RooFit Drawings

[atasattari]

Hi,
I’m producing these plots using PyROOT and RooFit (6.34.02).

The dashed lines are rendered poorly, and I couldn’t identify the issue. I wonder if this is a known issue and whether there’s a fix for it?
I’m sharing relevant snippet of my code:

**skipped a loop here**
frames[det,source]=w[f'ofl_{det}'].frame(Title=f" ",Bins=nbins,)

**skipped some code here**
w[f'pdf_{det}_{source}'].plotOn(frames[det,source],Components=f'exponential_{det}_eff',
                                LineWidth=10,
                                LineStyle = 10,
                                LineColor= 6,
                                NumCPU=1,
                                MoveToBack=False)
c = ROOT.TCanvas(det, det, 6000, 2400)
c.Draw()
pad1,pad2 = createHorizontalPads(c)
for pad, source in zip([pad1,pad2],['sig','bg']):
     inner_pads = createVerticalPads(pad,frames[det,source],pull_frames[det,source])

     frames[det,source].GetYaxis().SetTitleSize(0.1)
     frames[det,source].GetYaxis().SetLabelSize(0.1)
     frames[det,source].GetYaxis().SetTitleOffset(0.65)
     frames[det,source].GetYaxis().SetNdivisions(-205)
     frames[det,source].GetXaxis().SetLabelOffset(999);
     frames[det,source].GetXaxis().SetTitleOffset(999);
     frames[det,source].GetXaxis().SetNdivisions(508)            
     frames[det,source].GetYaxis().SetTitle('Counts')#           
     frames[det,source].GetXaxis().CenterTitle(True)
     frames[det,source].GetYaxis().CenterTitle(True)

     for pad_count, pad in enumerate(inner_pads):
          pt = pad.GetPrimitive('title');
          pt.SetTextSize(0.1);
          pt.SetX1NDC(0.48)
          pt.SetX2NDC(0.55)
          pt.SetY1NDC(0.94)
          pt.SetY2NDC(0.98)
                
          pad.cd()
          pad.Draw()
                
c.Draw()
C.SaveAs('---.png')
def createHorizontalPads(mother):
    mother.cd()
    # Upper histogram plot is pad1
    pad1 = ROOT.TPad('1', '1', 0, 0, 0.5, 1.0)
    set_margin(pad1, 0.00,0.0,0.0,0.)
    pad1.Draw()
    # Lower ratio plot is pad2
    mother.cd()  # returns to main canvas before defining pad2
    pad2 = ROOT.TPad('2','2', 0.5, 0., 1, 1)
    set_margin(pad2, 0.0,0.0,0.0,0.0)
    pad2.Draw()
    return pad1, pad2

def set_margin(mother, top,bottom,right,left):
    mother.SetTopMargin(top)  # joins upper and lower plot
    mother.SetBottomMargin(bottom)
    mother.SetRightMargin(right)
    mother.SetLeftMargin(left)

def createVerticalPads(mother,f1,f2):
    mother.cd()
    # Upper histogram plot is pad1
    pad1 = ROOT.TPad('1', '1', 0, 0.4, 1, 1.0)
    set_margin(pad1, 0.1,0.05,0.05,0.15)
    pad1.SetGridx()
    pad1.Draw()
    pad1.cd()
    f1.Draw()
    pad1.Draw()
    # Lower ratio plot is pad2
    mother.cd()  # returns to main canvas before defining pad2
    pad2 = ROOT.TPad('2','2', 0, 0.0, 1, 0.4)
    set_margin(pad2, 0.0,0.3,0.05,0.15)
    pad2.SetGridx()
    pad2.Draw()
    pad2.cd()
    f2.Draw()
    pad2.Draw()
    for sig in (-4,-2,0,2,4):
        line = get_axhline(pad2,sig)
        line.SetLineWidth(1)
        line.SetLineStyle(1)
        line.SetLineColor(ROOT.kBlack)
        line.DrawClone()
    return pad1, pad2

Thank you for any hint.

[dastudillo]

What do you mean by 'poorly"? What is wrong with them, exactly?

[atasattari]

Saving the figures as PDFs rather than PNGs fixed my issue. I attached a small section of the plot from a PDF file.

This is the same figure with similar settings except for the format of the saved file.

[couet]

Yes, PDF is vector graphics unlike png which is bitmap graphics therefore the definition is generally better with pdf. In particular it does not depend on the canvas resolution (size) unlike png.