Ed,
Thanks for the confirmation of the problem on your side.
TH1I.h has been added in version 3.10 (in CVS)
Rene Brun
Edward Moyse wrote:
>
> I'm sorry Rene, but I don't see how that can be it ... I'm using 2.96 which
> is less fussy about namespace, so I don't need to do those changes, but in
> any case.
>
> #include <TROOT.h>
> #include <TH1.h>
> #include <vector>
>
> class Analysis {
> private:
> std::vector<TH1S*>* m_particleLevelsHist;
> };
>
> compiles but :
>
> #include <TROOT.h>
> #include <TH1.h>
> #include <vector>
>
> class Analysis {
> private:
> std::vector<TH1I*>* m_particleLevelsHist;
> // ^^^^^^^
>
> };
>
> doesn't.
>
> However I have tracked down the source of my problem and confusion: the
> machine I was using to compile has
> ROOTSYS= /afs/cern.ch/sw/root/v3.05.05/rh73_gcc296/root
> whilst the machine I was browsing code on has:
> ROOTSYS=/afs/cern.ch/sw/root/v3.05.07/rh73_gcc296/root
> and I gather TH1I si a recent addition?
>
> I have no idea how this is possible, since both *should* be using the same
> .zshrc but I will investigate. Sorry about that though - I should have
> checked this first.
>
> Nevertheless, even in the latest version of ROOT(3.05/07) it appears that
> there's no TH1I.h file, like TH1S.h, TH1F.h etc. I know they don't do
> anything other than include TH1.h but it seems a bit inconsistent not to have
> a TH1I.h version too.
>
> Cheers,
>
> Ed
>
> On Monday 08 September 2003 2:26 pm, Rene Brun wrote:
> > Ed,
> >
> > The problem has nothing to do with TH1I, but with an incomplete declaration
> > of vector. Choose one of the two solutions:
> >
> > //solution1
> > ===========
> > #include <TROOT.h>
> > #include <TH1.h>
> > #include <vector>
> > using namespace std;
> >
> > class Analysis {
> > private:
> > vector<TH1S*>* m_particleLevelsHist;
> > };
> >
> > solution2
> > =========
> > #include <TROOT.h>
> > #include <TH1.h>
> > #include <vector>
> >
> > class Analysis {
> > private:
> > std::vector<TH1S*>* m_particleLevelsHist;
> > };
> >
> > Rene Brun
> >
> > Edward Moyse wrote:
> > > Okay. If you compile the following very, very simple example with :
> > > g++ -c -fPIC `root-config --cflags` test.cxx
> > > then it works.
> > >
> > > Change:
> > > vector<TH1S*>* m_particleLevelsHist;
> > > to:
> > > vector<TH1I*>* m_particleLevelsHist;
> > >
> > > then it doesn't and I get:
> > > test.cxx:8: `TH1I' was not declared in this scope
> > > test.cxx:8: parse error before `>'
> > >
> > > ---------------------------------------- test.cxx
> > > #include <TROOT.h>
> > > #include <TH1.h>
> > > #include <vector>
> > >
> > > class Analysis {
> > > private:
> > > vector<TH1S*>* m_particleLevelsHist;
> > > };
> > >
> > > On Monday 08 September 2003 1:38 pm, Rene Brun wrote:
> > > > Ed,
> > > >
> > > > TH1I is defined in TH1.h. I do not see any problems with your example.
> > > >
> > > > Send a short script reproducing the problem.
> > > >
> > > > Rene Brun
> > > >
> > > > Edward Moyse wrote:
> > > > > Hi,
> > > > >
> > > > > Can anyone explain why:
> > > > >
> > > > > #include <TH1.h>
> > > > > vector<TH1S*>* m_particleLevelsHist;
> > > > >
> > > > > works, but
> > > > > #include <TH1.h>
> > > > > vector<TH1I*>* m_particleLevelsHist;
> > > > > gives:
> > > > >
> > > > > In file included from Analysis.C:1:
> > > > > Analysis.h:70: `TH1I' was not declared in this scope
> > > > >
> > > > > I'd guess that TH1Is aren't supported somehow., especially since
> > > > > there's a TH1F header which just points to TH1.h, but there's no
> > > > > TH1I.h header doing the same thing.
> > > > >
> > > > > % ls $ROOTSYS/include/TH1*
> > > > > /afs/cern.ch/sw/root/v3.05.07/rh73_gcc296/root/include/TH1.h
> > > > > /afs/cern.ch/sw/root/v3.05.07/rh73_gcc296/root/include/TH1C.h
> > > > > /afs/cern.ch/sw/root/v3.05.07/rh73_gcc296/root/include/TH1D.h
> > > > > /afs/cern.ch/sw/root/v3.05.07/rh73_gcc296/root/include/TH1F.h
> > > > > /afs/cern.ch/sw/root/v3.05.07/rh73_gcc296/root/include/TH1K.h
> > > > > /afs/cern.ch/sw/root/v3.05.07/rh73_gcc296/root/include/TH1S.h
> > > > >
> > > > > But this page:
> > > > > http://root.cern.ch/root/html/TH1I.html
> > > > > definitely lists TH1Is as an available type of histogram.
> > > > >
> > > > > Cheers,
> > > > >
> > > > > Ed
This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:15 MET