SCT Offline Monitor SCT Offline Monitor Measuring Module Hit Measuring Module Hit

EfficienciesEfficienciesHelen HaywardHelen Hayward

University of LiverpoolUniversity of Liverpool

AimsAims

Demonstrate/verify proper functioning of SCT Demonstrate/verify proper functioning of SCT modules by tracking particles through the modules by tracking particles through the detector and comparing the hits found to those detector and comparing the hits found to those predictedpredicted Measure hit efficiency on a per module levelMeasure hit efficiency on a per module level

Looking at Barrel and forward RegionsLooking at Barrel and forward Regions Code should be versatile enough to be used for Code should be versatile enough to be used for

commisioning and long termcommisioning and long term Be able to read and write results to databaseBe able to read and write results to database Using ~18,000 minbias events, Using ~18,000 minbias events, Software version 10.0.2Software version 10.0.2

For a given module:For a given module: Eff =Eff = #hits #hits

ObservedObserved #Hits #Hits

ExpectedExpected

Numerator:Numerator: A simple map of all hits A simple map of all hits

associated with the associated with the reconstructed tracksreconstructed tracks

Plot is for 2nd barrel Plot is for 2nd barrel layerlayer

Hits On Track (Numerator)Hits On Track (Numerator)

PredictedHits (Denominator)PredictedHits (Denominator)

For a given track For a given track intersecting a given intersecting a given layer/disk:layer/disk: Remove hits (if present) Remove hits (if present)

from ifrom ithth disk and refit track disk and refit track (using xkalman)(using xkalman)

Extrapolate refitted track to Extrapolate refitted track to iithth disk and find intersection disk and find intersection pointpoint (Using TrkExtrapolation)(Using TrkExtrapolation)

Predict a hit should have Predict a hit should have been found at this pointbeen found at this point

This prediction becomes This prediction becomes the denominatorthe denominator

Can now calculate Can now calculate efficiency ……efficiency ……

Inefficiency PlotsInefficiency Plots 1-efficiency plots1-efficiency plots In this example, I have In this example, I have

introduced some inefficiencies:introduced some inefficiencies: one chip on side 0 (128 strips) one chip on side 0 (128 strips)

have been turned off in the phi=15 have been turned off in the phi=15 modules.modules.

Theoretically they should have Theoretically they should have an inefficiency of 8%an inefficiency of 8%

Some 100% efficient modules Some 100% efficient modules have been measured as being have been measured as being ~2% inefficient~2% inefficient

Plot is for 2nd barrel Plot is for 2nd barrel layerlayer

More DetailMore Detail

For any module with an efficiency less than For any module with an efficiency less than x (value set by user), strip detail is recorded.x (value set by user), strip detail is recorded. X=95% in these examples.X=95% in these examples.

Can look at Strip map to look for cause of Can look at Strip map to look for cause of inefficiencyinefficiency

Side 0

Side 1

Summary tableSummary table

Efficiency per layer for barrel SCT:Efficiency per layer for barrel SCT:

LayerLayer Number Number measuredmeasured

Number Number PredictedPredicted

Eff Eff measuredmeasured

EffEff

inputinput

00 8587885878 8705887058 0.986±0.0050.986±0.005 0.9950.995

11 7454874548 7525375253 0.991±0.0050.991±0.005 1.0001.000

22 6579265792 6637666376 0.991±0.0060.991±0.006 0.9970.997

33 5811658116 5815758157 0.999±0.0060.999±0.006 0.9990.999

Source of Mismeasurement?Source of Mismeasurement?

Appears to occur Appears to occur when only one hit when only one hit is measured, yet is measured, yet two are predictedtwo are predicted Edge of Module?Edge of Module?

Do original and Do original and refitted tracks have refitted tracks have similar paths?.similar paths?.

Measured Single StripMeasured Single Strip

Plotted is the Plotted is the distribution of all distribution of all observed single observed single hits (on either hits (on either side) of the side) of the barrel modulesbarrel modules

As expected As expected most are found most are found at the edge of at the edge of the modulethe module

Single HitMapSingle HitMap

Side 0Side 0 Side 1Side 1

Any Track quality effects?Any Track quality effects?

Conceivable that poorly fitted tracks, or Conceivable that poorly fitted tracks, or very soft tracks could cause problems very soft tracks could cause problems efficiency calculationefficiency calculation

For each hit:For each hit: Classify according to whether the efficiency Classify according to whether the efficiency

has been correctly or incorrectly measuredhas been correctly or incorrectly measured Plot various track quantitiesPlot various track quantities

TrackQuality ITrackQuality I

TrackQuality IITrackQuality II

Difference in P between original and Difference in P between original and refitted trackrefitted track

SCT standaloneSCT standalone Initially, monitoring with Initially, monitoring with

SCT alone?SCT alone?

Link to DatabaseLink to Database Once decided that strip is inefficient or dead, need to record it to Once decided that strip is inefficient or dead, need to record it to

database.database. Have written some code in 11.0.x developed from:Have written some code in 11.0.x developed from:

DatabaseTest/IOVDbTestAlg/IOVDbTestAlg.cxxDatabaseTest/IOVDbTestAlg/IOVDbTestAlg.cxx Can read and write to database in execute or finalize.Can read and write to database in execute or finalize.

Information written:Information written:

Also writing of monitoring histograms/files to Database on finalise?Also writing of monitoring histograms/files to Database on finalise? Convert algorithm to an AlgTool so can be used by anyone.Convert algorithm to an AlgTool so can be used by anyone. Would like to be able to read from database initially to remove from Would like to be able to read from database initially to remove from

efficiency calculation any tracks intercepting known dead channels?efficiency calculation any tracks intercepting known dead channels?

ChannelChannel

Barrel_ecBarrel_ec Layer_diskLayer_disk EtaEta PhiPhi SideSide StripStrip CommentComment

IntInt IntInt IntInt IntInt IntInt IntInt IntInt StringString

…… …… …… …… …… …… …… ……

SummarySummary

Measuring hit efficiency per module Measuring hit efficiency per module Enough statistics to see chip-level effectsEnough statistics to see chip-level effects Hit Efficiency working well in central regionHit Efficiency working well in central region Need to work out strategy for calculation near Need to work out strategy for calculation near

edgesedges Fiducial cut->loss of informationFiducial cut->loss of information

Requirement on minimum number of hits on Requirement on minimum number of hits on tracktrack

Appears to be robust enough to work with SCT Appears to be robust enough to work with SCT standalone tracksstandalone tracks

Coming Soon….Coming Soon…. Work on reducing mis-measurements…. Work on reducing mis-measurements….

Optimise TrackQuality/Module FiducialityOptimise TrackQuality/Module Fiduciality Look at separate “edge strategy”Look at separate “edge strategy” Update to 11.0.xUpdate to 11.0.x Study effect of misalignments on codeStudy effect of misalignments on code

OutliersOutliers Sensitive to mis-alignments?Sensitive to mis-alignments?

Replace denominator with “truth track” to see how my Replace denominator with “truth track” to see how my results compareresults compare

Look at Cosmic data from forthcoming runLook at Cosmic data from forthcoming run Real data with real issuesReal data with real issues Write to database (code available for online monitoring aswell)Write to database (code available for online monitoring aswell)

Other Samples: di-jet, high Pt muon samples, Other Samples: di-jet, high Pt muon samples,

TrackQuality ITrackQuality I