> "info" is a c-structure. The tree was created using ROOT 3.00/06.
> Compiling my class using 3.00/06 set of libraries solved the problem.
> So it looks like that "GetBranch" method ahs some problem with the 3.02/03
> version of ROOT.
>
> Thanks
>
> On Mon, 19 Nov 2001, Rene Brun wrote:
>
> > Hi Matthieu,
> >
> > Please try 3.02/03 and let me know if the problem persists.
> >
> > Rene Brun
> >
> > Matthieu Guillo wrote:
> > >
> > > Hello Rooters
> > >
> > > I know this topic has already been discussed her but somehow I am missing
> > > a point.
> > > I have a tree which contains c-structures and TLorentzVectors. I would
> > > like to copy all branches of the tree to another one but only a few
> > > selected entries.
> > > So, based on examples in the tutorials, I did:
> > >
> > > gSystem->Load("libPhysics");
> > > TFile* f = new TFile("oldfile.root");
> > > TTree* tree_old = (TTree*) f->Get("tree");
> > >
> > > info = new info_t; // info_t is a c-structure
> > > kine = new kine_t; // kine_t is a c-structure
> > > tdist = new tdist_t; // tdist_t is a c-structure
> > >
> > > TLorentzVector* v4P1 = new TLorentzVector;
> > > TLorentzVector* v4P2 = new TLorentzVector;
> > >
> > > tree_old->SetBranchAddress("info", &info);
> > > tree_old->SetBranchAddress("kine", &kine);
> > > tree_old->SetBranchAddress("tdist", &tdist);
> > > tree_old->SetBranchAddress("p1", &v4P1);
> > > tree_old->SetBranchAddress("p2", &v4P2);
> > >
> > > tree_old->SetBranchStatus("*", 1);
> > >
> > > TFile *newfile = new TFile("newfile.root","recreate");
> > > TTree *newtree = tree_old->CopyTree("complicated cut");
> > > newtree->SetName("tree");
> > >
> > > newfile->Write();
> > > newfile->Close();
> > >
> > > This fails! I compared the values in the c-structures in the old (using
> > > the cut) and new tree, they are identical, so that's good. The
> > > number of evens in the old tree after cut is the same than the
> > > number of events in the new tree. However the components of the
> > > TLorentzVectors are differents. I built the old tree in non-split mode for
> > > the TLorentzVector using ROOT 3.00/06 on Linux Red Hat 6.2 and I am
> > > trying to make the subselection using ROOT 3.02/00 on the same machine,
> > > same OS.
> > >
> > > What am I doing wrong?
> > >
> > > Thanks
> > >
> > > Matthieu Guillo
> > > University of South Carolina
> > > Thomas Jefferson National Laboratory
> > > Office 71 trailer 16
> > > Phone: 757-269-5551
> >
>
> Matthieu Guillo
> University of South Carolina
> Thomas Jefferson National Laboratory
> Office 71 trailer 16
> Phone: 757-269-5551
This archive was generated by hypermail 2b29 : Tue Jan 01 2002 - 17:51:09 MET