Ondra's TWiki

For 1D simultaneous fits go to SimFits

For 2D simultaneous fits go to SimFits2D

For full scan HLT B search go to FullScan

For tests of 2020 BDT go to NewBDT

For FR2 yield fitting go to YieldFits

For systematics go to SystematicsFR2

Individual fits are here.

The most useful information in the history

MINUIT covariance matrix quality codes:

1 - Approximation only, not accurate

2 - Full matrix, but forced positive-definite

3 - Full, acurate covariance matrix

Toy MC production

Individual Fits

In the 'individual mode', each bin of the reference sample is fitted in the following way:

Signal:

1. Single Gaussian, starting from: μ... sample mean, σ... sample rms
2. Two Gaussians, starting from: μ1, μ2... result of previous step, σ1 resp. σ2 ... 0.8 resp 1.2 x result of previous step, f1... 0.55
3. Three Gaussians with common μ1, μ2, starting from: μ1, μ2 ... resulting μ1 of previous step, σ1 resp. σ2 ... 0.8 resp 1.2 x resulting σ1 of previous step, σ3 ... resulting σ2 of previous step, f1... resulting f1 of previous step, f2... 0.5
4. Three Gaussians - unconstrained μ1, μ2, starting from: results of previous step

Background:

1. Binned fit - exponential, on lower sideband only, starting from: expConst... 0.001 MeV -1, binning... 50
2. Binned fit - Chebychev, on upper sideband only, starting from: slope... 0 MeV -1, binning... 50
3. Unbinned fit - exponential, on lower sideband only, starting from: result of the binned fit
4. Unbinned fit - Chebychev, on upper sideband only, starting from: result of the binned fit
5. Exponential + Chebychev, both sidebands, starting from: expConst... result of step 3., slope... very complicated combination of result of step 4, numbers of entries, fit range limits, etc.

Studies on the CombinatorialConstraint (constraining bkg slope) were performed after observing positive values in BDT bin 4 and η bin 2. Two forms of the constraints were tested:

1. Limiting the range of the parameter to [-10., 0.], resulting in: BDT bin 4: (-7.37293e-08 +/- 0.52879) MeV -1, η bin 2: (-7.9271e-07 +/- 0.445156) MeV -1
2. Forcing the model to use -Abs(slope) in the calculations, resulting in: BDT bin 4: (0 +/- 0.00515252) MeV -1, η bin 2: (6.37492e-05 +/- 0.042386) MeV -1
   (note that the actual fit parameter "slope" remains unaffected, but the MODEL is changed to use the -Abs(slope)

git repo: https://gitlab.cern.ch/ixiotidi/bs2mumu_fitclass/-/tree/developmentOndra/ClassExec

Parameter Evolution Studies

Based on the results of individual fits, the evolution of fit parameters wrt. η and BDT was studied. For evolution wrt. η, no binning in BDT was introduced and vice versa.

Evolution in η - signal

Invariant mass fits in individual η bins:

eta bin 1

eta bin 2

eta bin 3

Evolution of the fit parameters:

Mean1	Mean2	Mean3	Sigma1
Sigma2	Sigma3	f1	f2

Evolution wrt. η - background

Data side bands

Invariant mass fits in individual η bins:

eta bin 1

eta bin 2

eta bin 3

Evolution of the fit parameters:

Exponential constant

Slope

N Combinatorial

N Total

Fraction of combinatorial bkg

Background MC

Fit log: bkgMCEtaEv.txt

Invariant mass fits in individual η bins:

checkModel_BDTBin1ETAbin1_etaEvMC

checkModel_BDTBin1ETAbin2_etaEvMC

checkModel_BDTBin1ETAbin3_etaEvMC

Evolution of the fit parameters:

Exponential constant

Slope

N Combinatorial

N Total

Fraction of combinatorial bkg

Evolution wrt. BDT - signal

Invariant mass fits in individual BDT bins:

BDT bin 1

BDT bin 2

BDT bin 3

BDT bin 4

Evolution of the fit parameters:

Mean1	Mean2	Mean3	Sigma1
Sigma2	Sigma3	f1	f2

Evolution wrt. BDT - background

Data side bands:

Invariant mass fits in individual BDT bins:

BDT bin 1

BDTbin 2

BDT bin 3

BDT bin 4

Evolution of the fit parameters:

Exponential constant

Slope - bin 4 completely off the scale

N Combinatorial

N Total

Fraction of combinatorial bkg

Background MC:

Fit log: bkgMCBDTEv.txt

There is wrong BDT boundary in these plots between bins 3 and 4 - 0.415 instead of correct 0.4163. This shifted 6 events to the upper bin. In other cases on these pages this is correct.

Invariant mass fits in individual BDT bins:

checkModel_BDTBin1ETAbin1_bkgMCBDTEv

checkModel_BDTBin2ETAbin1_bkgMCBDTEv

checkModel_BDTBin3ETAbin1_bkgMCBDTEv

checkModel_BDTBin4ETAbin1_bkgMCBDTEv

Evolution of the fit parameters:

Exponential constant

Slope

N Combinatorial

N Total

Fraction of combinatorial bkg

Simultaneous Fits to B_s Invariant Mass MC

1D simultaneous fits: SimFits

2D simultaneous fits: SimFits2D

git repo: https://gitlab.cern.ch/okovanda/toymc_bsignal

Sample binned in η only

The simultaneous fits are performed by RooFit. When parameters in all bins are independent, we obtain the same results. Besides PDFs for each bin, the overall likelihood is multiplied by gaussian factor for each bin with μ = avEtaTrue, σ = 1. This factor is included to acommodate the trends of certain parameters as function of avEtaTrue. Initial values for the simultaneous fit are taken as the final values of the individual fits. The main result is the percentual change of the results of the simultaneous fit with respect to the initial values (100*(initial-final)/initial).

Results with all parameters free: resultEtaSigma=1_allFree

Results with Sigma1 linear on avEtaTrue: resultEtaSigma=1_sigma1Dep

Results with Sigma1 and Sigma2 linear on avEtaTrue: resultEtaSigma=1

Fullscan ID tracking studies

Main page is FullScan.

My Links

• ATasteOfTWiki - view a short introductory presentation on TWiki for beginners
• WelcomeGuest - starting points on TWiki
• TWikiUsersGuide - complete TWiki documentation, Quick Start to Reference
• Sandbox - try out TWiki on your own
• OndrejKovandaSandbox - just for me

My Personal Preferences

• Show tool-tip topic info on mouse-over of WikiWord links, on or off:
  • Set LINKTOOLTIPINFO = off

• Preference for the editor, default is the WYSIWYG editor. The options are raw, wysiwyg:
  • Set EDITMETHOD = wysiwyg
• users
  • Set ALLOWWEBEDIT = AlessandroCerri

• More preferences TWiki has system wide preferences settings defined in TWikiPreferences. You can customize preferences settings to your needs: To overload a system setting, (1) do a "raw view" on TWikiPreferences, (2) copy a Set VARIABLE = value bullet, (3) do a "raw edit" of your user profile page, (4) add the bullet to the bullet list above, and (5) customize the value as needed. Make sure the settings render as real bullets (in "raw edit", a bullet requires 3 or 6 spaces before the asterisk).

Related Topics

• ChangePassword for changing your password
• ChangeEmailAddress for changing your email address
• UserList has a list of other TWiki users
• UserDocumentationCategory is a list of TWiki user documentation
• UserToolsCategory lists all TWiki user tools