Re: [ROOT] line width in eps files

From: Rene Brun (Rene.Brun@cern.ch)
Date: Fri Feb 08 2002 - 23:41:49 MET


Hi Hannes,

The algorithm scaling the line-width as a function of the pad height was
implemented to provide a constant aspect ratio independent of the pad
size.
However, I realize that this has side-effects with TLegend and Postscript.
I have deleted this algorithm from the current version in CVS.

Rene Brun


On Fri, 8 Feb 2002, Hannes Sakulin wrote:

> Hi,
> 
> I have the following problems with setting the linewidth of histograms when
> I print them to an eps and later print the eps on the printer (ROOT
> 3.02/07):
> 
> 1) On the screen and printout the linewidth of the histogram and of the line
> in the legend are completely different. The one in the legend is much
> bigger. The example below isllustrates this.
> 
> 2) The line width on the printout depends on the size of the pad in which I
> plot the histogram. If I change the height from 5. to 4.5 cm in the example
> below than the readable line of width 2 becomes a hairline in the printout.
> 
> Could you please give me a hint how I can better control the line width in
> the printout (e.g. independent of the pad size) and how I can get the same
> line width in the histogram and legend.
> 
> Best Regards,
>   Hannes Sakulin.
> 
> 
> 
> void test_lw()
> {
>   gStyle->SetPaperSize(16.5, 5.);
>   TCanvas *cv_ovl= new TCanvas("cv_ovl","cv_ovl",1);
>   cv_ovl->Size(16.5,5.); // change pad height here
>   cv_ovl->cd();
>   gPad->SetFillColor(0);
>   gPad->SetBorderSize(0);
>   gPad->SetBorderMode(0);
>   gPad->SetGrid(1,1);
>   gPad->SetTicks(1,0);
>   gPad->Draw();
>   gPad->Divide(2,1);
>   gPad->cd(1);
>   gStyle->SetOptStat(0);
>   gPad->SetLeftMargin(0.18);
>   gPad->SetRightMargin(0.07);
>   gPad->SetBorderSize(0);
> 
>   TH1F* hGMT = new TH1F("xx","xx",10,0,10);
>   hGMT->Fill(3.,70.);
>   hGMT->Fill(4.,80.);
>   hGMT->Fill(5.,90.);
>   TH1F* hGMT1 = new TH1F("yy","yy",10,0,10);
>   hGMT1->Fill(3.,72.);
>   hGMT1->Fill(4.,78.);
>   hGMT1->Fill(5.,93.);
> 
>   hGMT->SetMaximum(110.);
>   hGMT->SetMinimum(60.);
>   hGMT->SetLabelSize(0.07,"x");
>   hGMT->SetLabelSize(0.07,"y");
>   hGMT->SetLabelOffset(-0.005,"x");
>   hGMT->SetTitle("");
>   hGMT->GetXaxis()->SetTitleSize(0.07);
>   hGMT->GetXaxis()->SetTitleOffset(0.5);
>   hGMT->GetYaxis()->SetTitleSize(0.07);
>   hGMT->SetXTitle("#eta_{gen}    ");
>   hGMT->SetYTitle("Efficiency (%)    ");
> 
> 
>   hGMT->SetLineColor(1);
>   hGMT->SetLineWidth(3);
>   hGMT->DrawCopy();
> 
>   hGMT1->SetLineColor(1);
>   hGMT1->SetLineWidth(2);
>   hGMT1->DrawCopy("SAME");
> 
>   TLegend* lab = new TLegend(0.64,0.16,0.91,0.42);
>   lab->SetFillColor(0);
>   lab->AddEntry(hGMT,"test test","L");
>   lab->AddEntry(hGMT1,"test test","L");
>   lab->Draw();
> 
> }
> 
> 
> 
> 
> 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Hannes Sakulin
> CERN / EP                           Phone: +41 22 767 7372
> CH-1211 Geneva 23, Switzerland      Fax:   +41 22 767 8940
> E-mail: Hannes.Sakulin@cern.ch
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 



This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:50:41 MET