Tracking Validation outline Giulia Casarosa Current Performances - - PowerPoint PPT Presentation

▶
tracking validation
SMART_READER_LITE
LIVE PREVIEW

Tracking Validation outline Giulia Casarosa Current Performances - - PowerPoint PPT Presentation

Nov 06, 2022 290 likes •603 views

Tracking Validation outline Giulia Casarosa Current Performances INFN - Sezione di Pisa Overall Efficiency & Purity Track Quality Eugenio Paoloni ~ unipi Pattern Recognition How Jakob Fix Looks Like Face2Face Tracking

slide-1
SLIDE 1

Tracking Validation

Face2Face Tracking Meeting Prague, January 19th 2015 Giulia Casarosa INFN - Sezione di Pisa Eugenio Paoloni ~ unipi

★ Current Performances

  • Overall Efficiency & Purity
  • Track Quality
  • Pattern Recognition

★ How Jakob Fix Looks Like

  • utline
slide-2
SLIDE 2

F2F Prague - 19th January 2015 Giulia Casarosa

Current Performances

2

track quality efficiency & purity pattern recognition

+ +

➡ used release r14908 (except otherwise stated), TrackingPerformanceEvaluation Module

  • Belle CDC pattern recognition (Trasan)
  • Merger of VXD TrackCand + CDC TrackCand uses MC truth information
  • no extrapolation from CDC to VXD and vice-versa
slide-3
SLIDE 3

F2F Prague - 19th January 2015 Giulia Casarosa

Integrated Efficiency & Purity

3

tracking(*) VXD only CDC only VXD+CDC purity (%) 95.72±0.08 77.1±0.1

  • efficiency (%)

68.4±0.2 75.3±0.1 82.0±0.1 efficiency’ (%) 77.8±0.1 91.1±0.1 91.8±0.1

ε = ε’ = # MCParticles with at least one associated Track # MCParticles ε = # MCTrackCands with at least one associated Track # MCTrackCands ε’ = purity = probability to find an MCParticle associated to a Track, given a Track physical efficiency geometrical acceptance and detector efficiency are factored out

(*) numbers and plots shown are based on 8k Y(4S) events

efficiency & purity not updated see next slide not updated

(r14200) (r14200)

slide-4
SLIDE 4

F2F Prague - 19th January 2015 Giulia Casarosa

Reconstruction run-time errors related to tracking

4

[ERROR] Uncaught exception encountered: Out-of-range access in StoreArray::operator[], for array 'TrackCands' (durability: event) { module: MCTrackMatcher } Traceback (most recent call last): File "/gpfs/home/belle2/casarosa/releases/rValidation2/workdir/../tracking/validation/ a02_trackingEfficiency_runTracking.py", line 85, in <module> process(path) File "/gpfs/home/belle2/casarosa/releases/rValidation2/lib/Linux_x86_64/debug/basf2_def.py", line 155, in process fw.process(path) IndexError: Out-of-range access in StoreArray::operator[], for array 'TrackCands' (durability: event) [ERROR] Uncaught exception encountered: Out-of-range access in StoreArray::operator[], for array 'TrackFitResults' (durability: event) { module: ParticleLoader } Traceback (most recent call last): File "/home/belle2/casarosa/releases/build-2014-10-18/workdir/scripts/skim.py", line 85, in <module> process(analysis_main) File "/sw/belle2/releases/build-2014-10-18/lib/Linux_x86_64/debug/basf2_def.py", line 153, in process fw.process(path) IndexError: Out-of-range access in StoreArray::operator[], for array 'TrackFitResults' (durability: event)

➡ r14908, during this study, after a few events, basf2 exited with this error: ➡ build-2014-10-18, Skimming MC-4.5 failed jobs are almost all due to:

slide-5
SLIDE 5

F2F Prague - 19th January 2015 Giulia Casarosa

Efficiency VS Transverse Momentum

5

legend:

  • ε, physical efficiency
  • ε’, geometrical acceptance and detector efficiency factored out

ε’ = 77% @ pt = 100 MeV/c

VXD+CDC VXD+CDC

1-ε’ = 3% @ pt = 1 GeV/c ε’ = 97% @ pt = 500 MeV/c

efficiency & purity

slide-6
SLIDE 6

F2F Prague - 19th January 2015 Giulia Casarosa

Efficiency VS Polar and Azimuthal Angles

6

legend:

  • ε, physical efficiency
  • ε’, geometrical acceptance and detector efficiency factored out

VXD+CDC VXD+CDC efficiency & purity

flat as expected clear effect of the geometrical acceptance lower ε in the forward and backward directions

slide-7
SLIDE 7

F2F Prague - 19th January 2015 Giulia Casarosa

Transverse Impact Parameter

➡ almost gaussian pulls distribution

7

VXD+CDC

⟨σd0⟩ = 131μm

VXD+CDC track quality

slide-8
SLIDE 8

F2F Prague - 19th January 2015 Giulia Casarosa

σd0 VS transverse momentum

➡ 3 families of tracks:

  • CDC only tracks
  • tracks with no PXD clusters

attached

  • tracks with PXD clusters

attached

8

VXD+CDC track quality

slide-9
SLIDE 9

F2F Prague - 19th January 2015 Giulia Casarosa

Comparison with BABAR Tracking

➡ Belle2 VXD Tracking when PXD

clusters are used in the track fit performs better than BABAR

➡ Belle2 Tracking when no PXD

clusters are used in the track fit performs worse than BABAR

9

BABAR σz0 BABAR σd0 B2 σd0 Tracks with PXD cluster B2 σd0 Tracks with NO PXD cluster

VXDonly r14200 r14200

slide-10
SLIDE 10

F2F Prague - 19th January 2015 Giulia Casarosa

Relating Track Parameters Errors (1)

10

track quality

σd0 ≃ rt · σϕ σd0 σϕ

track PXD layer beampipe VXD+CDC

➡ Can relate the errors on d0 and ϕ and estimate the distance between the origin

and the nearest hit (i.e. the hit that has more weight in the d0 estimation )

➡ valid when first hit is near the beam pipe ➡ neglecting detector hit resolution

MC-truth-information free plot!

pt

slide-11
SLIDE 11

F2F Prague - 19th January 2015 Giulia Casarosa

Relating Track Parameters Errors (2)

11

track quality

σz0 ≃ rt · σcotθ

VXD+CDC VXD+CDC

➡ Can also user the errors on z0 and cotθ to estimate the distance between the

  • rigin and the nearest hit (i.e. the hit that has more weight in the z0 estimation )

MC-truth-information free plot!

slide-12
SLIDE 12

F2F Prague - 19th January 2015 Giulia Casarosa

Relating Track Parameters Errors (3)

12

VXD+CDC track quality VXD+CDC

ProjectionY( ) of the left plot

beampipe

➡ Can also look at the longitudinal view and plot the distance between the

  • rigin and the nearest hit to the origin.

MC-truth-information free plots!

slide-13
SLIDE 13

F2F Prague - 19th January 2015 Giulia Casarosa

Effect of Multiple Scattering

➡ taking into account the error introduced by the multiple scattering of the

first layer crossed by the track:

13

VXD+CDC VXD+CDC track quality

ProfileX( ) of the left plot

pt·σd0 β p sin3/2θ = cost σd0 β p sin3/2θ·(13.6MeV/c)–1 VS pt σd0 β p sin3/2θ·(13.6MeV/c)–1 VS pt

slide-14
SLIDE 14

F2F Prague - 19th January 2015 Giulia Casarosa

Patter Recognition Efficiency & Purity

14

Pattern Recognition VXD only CDC only VXD+CDC purity (%) 94.88±0.08 75.5±0.1

  • efficiency (%)

79.3±0.2 91.3±0.1 94.3±0.1

# MCTrackCand with at least one associated TrackCand # MCTrackCand ε = ε = purity = probability to find an MCTrackCand associated to a TrackCand, given a TrackCand

pattern recognition

➡ above numbers are relative to release r14200

  • VXD only PR efficiency increased to 85% with the patch, same purity
  • VXD+CDC PR efficiency increased to 96.1% with the patch

not updated not updated

slide-15
SLIDE 15

F2F Prague - 19th January 2015 Giulia Casarosa

PR Efficiency on VXD Layers

15

  • Efficiency is maximum on layer3 (80%)
  • Efficiency decreases from inner to the outer layers (70% on layer1, 60% on layer6)

pattern recognition VXD+CDC

PR efficiency VS VXD Layer Number

layer1 layer2 layer3 layer4 layer5 layer6

slide-16
SLIDE 16

F2F Prague - 19th January 2015 Giulia Casarosa

VXD Clusters used in the Track Fit

16

L1 L2 L3

lower efficiency in PXD slanted ladders?

hits used in the Track Fit: transverse view

L2 L1

hits used in the Track Fit: longitudinal view

hot spots

pattern recognition VXD+CDC VXD+CDC

slide-17
SLIDE 17

F2F Prague - 19th January 2015 Giulia Casarosa

How Jakob Fix Looks Like

17

➡ compare VXD only tracking in r14200 (before the patch) and r14200+patch

patch = svn update of:

  • tracking/modules/VXDTF/{include,src}/VXDTFModule.{h,cc}
  • tracking/modules/VXDTFHelperTools/{include,src}/FilterCalculatorModule.{h,cc}
  • tracking/spacePointCreation/{include,src}/spacepoint.{h,cc}
  • tracking/vectorTools
slide-18
SLIDE 18

F2F Prague - 19th January 2015 Giulia Casarosa

Efficiency VS Polar and Azimuthal Angles

18

efficiency & purity VXDonly VXDonly

known modulation

  • f 2π/7 is still present

known drop for θ=π/2, still there

➡ improvement in efficiency but undesired feature are still present

slide-19
SLIDE 19

F2F Prague - 19th January 2015 Giulia Casarosa

Efficiency VS pt and VXD Layers

19

VXDonly VXDonly efficiency & purity

PR efficiency VS VXD Layer Number

layer1 layer2 layer3 layer4 layer5 layer6

➡ improvement in efficiency VS pt ➡ huge increase of the efficiency on layers 1 and 2, but an undesired 10%

decrease on outer layers shows up

slide-20
SLIDE 20

F2F Prague - 19th January 2015 Giulia Casarosa

Transverse Impact Parameter Resolution

20

track quality VXDonly VXDonly

Tracks with PXD cluster Tracks with NO PXD cluster

patch r14200

Tracks with PXD cluster Tracks with NO PXD cluster

slide-21
SLIDE 21

F2F Prague - 19th January 2015 Giulia Casarosa

σd0 with and without PXD clusters

21

track quality VXDonly VXDonly

slide-22
SLIDE 22

F2F Prague - 19th January 2015 Giulia Casarosa

Relating Track Parameters Errors

22

VXDonly track quality

➡ striking difference between the two curves

projection on rt of σd0/σϕ (v. slide 12)

slide-23
SLIDE 23

F2F Prague - 19th January 2015 Giulia Casarosa

Conclusions

➡ Good improvements in the tracking

performances after Jakob Fix!

➡ Some plots that may be implemented in

  • ur new validation:

23

Thank You!

σd0/σϕ z-rt view projected on rt (slide 12) σd0 β p sin3/2θ VS pt Profiled on pt (slide 13) PR efficiency VS VXD Layer Number (slide 15) σd0 VS pt w/wo PXD hits Profiled on pt (slide 21)

slide-24
SLIDE 24

F2F Prague - 19th January 2015 Giulia Casarosa

backup slides

24

https://agenda.infn.it/getFile.py/access?resId=0&materialId=slides&contribId=7&sessionId=3&subContId=1&confId=8680 r14200 validation plots here:

slide-25
SLIDE 25

F2F Prague - 19th January 2015 Giulia Casarosa

σd0 VS transverse momentum

➡ 3 families of tracks:

  • CDC only tracks
  • tracks with no PXD clusters

attached

  • tracks with PXD clusters

attached

25

VXD+CDC r14200

slide-26
SLIDE 26

F2F Prague - 19th January 2015 Giulia Casarosa

26

r14200

slide-27
SLIDE 27

F2F Prague - 19th January 2015 Giulia Casarosa

27

r14200

slide-28
SLIDE 28

F2F Prague - 19th January 2015 Giulia Casarosa

28

r14200

slide-29
SLIDE 29

F2F Prague - 19th January 2015 Giulia Casarosa

29

r14200

slide-30
SLIDE 30

F2F Prague - 19th January 2015 Giulia Casarosa

VXD Clusters used in the Track Fit

30

L1 L2 L3

holes in PXD slanted ladders

used in the Track Fit: transverse view

L3 L2 L1

backward modules?

hits used in the Track Fit: longitudinal view

T-shape due to 1D SVD clusters

r14200