Hi Dariusz,
Thanks for reporting this problem with a simple test case.
The problem is now fixed in the CVS version.
Rene Brun
On Mon, 18 Nov 2002, Dariusz Miskowiec wrote:
> Dear Root experts,
>
> I have the impression that performing a fit with option "E"
> affects not only the actual fit but also all fits which come
> after it. Particularly, FitSlices starts producing unreasonably
> small error bars for the second parameter. This is shown in
>
> http://www-linux.gsi.de/~misko/eoption.gif
>
> which was produced with root 3.03/09 under debian linux using
> enclosed macro. Many thanks and best regards
>
> Dariusz Miskowiec
>
> {
> gROOT->Reset();
> gStyle->SetPalette(1,0);
>
> TF1 *ga1 = new TF1("ga1","gaus",-4,4);
> TF2 *ga2 = new TF2("ga2","exp(-(y-200)*(y-200)/2/50/50)",-4,4,0,500);
> ga1->SetParameters(100.0,0.0,1.0);
> ga1->SetLineColor(2);
>
> TH1F *his1 = new TH1F("his1","one dim",100,-4,4);
> TH2F *his2 = new TH2F("his2","two dim",48,-4,4,50,0,500);
> his1->FillRandom("ga1",1000);
> his2->FillRandom("ga2",10000);
>
> TCanvas *c1 = new TCanvas("c1","E-option problem",10,10,500,400);
> c1->Divide(3,2);
>
> c1->cd(1); his1->Fit("ga1");
> c1->cd(2); his2->FitSlicesY();
> c1->cd(2); his2->Draw("colz"); gROOT->Get("his2_1")->Draw("same");
> c1->cd(3); gROOT->Get("his2_2")->Draw();
>
> c1->cd(4); his1->Fit("ga1","E");
> c1->cd(5); his2->FitSlicesY();
> c1->cd(5); his2->Draw("colz"); gROOT->Get("his2_1")->Draw("same");
> c1->cd(6); gROOT->Get("his2_2")->Draw();
> }
>
>
>
> --
> +-----------------------------------------------------------------+
> + Dariusz Miskowiec E-mail: D.Miskowiec@gsi.de +
> + GSI, Planckstr. 1 Phone: 0049-6159-712-133 +
> + D-64291 Darmstadt Fax: 0049-6159-712-989 +
> +-----------------------------------------------------------------+
>
This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:51:18 MET