Re: pointer defined but ROOT complain "not defined"

Marcelo Zimbres wrote:
> Hi wxie,
>
> >{
> > TFile f("oo.root");
> > h->Divide(h0);
> > h->Divide(h1);
> >}
>
> Since this is not valid C++ code, I think it would be a better idea to
> first get the hists from the file(I do not know why cint is not doing it
> itself):
>
> {
> TFile f("oo.root");
> h->Divide((TH1 *) f.Get("h0"));
> h->Divide((TH1 *) f.Get("h1"));
> }

Note that a similar thing should be done for h as well, since it should also not be defined in this scope, since it is created in the first scope

{

   TH1F* h = new TH1F("h", "", 1, 0, 1); ....
}
// h (should not be) not define

and thus is only valid there, not in the second scope.

For reasons like this I prefer to compile my macros, not use cint interpret them (or use PyRoot), since it tends to let though quite a few non-valid C++ constructs ...

Chris

>
> Cheers,
> Marcelo
>
> 2012/1/6 wxie <wxie_at_purdue.edu <mailto:wxie_at_purdue.edu>>
>
>
> One more input is that this issue happens only when the histograms
> are read from a file. The following macro have no problem to run:
>
>
> {
> TH1F* h = new TH1F("h", "", 1, 0, 1);
> h->Fill(0.5);
> h->Sumw2();
>
> TH1F* h0 = new TH1F("h0", "", 1, 0, 1);
> h0->Sumw2();
> h0->Fill(0.5);
>
> TH1F* h1 = new TH1F("h1", "", 1, 0, 1);
> h1->Sumw2();
> h1->Fill(0.5);
>
> h->Divide(h0);
> h->Divide(h1);
>
> }
>
>
>
> On 1/5/2012 5:40 PM, wxie wrote:
>
> I run the following macro producing a file "oo.root".
>
> {
> TFile f("oo.root", "recreate");
> TH1F* h = new TH1F("h", "", 1, 0, 1);
> h->Fill(0.5);
> h->Sumw2();
>
> TH1F* h0 = new TH1F("h0", "", 1, 0, 1);
> h0->Sumw2();
> h0->Fill(0.5);
>
> TH1F* h1 = new TH1F("h1", "", 1, 0, 1);
> h1->Sumw2();
> h1->Fill(0.5);
>
> h->Write();
> h0->Write();
> h1->Write();
> f.Close();
> }
>
> then I run the following macro:
> {
> TFile f("oo.root");
> h->Divide(h0);
> h->Divide(h1);
> }
>
> and got an error message:
>
> root.exe [1] h->Divide(h0);
> root.exe [2] h->Divide(h1);
> Error: Symbol h1 is not defined in current scope (tmpfile):1:
> *** Interpreter error recovered ***
>
> People usually bypass it by defining another pointer, e.g. the
> following macro:
> {
> TFile f("oo.root");
> h->Divide(h0);
> TH1F* h2 = h1;
> h->Divide(h2);
> }
>
> Is there a way not to redefine this new pointer?
>
> Thanks
> --Wei
>
>
>
>
>
>
>
>
>
Received on Fri Jan 06 2012 - 11:58:55 CET