TWiki
>
LHCb Web
>
MassFitting
(revision 12) (raw view)
Edit
Attach
---+ G-Fact Mass Fitting ---++ Validation of signal fraction fitter (Lars) To validate the signal fraction fitter (aka mass fitter) a series of data sets were generated with different mixes of signals and combinatoric background were generated. Each data set has approximately 1000 toy data sets with 100k events each. All data sets are located in /data/lhcb01/eklund/B2hhToyData/, in sub-directories as given below. ---+++ Mass Peaks from real data Mass peaks of the different signal classes from approximately 2 pb^-1 of data. The difference between the plots is the use of the K-pi separation. The cut on DLL_PIDK is done at 0 for these plots. [[%ATTACHURL%/B2hhMassPeaks.pdf][<img width="999" alt="B2hhMassPeaks.png" src="https://ppes8.physics.gla.ac.uk/twiki/pub/LHCb/MassFitting/B2hhMassPeaks.png" title="B2hhMassPeaks.png" height="773" />]] ---+++ Event mix: Bd -> π π, Bd -> K π and combinatoric background *Directory:* BdMix _561pipi_103Kpi/, 1000 data sets, 100k events each <table align="center" style="height: 150px; width: 600px;" border="0"> <thead> <tr valign="middle"> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle">True S.F.</td> <td align="center" valign="middle">Initial value</td> <td align="center" valign="middle">Mean [%]<br /></td> <td align="center" valign="middle">Sigma [%]<br /></td> <td align="center" valign="middle">Pull mean</td> <td align="center" valign="middle">Pull sigma</td> </tr> </thead> <tbody> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix1_Bd2PiPi_pull.png][Bd->π+π-]]</td> <td align="center" valign="middle">56.1</td> <td align="center" valign="middle">55</td> <td align="center" valign="middle">56.12</td> <td align="center" valign="middle">0.19</td> <td align="center" valign="middle">0.11+/-0.03</td> <td align="center" valign="middle">1.00+/-0.02</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix1_Bd2KplusPiminus_pull.png][Bd->K+π-]]</td> <td align="center" valign="middle">5.665</td> <td align="center" valign="middle">5</td> <td align="center" valign="middle">5.657</td> <td align="center" valign="middle">0.084</td> <td align="center" valign="middle">-0.12+/-0.03</td> <td align="center" valign="middle">1.03+/-0.02</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix1_Bd2KminusPiplus_pull.png][Bd->K-π+]]</td> <td align="center" valign="middle">4.635</td> <td align="center" valign="middle">5</td> <td align="center" valign="middle">4.619</td> <td align="center" valign="middle">0.078</td> <td align="center" valign="middle">-0.21+/-0.03</td> <td align="center" valign="middle">0.99+/-0.02</td> </tr> <tr> <td align="left" valign="middle">Combinatoric</td> <td align="center" valign="middle">33.6</td> <td align="center" valign="middle">35</td> <td align="center" valign="middle">33.60</td> <td align="center" valign="middle"> </td> <td align="center" valign="middle"> </td> <td align="center" valign="middle"></td> </tr> </tbody> </table> ---+++ Event mix: Bd -> π π, Bd -> K π, Bs -> K K and combinatoric background *Directory:* Mix_Bd2pipiKpiNBs2KK/ , 940 data sets, 100k events each <table align="center" style="height: 180px; width: 600px;" border="0"> <thead> <tr valign="middle"> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle">True S.F.</td> <td align="center" valign="middle">Initial value</td> <td align="center" valign="middle">Mean [%]<br /></td> <td align="center" valign="middle">Sigma [%]<br /></td> <td align="center" valign="middle">Pull mean</td> <td align="center" valign="middle">Pull sigma</td> </tr> </thead> <tbody> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix2_Bd2PiPi_pull.png][Bd->π+π-]]</td> <td align="center" valign="middle">24<br /></td> <td align="center" valign="middle">20</td> <td align="center" valign="middle">23.99</td> <td align="center" valign="middle">0.18</td> <td align="center" valign="middle">-0.04+/-0.03</td> <td align="center" valign="middle">1.01+/-0.03</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix2_Bd2KplusPiminus_pull.png][Bd->K+π-]]</td> <td align="center" valign="middle">11<br /></td> <td align="center" valign="middle">10</td> <td align="center" valign="middle">10.98</td> <td align="center" valign="middle">0.11</td> <td align="center" valign="middle">-0.16+/-0.03</td> <td align="center" valign="middle">1.01+/-0.02</td> </tr> <tr align="center" valign="middle"> <td align="left">[[%ATTACHURL%/Mix2_Bd2KminusPiplus_pull.png][Bd->K-π+]]</td> <td>9</td> <td>10</td> <td>8.98</td> <td>0.11</td> <td>-0.21+/-0.04</td> <td>1.05+/-0.03</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix2_Bs2KK_pull.png][Bs->K+K-]]</td> <td align="center" valign="middle">21<br /></td> <td align="center" valign="middle">25</td> <td align="center" valign="middle">21.04</td> <td align="center" valign="middle">0.14</td> <td align="center" valign="middle">0.27+/-0.03</td> <td align="center" valign="middle">1.00+/-0.03</td> </tr> <tr> <td align="left" valign="middle">Combinatoric</td> <td align="center" valign="middle">35</td> <td align="center" valign="middle">35</td> <td align="center" valign="middle">35.02</td> <td align="center" valign="middle"> </td> <td align="center" valign="middle"> </td> <td align="center" valign="middle"><br /></td> </tr> </tbody> </table> ---+++ Event mix: Bd -> π π, Bd -> K π, Bs -> K K, Bs -> K π and combinatoric background *Directory:* Mix_BdNBs2pipiKpiKK/ , 860 data sets, 100k events each <table align="center" style="height: 210px; width: 600px;" border="0"> <thead> <tr valign="middle"> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle">True S.F.</td> <td align="center" valign="middle">Initial value</td> <td align="center" valign="middle">Mean [%]<br /></td> <td align="center" valign="middle">Sigma [%]<br /></td> <td align="center" valign="middle">Pull mean</td> <td align="center" valign="middle">Pull sigma</td> </tr> </thead> <tbody> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix3_Bd2PiPi_pull.png][Bd->π+π-]]</td> <td align="center" valign="middle">20<br /></td> <td align="center" valign="middle">23</td> <td align="center" valign="middle">19.98</td> <td align="center" valign="middle">0.17</td> <td align="center" valign="middle">-0.12+/-0.03</td> <td align="center" valign="middle">0.98+/-0.03</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix3_Bd2KplusPiminus_pull.png][Bd->K+π-]]</td> <td align="center" valign="middle">13.75<br /></td> <td align="center" valign="middle">12</td> <td align="center" valign="middle">13.75</td> <td align="center" valign="middle">0.14</td> <td align="center" valign="middle">-0.01+/-0.03</td> <td align="center" valign="middle">0.99+/-0.02</td> </tr> <tr align="center" valign="middle"> <td align="left">[[%ATTACHURL%/Mix3_Bd2KminusPiplus_pull.png][Bd->K-π+]]</td> <td>11.25</td> <td>12</td> <td>11.24</td> <td>0.13</td> <td>-0.08+/-0.03</td> <td>0.97+/-0.03</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix3_Bs2KK_pull.png][Bs->K+K-]]</td> <td align="center" valign="middle">15<br /></td> <td align="center" valign="middle">12</td> <td align="center" valign="middle">15.03</td> <td align="center" valign="middle">0.14</td> <td align="center" valign="middle">0.21+/-0.03</td> <td align="center" valign="middle">0.99+/-0.03</td> </tr> <tr align="center"> <td align="left">[[%ATTACHURL%/Mix3_Bs2KplusPiminus_pull.png][Bs->K+π-]]</td> <td>5.56</td> <td>5</td> <td>5.56</td> <td>0.10</td> <td>0.01+/-0.04</td> <td>1.02+/-0.03</td> </tr> <tr align="center" valign="middle"> <td align="left">[[%ATTACHURL%/Mix3_Bs2KminusPiplus_pull.png][Bs->K-π+]]</td> <td>2.44</td> <td>5</td> <td>2.43</td> <td>0.07</td> <td>-0.11+/-0.03</td> <td>0.98+/-0.03</td> </tr> <tr> <td align="left" valign="middle">Combinatoric</td> <td align="center" valign="middle">32</td> <td align="center" valign="middle">35</td> <td align="center" valign="middle">35.02</td> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle"><br /></td> </tr> </tbody> </table> ---+++ Event mix: Bd -> π π, Bd -> K π, Bs -> K K, Bs -> K π, Bd -> π π π and combinatoric background *Directory:* fullMFMixAug2010/ , 1000 data sets, 100k events each <table align="center" style="height: 240px; width: 600px;" border="0"> <thead> <tr valign="middle"> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle">True S.F.</td> <td align="center" valign="middle">Initial value</td> <td align="center" valign="middle">Mean [%]<br /></td> <td align="center" valign="middle">Sigma [%]<br /></td> <td align="center" valign="middle">Pull mean</td> <td align="center" valign="middle">Pull sigma</td> </tr> </thead> <tbody> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix4_Bd2PiPi_pull.png][Bd->π+π-]]</td> <td align="center" valign="middle">8.47<br /></td> <td align="center" valign="middle">10</td> <td align="center" valign="middle">8.45</td> <td align="center" valign="middle">0.13</td> <td align="center" valign="middle">-0.13+/-0.03</td> <td align="center" valign="middle">1.02+/-0.02</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix4_Bd2KplusPiminus_pull.png][Bd->K+π-]]</td> <td align="center" valign="middle">17.82<br /></td> <td align="center" valign="middle">15</td> <td align="center" valign="middle">17.84</td> <td align="center" valign="middle">0.14</td> <td align="center" valign="middle">0.12+/-0.03</td> <td align="center" valign="middle">0.99+/-0.03</td> </tr> <tr align="center" valign="middle"> <td align="left">[[%ATTACHURL%/Mix4_Bd2KminusPiplus_pull.png][Bd->K-π+]]</td> <td>14.58</td> <td>15</td> <td>14.58</td> <td>0.13</td> <td>-0.02+/-0.03</td> <td>1.01+/-0.03</td> </tr> <tr> <td align="left" valign="middle">[[%ATTACHURL%/Mix4_Bs2KK_pull.png][Bs->K+K-]]</td> <td align="center" valign="middle">8.47<br /></td> <td align="center" valign="middle">10</td> <td align="center" valign="middle">8.47</td> <td align="center" valign="middle">0.10</td> <td align="center" valign="middle">0.01+/-0.03</td> <td align="center" valign="middle">1.02+/-0.03</td> </tr> <tr align="center"> <td align="left">[[%ATTACHURL%/Mix4_Bs2KplusPiminus_pull.png][Bs->K+π-]]</td> <td>1.62</td> <td>1</td> <td>1.62</td> <td>0.07</td> <td>-0.03+/-0.04</td> <td>0.96+/-0.03</td> </tr> <tr align="center" valign="middle"> <td align="left">[[%ATTACHURL%/Mix4_Bs2KminusPiplus_pull.png][Bs->K-π+]]</td> <td>0.72</td> <td>1</td> <td>0.71</td> <td>0.05</td> <td>-0.32+/-0.03</td> <td>1.01+/-0.02</td> </tr> <tr align="center" valign="middle"> <td align="left">[[%ATTACHURL%/Mix4_Bd2PiPiPi_pull.png][Bd->π+π-π0]]</td> <td>15.0</td> <td>10</td> <td>14.98</td> <td>0.16</td> <td>-0.11+/-0.03</td> <td>1.02+/-0.02</td> </tr> <tr> <td align="left" valign="middle">Combinatoric</td> <td align="center" valign="middle">33.32</td> <td align="center" valign="middle">38</td> <td align="center" valign="middle">35.02</td> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle"><br /></td> <td align="center" valign="middle"><br /></td> </tr> </tbody> </table> ---+++ Sensitivity to number of events A series of fits to the data set Mix_Bd2pipiKpiNBs2KK (above) was done using a decreasing number of events, from 100 000 to 100 events. The initial values and fit configuration was identical to the fit described above. The results are summarised in the following graphs. * The error from the fit as reported by MINUIT versus number of events is show in this [[%ATTACHURL%/FitErrorVsNEvents.pdf][graph]]. As expected, the statistical error from the fit follows well the 1/sqrt(N) law. The unit here is % signal fraction, hence 1% error means that the signal fraction Bd -> pi pi is estimated to be 24 +/- 1 %. * The mean of the pull distribution is shown in this [[%ATTACHURL%/MeanOfPullVsNEvents.pdf][graph]]. The mean is less that 0.1 apart from the fits with the largest number of events. Since the statistical error is very small (<2*10^-3 for 100k evt) even the smallest bias is seen. At 100k event the observed absolute bias is < 4*10^-4. * The bias in absolute numbers is shown in this [[%ATTACHURL%/AbsBiasVsNEvents.pdf][graph]]. This is calculated as the statistical error of the fit times the mean of the pull distribution. Hence this gives the bias in absoulte units. The bias is less than 0.1 % if more than 1000 events are used, below that number a measurable bias is seen. * The sigma of the pull distribution is seen in this [[%ATTACHURL%/SigmaOfPullVsNEvents.pdf][graph]]. Seems to be pretty independent of the number of events. ---+++ Sensitivity to initial values of the fit To study how sensitive the mass fitter is to the initial values given to the fit, a series of 15 fits were made to all 940 Mix_Bd2pipiKpiNBs2KK toy data sets. Four sets of completely random intitial values were used and one set of 'sensible' values for reference. And the fits were perfomed with 10000, 1000 and 100 events (5x3=15 sets of fits). The table below shos the five sets of initial values and the true values of the signal fractions. <table align="center" style="height: 150px; width: 500px;" border="0"> <tbody> <tr> <td align="left" valign="middle">Signal class</td> <td align="center" valign="middle">random 1</td> <td align="center" valign="middle">random 2</td> <td align="center" valign="middle">random 3</td> <td align="center" valign="middle">random 4</td> <td align="center" valign="middle">sensible</td> <td align="center" valign="middle">true value</td> </tr> <tr> <td align="left" valign="middle">Bd -> pi+ pi1</td> <td align="center" valign="middle">0.020</td> <td align="center" valign="middle">0.245</td> <td align="center" valign="middle">0.279</td> <td align="center" valign="middle">0.157</td> <td align="center" valign="middle">0.200</td> <td align="center" valign="middle">0.240</td> </tr> <tr> <td align="left" valign="middle">Bd -> K+ pi</td> <td align="center" valign="middle">0.428</td> <td align="center" valign="middle">0.231</td> <td align="center" valign="middle">0.125</td> <td align="center" valign="middle">0.254</td> <td align="center" valign="middle">0.150</td> <td align="center" valign="middle">0.110</td> </tr> <tr> <td align="left" valign="middle">Bd -> K- pi+</td> <td align="center" valign="middle">0.253</td> <td align="center" valign="middle">0.354</td> <td align="center" valign="middle">0.121</td> <td align="center" valign="middle">0.099</td> <td align="center" valign="middle">0.150</td> <td align="center" valign="middle">0.090</td> </tr> <tr> <td align="left" valign="middle">Bs -> K+ K-</td> <td align="center" valign="middle">0.194</td> <td align="center" valign="middle">0.061</td> <td align="center" valign="middle">0.250</td> <td align="center" valign="middle">0.240</td> <td align="center" valign="middle">0.250</td> <td align="center" valign="middle">0.210</td> </tr> <tr> <td align="left" valign="middle">Combinatoric</td> <td align="center" valign="middle">0.195</td> <td align="center" valign="middle">0.110</td> <td align="center" valign="middle">0.225</td> <td align="center" valign="middle">0.250</td> <td align="center" valign="middle">0.250</td> <td align="center" valign="middle">0.350</td> </tr> </tbody> </table> ---+ The results are summarised in the following graphs: * The error from the fit as reported by MINUIT versus number of events is show in this [[%ATTACHURL%/FitErrorVsInitValues.pdf][graph]]. The statistical error is as expected completely independent of the initial values and depend only on the number of events. * The mean of the pull distribution is shown in this [[%ATTACHURL%/MeanOfPullVsInitValues.pdf][graph]]. The mean of the pull scatters between +/-0.1 for all signal classes and initial values, with a slightly larger spread when only using 100 events. * The bias in absolute numbers is shown in this [[%ATTACHURL%/AbsBiasVsInitValues.pdf][graph]]. This is calculated as the statistical error of the fit times the mean of the pull distribution. Hence this gives the bias in absoulte units (of %). As seen from the graph, the bias in signal fractions are less than 0.1%-units if 1000 or more events are used and less than 0.5%-units if 100 events are used. * The sigma of the pull distribution is seen in this [[%ATTACHURL%/SigmaOfPullVsInitValues.pdf][graph]]. Seems to be pretty independent of the number of events and initial values. -- Main.AlisonBates - 2010-08-06
Attachments
Attachments
Topic attachments
I
Attachment
History
Action
Size
Date
Who
Comment
pdf
AbsBiasVsInitValues.pdf
r1
manage
10.4 K
2010-09-24 - 17:45
LarsEklund
pdf
AbsBiasVsNEvents.pdf
r2
r1
manage
8.8 K
2010-09-24 - 17:24
LarsEklund
pdf
B2hhMassPeaks.pdf
r1
manage
31.9 K
2010-09-26 - 17:26
LarsEklund
png
B2hhMassPeaks.png
r1
manage
53.1 K
2010-09-26 - 17:26
LarsEklund
pdf
FitErrorVsInitValues.pdf
r1
manage
9.9 K
2010-09-24 - 17:45
LarsEklund
pdf
FitErrorVsNEvents.pdf
r2
r1
manage
8.8 K
2010-09-24 - 17:24
LarsEklund
pdf
MeanOfPullVsInitValues.pdf
r1
manage
9.5 K
2010-09-24 - 17:45
LarsEklund
pdf
MeanOfPullVsNEvents.pdf
r2
r1
manage
8.1 K
2010-09-24 - 17:24
LarsEklund
png
Mix1_Bd2KminusPiplus_pull.png
r1
manage
21.9 K
2010-09-21 - 10:16
LarsEklund
png
Mix1_Bd2KplusPiminus_pull.png
r1
manage
23.7 K
2010-09-21 - 10:16
LarsEklund
png
Mix1_Bd2PiPi_pull.png
r1
manage
20.2 K
2010-09-21 - 10:16
LarsEklund
png
Mix2_Bd2KminusPiplus_pull.png
r1
manage
21.5 K
2010-09-21 - 11:23
LarsEklund
png
Mix2_Bd2KplusPiminus_pull.png
r1
manage
22.1 K
2010-09-21 - 11:23
LarsEklund
png
Mix2_Bd2PiPi_pull.png
r1
manage
20.3 K
2010-09-21 - 11:23
LarsEklund
png
Mix2_Bs2KK_pull.png
r1
manage
20.9 K
2010-09-21 - 12:41
LarsEklund
png
Mix3_Bd2KminusPiplus_pull.png
r1
manage
21.9 K
2010-09-21 - 13:00
LarsEklund
png
Mix3_Bd2KplusPiminus_pull.png
r1
manage
21.5 K
2010-09-21 - 13:00
LarsEklund
png
Mix3_Bd2PiPi_pull.png
r1
manage
21.1 K
2010-09-21 - 13:00
LarsEklund
png
Mix3_Bs2KK_pull.png
r1
manage
22.2 K
2010-09-21 - 13:01
LarsEklund
png
Mix3_Bs2KminusPiplus_pull.png
r1
manage
22.7 K
2010-09-21 - 13:01
LarsEklund
png
Mix3_Bs2KplusPiminus_pull.png
r1
manage
21.9 K
2010-09-21 - 13:01
LarsEklund
png
Mix4_Bd2KminusPiplus_pull.png
r1
manage
22.8 K
2010-09-21 - 15:07
LarsEklund
png
Mix4_Bd2KplusPiminus_pull.png
r1
manage
23.4 K
2010-09-21 - 15:07
LarsEklund
png
Mix4_Bd2PiPiPi_pull.png
r1
manage
23.1 K
2010-09-21 - 15:07
LarsEklund
png
Mix4_Bd2PiPi_pull.png
r1
manage
23.8 K
2010-09-21 - 15:07
LarsEklund
png
Mix4_Bs2KK_pull.png
r1
manage
23.3 K
2010-09-21 - 15:11
LarsEklund
png
Mix4_Bs2KminusPiplus_pull.png
r1
manage
23.0 K
2010-09-21 - 15:07
LarsEklund
png
Mix4_Bs2KplusPiminus_pull.png
r1
manage
23.0 K
2010-09-21 - 15:07
LarsEklund
pdf
SigmaOfPullVsInitValues.pdf
r1
manage
9.7 K
2010-09-24 - 17:45
LarsEklund
pdf
SigmaOfPullVsNEvents.pdf
r2
r1
manage
8.2 K
2010-09-24 - 17:25
LarsEklund
gif
c1.gif
r1
manage
7.7 K
2010-08-16 - 14:27
AlisonBates
png
mcEventClasses.png
r1
manage
20.1 K
2010-08-10 - 08:14
LarsEklund
png
mcEventRatios.png
r1
manage
14.4 K
2010-08-10 - 08:14
LarsEklund
Edit
|
Attach
|
Watch
|
P
rint version
|
H
istory
:
r14
<
r13
<
r12
<
r11
<
r10
|
B
acklinks
|
V
iew topic
|
Raw edit
|
More topic actions...
Topic revision: r12 - 2010-09-26
-
LarsEklund
LHCb
Log In
or
Register
LHCb Web
Create New Topic
Index
Search
Changes
Notifications
RSS Feed
Statistics
Preferences
Webs
ATLAS
PUUKA
DetDev
Gridmon
IT
LHCb
LinearCollider
Main
NA62
Sandbox
TWiki
Copyright © 2008-2025 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki?
Send feedback