Logo ROOT  
Reference Guide
rf110_normintegration.py
Go to the documentation of this file.
1 ## \file
2 ## \ingroup tutorial_roofit
3 ## \notebook
4 ##
5 ## Basic functionality: examples on normalization and integration of p.d.fs, construction
6 ## of cumulative distribution functions from monodimensional p.d.f.s
7 ##
8 ## \macro_code
9 ##
10 ## \date February 2018
11 ## \authors Clemens Lange, Wouter Verkerke (C++ version)
12 
13 from __future__ import print_function
14 import ROOT
15 
16 # Set up model
17 # ---------------------
18 
19 # Create observables x,y
20 x = ROOT.RooRealVar("x", "x", -10, 10)
21 
22 # Create p.d.f. gaussx(x,-2,3)
23 gx = ROOT.RooGaussian(
24  "gx", "gx", x, ROOT.RooFit.RooConst(-2), ROOT.RooFit.RooConst(3))
25 
26 # Retrieve raw & normalized values of RooFit p.d.f.s
27 # --------------------------------------------------------------------------------------------------
28 
29 # Return 'raw' unnormalized value of gx
30 print("gx = ", gx.getVal())
31 
32 # Return value of gx normalized over x in range [-10,10]
33 nset = ROOT.RooArgSet(x)
34 print("gx_Norm[x] = ", gx.getVal(nset))
35 
36 # Create object representing integral over gx
37 # which is used to calculate gx_Norm[x] == gx / gx_Int[x]
38 igx = gx.createIntegral(ROOT.RooArgSet(x))
39 print("gx_Int[x] = ", igx.getVal())
40 
41 # Integrate normalized pdf over subrange
42 # ----------------------------------------------------------------------------
43 
44 # Define a range named "signal" in x from -5,5
45 x.setRange("signal", -5, 5)
46 
47 # Create an integral of gx_Norm[x] over x in range "signal"
48 # ROOT.This is the fraction of of p.d.f. gx_Norm[x] which is in the
49 # range named "signal"
50 xset = ROOT.RooArgSet(x)
51 igx_sig = gx.createIntegral(xset, ROOT.RooFit.NormSet(xset), ROOT.RooFit.Range("signal"))
52 print("gx_Int[x|signal]_Norm[x] = ", igx_sig.getVal())
53 
54 # Construct cumulative distribution function from pdf
55 # -----------------------------------------------------------------------------------------------------
56 
57 # Create the cumulative distribution function of gx
58 # i.e. calculate Int[-10,x] gx(x') dx'
59 gx_cdf = gx.createCdf(ROOT.RooArgSet(x))
60 
61 # Plot cdf of gx versus x
62 frame = x.frame(ROOT.RooFit.Title("c.d.f of Gaussian p.d.f"))
63 gx_cdf.plotOn(frame)
64 
65 # Draw plot on canvas
66 c = ROOT.TCanvas("rf110_normintegration",
67  "rf110_normintegration", 600, 600)
68 ROOT.gPad.SetLeftMargin(0.15)
69 frame.GetYaxis().SetTitleOffset(1.6)
70 frame.Draw()
71 
72 c.SaveAs("rf110_normintegration.png")