> The changing of the mean confuses people a little bit. If accept , that ROOT always use the most precise approach, then no confusion at all. Victor Victor M. Perevoztchikov perev@bnl.gov Brookhaven National Laboratory MS 510A PO Box 5000 Upton NY 11973-5000 tel office : 631-344-7894; fax 631-344-4206; ----- Original Message ----- From: "Andrei Kuznetsov" <Andrei.Kouznetsov@cern.ch> To: "Rene Brun" <Rene.Brun@cern.ch> Cc: "Andrei Kuznetsov" <Andrei.Kouznetsov@cern.ch>; "Root Mailinglist" <roottalk@cern.ch> Sent: Wednesday, September 15, 2004 12:36 PM Subject: Re: [ROOT] Mean calculation in TH1 > > { > gROOT->Reset(); > TCanvas *cr = new TCanvas("cr","Amplitude",200,10,800,600); > TH1D *h=new TH1D("Pixel","Pixel",13,0,13); > for(Int_t i=0;i<6;i++) h->Fill(10*gRandom->Rndm()); > cout<<"Mean befor SetRange ="<<h->GetMean()<<endl; > h->GetXaxis()->SetRange(0,12); > cout<<"Mean after SetRange ="<<h->GetMean()<<endl; > h->Draw(); > cr->Update(); > cr->Pop(); > } > > The changing of the mean confuses people a little bit. > > Best regards, > Andrei. > > > On Wed, 15 Sep 2004, Rene Brun wrote: > > > Andrei, > > > > The member fTsumwx is incremented in the TH1::Fill methods using the true > > value of the x coordinate. TH1::GetMean returns the mean using this value. > > > > However, when you zoom via TAxis::SetRange, ROOT recomputes the mean value > > using the bin contents and the center of each bin. > > > > I do not see any problems with this method. If you have evidence of the > > contrary, please send the shortest possible running script showing the > > problem. > > > > Rene Brun > > > > Andrei Kuznetsov wrote: > > > > > > Hello, > > > > > > it looks like ROOT calculates fTsumwx (and hence Mean of a histogram) in > > > TH1 incorrectly. > > > > > > But if you perform > > > your_histogram->GetXaxis()->SetRange(a,b); > > > after the filling (or "zoom" by hands) it works correctly. > > > > > > It looks likes in the first case it is not used the center of bin when > > > fTsumwx is calculated. > > > > > > Best regards, > > > Andrei. > > >
This archive was generated by hypermail 2b29 : Sun Jan 02 2005 - 05:50:09 MET