Hi root team,
I'm wondering about the status of plans for implementing MakeClass
support for Friend trees, and also the status of the "AddFriend By Key"
mechanism discussed on a separate thread:
http://root.cern.ch/root/roottalk/roottalk04/0854.html
Some collaborators on our experiment make extensive use of the MakeClass
mechanism, and since we store MC & reconstruction results on separate
ntuples,
are anxious to have the ability to run MakeClass on Friend trees so as to
facilitate the use of analyzing the results of the two trees in parallel.
Is the plan still to have these features implemented as part of the final
version of 4.00?
Thanks,
-Sue
Rene Brun wrote:
>Hi Sue,
>
>Yes, I agree that friends should be supported by MakeClass.
>This requires some work, in particular in case of MakeSelector.
>On our todo list.
>
>Rene Brun
>
>On Tue, 30
>Dec
>2003, Sue Kasahara wrote:
>
>
>
>>Hi roottalk,
>>I am trying to use TTree::MakeClass with a set of friend trees:
>>{
>>TFile* file = new TFile("ntuple.root","READ");
>>TTree* treeA = (TTree*)(file -> Get("NtpA"));
>>TTree* treeB = (TTree*)(file -> Get("NtpB"));
>>
>>treeA -> AddFriend(treeB);
>>treeA -> MakeClass();
>>}
>>I'm finding that the Class generated by MakeClass only
>>reflects the contents of treeA and not of its friend treeB.
>>Is there a way to set this up so that MakeClass will generate
>>a class with reference to both trees? Otherwise, is it possible
>>to modify the MakeClass method to support this?
>>We have a set of users that find the MakeClass method useful,
>>and would like to make the use of friend trees with MakeClass
>>as transparent as possible.
>>I am using root version 3.10/01 updated 12/9/03 with gcc 3.2.2
>>on rh linux v9.
>>Thanks for your help,
>>-Sue K.
>>
>>
>>
>>
>>
This archive was generated by hypermail 2b29 : Sun Jan 02 2005 - 05:50:08 MET