At the end of a fit, RooFit prints a summary table with parameter values and errors. The errors printed in this table are different from the stored errors when using SumW2Error, dramatically so for binned fits.
Why do the printed and stored values differ and which is correct? Moreover, for an unweighted dataset, shouldn’t SumW2Error have no effect?
Similar considerations apply for AsymptoticError.
Reproducer follows.
EDIT:
The excessive difference for binned fits seems to disappear when BatchMode is disabled. So the differences are smaller for unbinned or non-batch fits, though they are still present.
import ROOT as r
ws = r.RooWorkspace("workspace")
x = ws.factory("x[-10, 10]")
sig = ws.factory("Gaussian::sig(x, mu[-1, 1], s[0.1, 5])")
bkg = ws.factory("Chebychev::bkg(x, {c1[0.1, -1, 1]})")
shp = ws.factory("SUM::shp(Nsig[0, 200] * sig, Nbkg[0, 200] * bkg)")
data = shp.generate(r.RooArgSet(x))
ws.Import(data)
print("with SumW2Error:")
res = shp.fitTo(
data,
r.RooFit.Extended(),
r.RooFit.Save(),
r.RooFit.SumW2Error(True),
r.RooFit.Strategy(1),
r.RooFit.BatchMode(True),
)
res.Print()
print("without SumW2Error:")
res = shp.fitTo(
data,
r.RooFit.Extended(),
r.RooFit.Save(),
r.RooFit.SumW2Error(False),
r.RooFit.Strategy(1),
r.RooFit.BatchMode(True),
)
res.Print()
Output:
RooFit v3.60 -- Developed by Wouter Verkerke and David Kirkby
Copyright (C) 2000-2013 NIKHEF, University of California & Stanford University
All rights reserved, please read http://roofit.sourceforge.net/license.txt
[#1] INFO:ObjectHandling -- RooWorkspace::import(workspace) importing dataset shpData
with SumW2Error:
[#1] INFO:Minization -- createNLL: caching constraint set under name CONSTR_OF_PDF_shp_FOR_OBS_x with 0 entries
[#1] INFO:Minization -- RooMinimizer::optimizeConst: activating const optimization
**********
** 1 **SET PRINT 1
**********
**********
** 2 **SET NOGRAD
**********
PARAMETER DEFINITIONS:
NO. NAME VALUE STEP SIZE LIMITS
1 Nbkg 1.00000e+02 2.00000e+01 0.00000e+00 2.00000e+02
2 Nsig 1.00000e+02 2.00000e+01 0.00000e+00 2.00000e+02
3 c1 1.00000e-01 2.00000e-01 -1.00000e+00 1.00000e+00
4 mu 0.00000e+00 2.00000e-01 -1.00000e+00 1.00000e+00
5 s 2.55000e+00 4.90000e-01 1.00000e-01 5.00000e+00
**********
** 3 **SET ERR 0.5
**********
**********
** 4 **SET PRINT 1
**********
**********
** 5 **SET STR 1
**********
NOW USING STRATEGY 1: TRY TO BALANCE SPEED AGAINST RELIABILITY
**********
** 6 **MIGRAD 2500 1
**********
FIRST CALL TO USER FUNCTION AT NEW START POINT, WITH IFLAG=4.
START MIGRAD MINIMIZATION. STRATEGY 1. CONVERGENCE WHEN EDM .LT. 1.00e-03
FCN=-299.19 FROM MIGRAD STATUS=INITIATE 16 CALLS 17 TOTAL
EDM= unknown STRATEGY= 1 NO ERROR MATRIX
EXT PARAMETER CURRENT GUESS STEP FIRST
NO. NAME VALUE ERROR SIZE DERIVATIVE
1 Nbkg 1.00000e+02 2.00000e+01 2.01358e-01 6.00895e+00
2 Nsig 1.00000e+02 2.00000e+01 2.01358e-01 -6.00904e+00
3 c1 1.00000e-01 2.00000e-01 2.02430e-01 -5.01767e+00
4 mu 0.00000e+00 2.00000e-01 2.01358e-01 -4.22636e+00
5 s 2.55000e+00 4.90000e-01 2.01358e-01 -1.28338e+00
ERR DEF= 0.5
MIGRAD MINIMIZATION HAS CONVERGED.
MIGRAD WILL VERIFY CONVERGENCE AND ERROR MATRIX.
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=-303.213 FROM MIGRAD STATUS=CONVERGED 118 CALLS 119 TOTAL
EDM=7.49028e-05 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER STEP FIRST
NO. NAME VALUE ERROR SIZE DERIVATIVE
1 Nbkg 5.41047e+01 2.21287e+01 1.51558e-03 -6.00593e-03
2 Nsig 1.45931e+02 2.40308e+01 1.85824e-03 2.40501e-02
3 c1 2.09549e-01 3.36596e-01 3.51927e-03 1.96629e-02
4 mu 4.64253e-01 3.92685e-01 4.65228e-03 1.17958e-02
5 s 3.17954e+00 4.86056e-01 1.49199e-03 -1.38766e-02
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
5.002e+02 -4.460e+02 -6.546e-01 3.102e-01 -8.592e+00
-4.460e+02 5.921e+02 6.549e-01 -3.104e-01 8.594e+00
-6.546e-01 6.549e-01 1.181e-01 -7.654e-02 2.481e-02
3.102e-01 -3.104e-01 -7.654e-02 1.655e-01 -1.509e-02
-8.592e+00 8.594e+00 2.481e-02 -1.509e-02 2.396e-01
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.86612 1.000 -0.820 -0.085 0.034 -0.785
2 0.82931 -0.820 1.000 0.078 -0.031 0.721
3 0.55844 -0.085 0.078 1.000 -0.547 0.147
4 0.54770 0.034 -0.031 -0.547 1.000 -0.076
5 0.80050 -0.785 0.721 0.147 -0.076 1.000
**********
** 7 **SET ERR 0.5
**********
**********
** 8 **SET PRINT 1
**********
**********
** 9 **HESSE 2500
**********
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=-303.213 FROM HESSE STATUS=OK 31 CALLS 150 TOTAL
EDM=7.51899e-05 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER INTERNAL INTERNAL
NO. NAME VALUE ERROR STEP SIZE VALUE
1 Nbkg 5.41047e+01 2.22703e+01 3.03115e-04 -4.76816e-01
2 Nsig 1.45931e+02 2.41666e+01 3.71647e-04 4.77220e-01
3 c1 2.09549e-01 3.36686e-01 7.03855e-04 2.11114e-01
4 mu 4.64253e-01 3.92862e-01 9.30455e-04 4.82791e-01
5 s 3.17954e+00 4.88848e-01 2.98397e-04 2.59870e-01
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
5.067e+02 -4.528e+02 -6.466e-01 3.055e-01 -8.730e+00
-4.528e+02 5.990e+02 6.469e-01 -3.056e-01 8.734e+00
-6.466e-01 6.469e-01 1.181e-01 -7.668e-02 2.467e-02
3.055e-01 -3.056e-01 -7.668e-02 1.657e-01 -1.509e-02
-8.730e+00 8.734e+00 2.467e-02 -1.509e-02 2.424e-01
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.86799 1.000 -0.822 -0.084 0.033 -0.788
2 0.83147 -0.822 1.000 0.077 -0.031 0.725
3 0.55878 -0.084 0.077 1.000 -0.548 0.146
4 0.54833 0.033 -0.031 -0.548 1.000 -0.075
5 0.80309 -0.788 0.725 0.146 -0.075 1.000
[#1] INFO:Fitting -- RooAbsPdf::fitTo(shp) Calculating sum-of-weights-squared correction matrix for covariance matrix
**********
** 10 **SET ERR 0.5
**********
**********
** 11 **SET PRINT 1
**********
**********
** 12 **HESSE 2500
**********
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=-303.213 FROM HESSE STATUS=OK 31 CALLS 181 TOTAL
EDM=7.52462e-05 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER INTERNAL INTERNAL
NO. NAME VALUE ERROR STEP SIZE VALUE
1 Nbkg 5.41047e+01 2.22992e+01 6.06231e-05 -4.76816e-01
2 Nsig 1.45931e+02 2.41943e+01 7.43294e-05 4.77220e-01
3 c1 2.09549e-01 3.36704e-01 1.40771e-04 2.11114e-01
4 mu 4.64253e-01 3.92897e-01 1.86091e-04 4.82791e-01
5 s 3.17954e+00 4.89416e-01 5.96794e-05 2.59870e-01
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
5.081e+02 -4.542e+02 -6.450e-01 3.045e-01 -8.758e+00
-4.542e+02 6.004e+02 6.453e-01 -3.047e-01 8.762e+00
-6.450e-01 6.453e-01 1.182e-01 -7.671e-02 2.464e-02
3.045e-01 -3.047e-01 -7.671e-02 1.657e-01 -1.509e-02
-8.758e+00 8.762e+00 2.464e-02 -1.509e-02 2.430e-01
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.86836 1.000 -0.822 -0.083 0.033 -0.788
2 0.83191 -0.822 1.000 0.077 -0.031 0.725
3 0.55885 -0.083 0.077 1.000 -0.548 0.145
4 0.54845 0.033 -0.031 -0.548 1.000 -0.075
5 0.80361 -0.788 0.725 0.145 -0.075 1.000
[#1] INFO:Minization -- RooMinimizer::optimizeConst: deactivating const optimization
RooFitResult: minimized FCN value: -303.213, estimated distance to minimum: 7.52462e-05
covariance matrix quality: Full, accurate covariance matrix
Status : MINIMIZE=0 HESSE=0 HESSE=0
Floating Parameter FinalValue +/- Error
-------------------- --------------------------
Nbkg 5.4105e+01 +/- 2.25e+01
Nsig 1.4593e+02 +/- 2.44e+01
c1 2.0955e-01 +/- 3.44e-01
mu 4.6425e-01 +/- 4.07e-01
s 3.1795e+00 +/- 4.92e-01
without SumW2Error:
[#1] INFO:Minization -- createNLL picked up cached constraints from workspace with 0 entries
[#1] INFO:Minization -- RooMinimizer::optimizeConst: activating const optimization
**********
** 13 **SET PRINT 1
**********
**********
** 14 **SET NOGRAD
**********
PARAMETER DEFINITIONS:
NO. NAME VALUE STEP SIZE LIMITS
1 Nbkg 5.41047e+01 2.24807e+01 0.00000e+00 2.00000e+02
2 Nsig 1.45931e+02 2.44464e+01 0.00000e+00 2.00000e+02
3 c1 2.09549e-01 3.43702e-01 -1.00000e+00 1.00000e+00
4 mu 4.64253e-01 4.06999e-01 -1.00000e+00 1.00000e+00
5 s 3.17954e+00 4.91809e-01 1.00000e-01 5.00000e+00
**********
** 15 **SET ERR 0.5
**********
**********
** 16 **SET PRINT 1
**********
**********
** 17 **SET STR 1
**********
NOW USING STRATEGY 1: TRY TO BALANCE SPEED AGAINST RELIABILITY
**********
** 18 **MIGRAD 2500 1
**********
FIRST CALL TO USER FUNCTION AT NEW START POINT, WITH IFLAG=4.
START MIGRAD MINIMIZATION. STRATEGY 1. CONVERGENCE WHEN EDM .LT. 1.00e-03
FCN=-303.213 FROM MIGRAD STATUS=INITIATE 18 CALLS 19 TOTAL
EDM= unknown STRATEGY= 1 NO ERROR MATRIX
EXT PARAMETER CURRENT GUESS STEP FIRST
NO. NAME VALUE ERROR SIZE DERIVATIVE
1 Nbkg 5.41047e+01 2.24807e+01 2.58281e-01 -6.01700e-03
2 Nsig 1.45931e+02 2.44464e+01 2.82082e-01 2.39913e-02
3 c1 2.09549e-01 3.43702e-01 3.60410e-01 1.96540e-02
4 mu 4.64253e-01 4.06999e-01 5.00231e-01 1.17591e-02
5 s 3.17954e+00 4.91809e-01 2.09576e-01 -1.39610e-02
ERR DEF= 0.5
MIGRAD MINIMIZATION HAS CONVERGED.
MIGRAD WILL VERIFY CONVERGENCE AND ERROR MATRIX.
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=-303.213 FROM MIGRAD STATUS=CONVERGED 61 CALLS 62 TOTAL
EDM=2.58614e-05 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER STEP FIRST
NO. NAME VALUE ERROR SIZE DERIVATIVE
1 Nbkg 5.41106e+01 2.21198e+01 1.51410e-03 -8.71965e-03
2 Nsig 1.45896e+02 2.40153e+01 1.84218e-03 7.53948e-03
3 c1 2.08415e-01 3.36463e-01 3.50827e-03 -3.73930e-04
4 mu 4.63185e-01 3.92460e-01 4.62498e-03 -1.91327e-03
5 s 3.17989e+00 4.85880e-01 1.49120e-03 5.42675e-03
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
4.997e+02 -4.455e+02 -6.342e-01 2.976e-01 -8.582e+00
-4.455e+02 5.914e+02 6.342e-01 -2.977e-01 8.581e+00
-6.342e-01 6.342e-01 1.180e-01 -7.641e-02 2.438e-02
2.976e-01 -2.977e-01 -7.641e-02 1.653e-01 -1.486e-02
-8.582e+00 8.581e+00 2.438e-02 -1.486e-02 2.395e-01
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.86598 1.000 -0.819 -0.083 0.033 -0.785
2 0.82916 -0.819 1.000 0.076 -0.030 0.721
3 0.55778 -0.083 0.076 1.000 -0.547 0.145
4 0.54737 0.033 -0.030 -0.547 1.000 -0.075
5 0.80018 -0.785 0.721 0.145 -0.075 1.000
**********
** 19 **SET ERR 0.5
**********
**********
** 20 **SET PRINT 1
**********
**********
** 21 **HESSE 2500
**********
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=-303.213 FROM HESSE STATUS=OK 31 CALLS 93 TOTAL
EDM=2.62244e-05 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER INTERNAL INTERNAL
NO. NAME VALUE ERROR STEP SIZE VALUE
1 Nbkg 5.41106e+01 2.22608e+01 3.02820e-04 -4.76750e-01
2 Nsig 1.45896e+02 2.41505e+01 3.68436e-04 4.76830e-01
3 c1 2.08415e-01 3.36552e-01 7.01654e-04 2.09954e-01
4 mu 4.63185e-01 3.92635e-01 9.24996e-04 4.81586e-01
5 s 3.17989e+00 4.88661e-01 2.98240e-04 2.60018e-01
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
5.063e+02 -4.522e+02 -6.259e-01 2.927e-01 -8.719e+00
-4.522e+02 5.982e+02 6.260e-01 -2.927e-01 8.721e+00
-6.259e-01 6.260e-01 1.180e-01 -7.655e-02 2.423e-02
2.927e-01 -2.927e-01 -7.655e-02 1.655e-01 -1.486e-02
-8.719e+00 8.721e+00 2.423e-02 -1.486e-02 2.423e-01
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.86784 1.000 -0.822 -0.081 0.032 -0.787
2 0.83132 -0.822 1.000 0.075 -0.029 0.724
3 0.55812 -0.081 0.075 1.000 -0.548 0.143
4 0.54799 0.032 -0.029 -0.548 1.000 -0.074
5 0.80277 -0.787 0.724 0.143 -0.074 1.000
[#1] INFO:Minization -- RooMinimizer::optimizeConst: deactivating const optimization
RooFitResult: minimized FCN value: -303.213, estimated distance to minimum: 2.62244e-05
covariance matrix quality: Full, accurate covariance matrix
Status : MINIMIZE=0 HESSE=0
Floating Parameter FinalValue +/- Error
-------------------- --------------------------
Nbkg 5.4111e+01 +/- 2.23e+01
Nsig 1.4590e+02 +/- 2.42e+01
c1 2.0842e-01 +/- 3.37e-01
mu 4.6319e-01 +/- 3.93e-01
s 3.1799e+00 +/- 4.89e-01
Notice that with SumW2Error the printed error on Nbkg is 2.23e+01 while that stored in the result is 2.25e+01. Without SumW2Error, they are both 2.23e+01. A similar effect can be seen with AsymptoticError(True).
The differences are much worse with binned data:
datahist = r.RooDataHist("datahist", "datahist", data.get(), data)
res = shp.fitTo(
datahist,
r.RooFit.Extended(),
r.RooFit.Save(),
r.RooFit.SumW2Error(True),
r.RooFit.Strategy(1),
r.RooFit.BatchMode(True),
)
res.Print()
Output:
[#1] INFO:Minization -- createNLL picked up cached constraints from workspace with 0 entries
[#1] INFO:Minization -- RooMinimizer::optimizeConst: activating const optimization
**********
** 73 **SET PRINT 1
**********
**********
** 74 **SET NOGRAD
**********
PARAMETER DEFINITIONS:
NO. NAME VALUE STEP SIZE LIMITS
1 Nbkg 5.41849e+01 2.31459e+01 0.00000e+00 2.00000e+02
2 Nsig 1.45814e+02 2.31472e+01 0.00000e+00 2.00000e+02
3 c1 2.08299e-01 3.48104e-01 -1.00000e+00 1.00000e+00
4 mu 4.63687e-01 4.11892e-01 -1.00000e+00 1.00000e+00
5 s 3.17833e+00 5.28331e-01 1.00000e-01 5.00000e+00
**********
** 75 **SET ERR 0.5
**********
**********
** 76 **SET PRINT 1
**********
**********
** 77 **SET STR 1
**********
NOW USING STRATEGY 1: TRY TO BALANCE SPEED AGAINST RELIABILITY
**********
** 78 **MIGRAD 2500 1
**********
FIRST CALL TO USER FUNCTION AT NEW START POINT, WITH IFLAG=4.
START MIGRAD MINIMIZATION. STRATEGY 1. CONVERGENCE WHEN EDM .LT. 1.00e-03
FCN=-303.383 FROM MIGRAD STATUS=INITIATE 18 CALLS 19 TOTAL
EDM= unknown STRATEGY= 1 NO ERROR MATRIX
EXT PARAMETER CURRENT GUESS STEP FIRST
NO. NAME VALUE ERROR SIZE DERIVATIVE
1 Nbkg 5.41849e+01 2.31459e+01 2.66135e-01 9.55564e-02
2 Nsig 1.45814e+02 2.31472e+01 2.66149e-01 -3.64247e-02
3 c1 2.08299e-01 3.48104e-01 3.65158e-01 4.70599e-02
4 mu 4.63687e-01 4.11892e-01 5.07407e-01 3.40515e-02
5 s 3.17833e+00 5.28331e-01 2.25426e-01 1.83695e-01
ERR DEF= 0.5
MIGRAD MINIMIZATION HAS CONVERGED.
MIGRAD WILL VERIFY CONVERGENCE AND ERROR MATRIX.
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=-303.384 FROM MIGRAD STATUS=CONVERGED 61 CALLS 62 TOTAL
EDM=2.87246e-05 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER STEP FIRST
NO. NAME VALUE ERROR SIZE DERIVATIVE
1 Nbkg 5.40861e+01 2.19506e+01 1.51525e-03 -1.95993e-02
2 Nsig 1.45869e+02 2.38509e+01 1.83954e-03 -1.99599e-02
3 c1 2.05423e-01 3.36385e-01 3.51059e-03 1.31456e-03
4 mu 4.60414e-01 3.91325e-01 4.62417e-03 -6.22323e-03
5 s 3.17342e+00 4.81816e-01 1.49271e-03 1.34641e-02
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
4.920e+02 -4.375e+02 -7.076e-01 4.250e-01 -8.414e+00
-4.375e+02 5.831e+02 7.073e-01 -4.248e-01 8.409e+00
-7.076e-01 7.073e-01 1.179e-01 -7.607e-02 2.596e-02
4.250e-01 -4.248e-01 -7.607e-02 1.642e-01 -1.768e-02
-8.414e+00 8.409e+00 2.596e-02 -1.768e-02 2.354e-01
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.86398 1.000 -0.817 -0.093 0.047 -0.782
2 0.82672 -0.817 1.000 0.085 -0.043 0.718
3 0.55772 -0.093 0.085 1.000 -0.547 0.156
4 0.54680 0.047 -0.043 -0.547 1.000 -0.090
5 0.79798 -0.782 0.718 0.156 -0.090 1.000
**********
** 79 **SET ERR 0.5
**********
**********
** 80 **SET PRINT 1
**********
**********
** 81 **HESSE 2500
**********
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=-303.384 FROM HESSE STATUS=OK 31 CALLS 93 TOTAL
EDM=2.88088e-05 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER INTERNAL INTERNAL
NO. NAME VALUE ERROR STEP SIZE VALUE
1 Nbkg 5.40861e+01 2.20876e+01 3.03051e-04 -4.77026e-01
2 Nsig 1.45869e+02 2.39821e+01 3.67907e-04 4.76523e-01
3 c1 2.05423e-01 3.36470e-01 7.02118e-04 2.06896e-01
4 mu 4.60414e-01 3.91498e-01 9.24834e-04 4.78461e-01
5 s 3.17342e+00 4.84517e-01 2.98543e-04 2.57285e-01
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
4.983e+02 -4.440e+02 -7.011e-01 4.227e-01 -8.546e+00
-4.440e+02 5.897e+02 7.009e-01 -4.225e-01 8.543e+00
-7.011e-01 7.009e-01 1.180e-01 -7.621e-02 2.585e-02
4.227e-01 -4.225e-01 -7.621e-02 1.644e-01 -1.773e-02
-8.546e+00 8.543e+00 2.585e-02 -1.773e-02 2.381e-01
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.86583 1.000 -0.819 -0.091 0.047 -0.785
2 0.82886 -0.819 1.000 0.084 -0.043 0.721
3 0.55805 -0.091 0.084 1.000 -0.547 0.154
4 0.54741 0.047 -0.043 -0.547 1.000 -0.090
5 0.80054 -0.785 0.721 0.154 -0.090 1.000
[#1] INFO:Fitting -- RooAbsPdf::fitTo(shp) Calculating sum-of-weights-squared correction matrix for covariance matrix
**********
** 82 **SET ERR 0.5
**********
**********
** 83 **SET PRINT 1
**********
**********
** 84 **HESSE 2500
**********
COVARIANCE MATRIX CALCULATED SUCCESSFULLY
FCN=1171.67 FROM HESSE STATUS=OK 39 CALLS 132 TOTAL
EDM=268.711 STRATEGY= 1 ERROR MATRIX ACCURATE
EXT PARAMETER INTERNAL INTERNAL
NO. NAME VALUE ERROR STEP SIZE VALUE
1 Nbkg 5.40861e+01 1.18686e+01 6.06101e-05 -4.77026e-01
2 Nsig 1.45869e+02 9.38736e+00 7.50531e-03 4.76523e-01
3 c1 2.05423e-01 2.47582e-01 1.43232e-02 2.06896e-01
4 mu 4.60414e-01 1.57589e-01 1.88666e-02 4.78461e-01
5 s 3.17342e+00 2.42483e-01 6.09027e-03 2.57285e-01
ERR DEF= 0.5
EXTERNAL ERROR MATRIX. NDIM= 25 NPAR= 5 ERR DEF=0.5
1.417e+02 1.961e+00 3.799e-01 -1.359e-01 -7.537e-01
1.961e+00 8.845e+01 -8.241e-02 2.952e-02 1.651e-01
3.799e-01 -8.241e-02 6.265e-02 -1.659e-02 5.003e-03
-1.359e-01 2.952e-02 -1.659e-02 2.510e-02 -4.439e-04
-7.537e-01 1.651e-01 5.003e-03 -4.439e-04 5.900e-02
PARAMETER CORRELATION COEFFICIENTS
NO. GLOBAL 1 2 3 4 5
1 0.30377 1.000 0.018 0.128 -0.072 -0.261
2 0.09442 0.018 1.000 -0.035 0.020 0.072
3 0.44418 0.128 -0.035 1.000 -0.418 0.082
4 0.41915 -0.072 0.020 -0.418 1.000 -0.012
5 0.29742 -0.261 0.072 0.082 -0.012 1.000
[#1] INFO:Minization -- RooMinimizer::optimizeConst: deactivating const optimization
RooFitResult: minimized FCN value: 1171.67, estimated distance to minimum: 268.711
covariance matrix quality: Full, accurate covariance matrix
Status : MINIMIZE=0 HESSE=0 HESSE=0
Floating Parameter FinalValue +/- Error
-------------------- --------------------------
Nbkg 5.4086e+01 +/- 6.76e+01
Nsig 1.4587e+02 +/- 7.70e+01
c1 2.0542e-01 +/- 5.88e-01
mu 4.6041e-01 +/- 1.05e+00
s 3.1734e+00 +/- 1.40e+00
The printed error is 1.19e+01 and the stored error is 6.76e+01.