Page 1
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 1
GLAST LAT Project IA Workshop 5 • August 29, 2005
Raiders of the Raiders of the lost photonslost photons
Elisabetta Bissaldi
University & INFN Trieste, ItalySLAC, USA
Gamma-ray Large AreaGamma-ray Large AreaSpace TelescopeSpace Telescope
Page 2
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 2
GLAST LAT Project IA Workshop 5 • August 29, 2005
OutlineOutline
• Introduction
• Photon selection with MC algorithm
• Results from cosmic ray data for integrated 2, 4 and 6 towers configurations
• Study of vertex topologies
• Summary
Page 3
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 3
GLAST LAT Project IA Workshop 5 • August 29, 2005
Photons
2 tracks = VERTEX
W
SI
SI
e+ e-
Topologies for DummiesTopologies for Dummies
Muons
NO VERTEX ?
W
SI
SI
In TkrRecon: 1 track ≡ VERTEX
Page 4
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 4
GLAST LAT Project IA Workshop 5 • August 29, 2005
Often one of the 2 tracks is missing!
No Si hits No Si hits (edge)
Narrow separation Short track Compton angle scattering
Topologies in real lifeTopologies in real life
W
SISI
W
SISI
W
SISI
W
SISI
W
SISI
Page 5
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 5
GLAST LAT Project IA Workshop 5 • August 29, 2005
Original idea...Original idea...
• Bill Atwood’s talk presented at the 3rd IA Workshop in march 2005
• Develop finding algorithm:– Classification trees– Monte Carlo samples
I. AllGammaII. Surface Muons
– Cosmic ray data sampleIII. Test RUN 140001338
for single tower
candidates sample
(5% of triggers)
Page 6
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 6
GLAST LAT Project IA Workshop 5 • August 29, 2005
BB ChallengeBB Challenge
MC for 1 tower configuration
• Developed Initial selection cuts
• Defined 2 topologies
• Developed algorithm– classification trees
• Applied algorithm in data
MC for 2, 4, 6 towers configurations
Same
• Used only 1 topology
• Searched for simple cuts using main variables
• Applied cuts in data
• Studied vertex topologies
More details in the next slides...
BILL BETTA
Page 7
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 7
GLAST LAT Project IA Workshop 5 • August 29, 2005
• Same
• Focus on first topology
VtxAngle > 0.“VERTEX”
• Same
Initial selectionsInitial selections
• Initial conditions1. TkrNumTracks > 02. CalEnergySum > 10 MeV
• Definition of 2 topologies
3. VtxAngle > 0. VtxAngle ≡ 0. “VERTEX” “1TRACK”
4. Tkr1ToTFirst > 1 MIP
W
SI
SI
W
SI
SI
W
SI
SI
Page 8
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 8
GLAST LAT Project IA Workshop 5 • August 29, 2005
Additional variablesAdditional variables
5. Tkr1SSDVeto > 1
Use Tracker in self Veto Mode to replace ACD At least one hit plane before start of track
Calorimeter
Electronics
Found track
Track projection
same
Page 9
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 9
GLAST LAT Project IA Workshop 5 • August 29, 2005
Bill’s Classification TreesBill’s Classification Trees
Variables used to select :
a) Tkr1ToTTrAveb) Tkr1Chisqc) Tkr1SSDVetod) Tkr1TotFirste) Tkr2Chisqf) Tkr2FirstChisqg) CalEnergySum
* From Bill’s talk
*
VERTEX TOPOLOGY
Page 10
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 10
GLAST LAT Project IA Workshop 5 • August 29, 2005
Betta’s MC/DATA samplesBetta’s MC/DATA samples
I. MC AllGamma
– 1 x 106 events– 33341 trig. events– Isotropic– 18 MeV – 18 GeV
v5r0608p7
II. MC Surface Muons
– 4 x 106 events– 219322 trig. events– Isotropic– PDG formula and low
energy extension v5r0608p7
III. DATA Cosmic Rays
– 2 Towers: RUN 135002134 (462678 trig. events) v5r0608p6
– 4 Towers: RUN 135002778 (61996 trig. events) v4r060302p23
– 6 Towers: RUN 135004075 (390035 trig. events) v5r0608p6
Page 11
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 11
GLAST LAT Project IA Workshop 5 • August 29, 2005
all cuts
no cuts
Tkr1SSDVetoTkr1SSDVeto
MC Muons
all cuts
no cuts
MC All Gamma
Tkr1SSDVeto>2 ?
TkrNumTracks>0CalEnergySum>10. VtxAngle>0. Tkr1TotFirst>1Tkr1SSDVeto>1
Page 12
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 12
GLAST LAT Project IA Workshop 5 • August 29, 2005
Tkr1SSDVetoTkr1SSDVeto
no cuts
MC Muons
no cuts
MC All Gamma
all cutsall cuts
We lose 10% of photons and 70% of muons
TkrNumTracks>0CalEnergySum>10. VtxAngle>0. Tkr1TotFirst>1Tkr1SSDVeto>2
Page 13
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 13
GLAST LAT Project IA Workshop 5 • August 29, 2005
Tkr1SSDVetoTkr1SSDVeto
no cuts
all cutsall cuts
no cuts
MC All Gamma DATA
We lose 10% of cosmic ray data
TkrNumTracks>0CalEnergySum>10. VtxAngle>0. Tkr1TotFirst>1Tkr1SSDVeto>2
Page 14
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 14
GLAST LAT Project IA Workshop 5 • August 29, 2005
Tkr1ToTTrAveTkr1ToTTrAve
no cuts no cuts
all cuts all cuts
MC All Gamma MC Muons
Tkr1ToTTrAve>1.3?
TkrNumTracks>0CalEnergySum>10. VtxAngle>0. Tkr1TotFirst>1Tkr1SSDVeto>2
Page 15
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 15
GLAST LAT Project IA Workshop 5 • August 29, 2005
no cuts no cuts
MC MuonsMC All Gamma
all cuts all cuts
We lose 5% of photons and 70% of muons
TkrNumTracks>0CalEnergySum>10. VtxAngle>0. Tkr1TotFirst>1Tkr1SSDVeto>1Tkr1ToTTrAve>1.3
Tkr1ToTTrAveTkr1ToTTrAve
Page 16
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 16
GLAST LAT Project IA Workshop 5 • August 29, 2005
no cutsno cuts
MC All Gamma DATA
all cuts all cuts
We lose 20% of cosmic ray data
TkrNumTracks>0CalEnergySum>10. VtxAngle>0. Tkr1TotFirst>1Tkr1SSDVeto>1Tkr1ToTTrAve>1.3
Tkr1ToTTrAveTkr1ToTTrAve
Page 17
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 17
GLAST LAT Project IA Workshop 5 • August 29, 2005
2 Towers Results2 Towers Results
MCAllGamma
MCSurface muons
DATACosmic rays
STEPSN°
events%
N° events
%N°
events%
No cuts 33341 100 219322 100 462676 100
1 47 98 97
2 19 33 32
3 9 1.5 3
4 5.7 0.05 1.6
5 5.3 0.03 1.3
6 1672 5.0 25 0.01 4715 1.0Cuts are cumulative!
TkrNumTracks>0+CalEnergySum>10. +VtxAngle>0. +Tkr1TotFirst>1+Tkr1SSDVeto>2+Tkr1ToTTrAve>1.3
Similar results were obtainedfor 4 and 6 towers!(ask me later)
Page 18
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 18
GLAST LAT Project IA Workshop 5 • August 29, 2005
2 Towers Results2 Towers Results
100 %
10 %
1 %
0.1 %
0.01 %
STEPS 1 2 3 4 5 6
5
1
0.01
MC AllGamma
DATA
MC Muons
VtxAngle>0.
Tkr1SSDVeto>2
Cuts are cumulative!
Tkr1ToTTrAve>1.3
Page 19
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 19
GLAST LAT Project IA Workshop 5 • August 29, 2005
candidates with Event Displaycandidates with Event Display
tracks
dir
hitW
SI
SI
tracks
dir
Page 20
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 20
GLAST LAT Project IA Workshop 5 • August 29, 2005
tracks
dir
candidates with Event Displaycandidates with Event Display
W
SI
SI
tracks
dir
Page 21
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 21
GLAST LAT Project IA Workshop 5 • August 29, 2005
DifferentVertexTopologies tracks
dir
candidates with Event Displaycandidates with Event Display
W
SI
SI
tracks
dir
Page 22
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 22
GLAST LAT Project IA Workshop 5 • August 29, 2005
Topologies à la carteTopologies à la carte
W
SI
SI
tracks
dir
W
SI
SI
tracks
dir
W
SI
SI
tracks
dir
DOCA = Distance Of Closest Approach
Page 23
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 23
GLAST LAT Project IA Workshop 5 • August 29, 2005
Vertexing Variable in the Merit Ntuple Summary of track composition and topology
?
VtxStatusVtxStatus
no cuts
MC All Gamma 2 Towers
Page 24
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 24
GLAST LAT Project IA Workshop 5 • August 29, 2005
Single track vertex
2 tracks vertex
More than 2 tracks in
vertex
Vertex positionset by
DOCA point
Vertex tracks share first hit
Vertex tracks start in
different planes
DOCA pointlies inside track hits
VtxStatus Values
0 1 2 3 4 5 6 7
01 1
1 161 1 34
1 641 1 80
1 1281 1 1 162
1 1 1921 1 1 208
VtxStatus bit patternVtxStatus bit pattern
21 + 25 + 27 = 162
1=true
Page 25
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 25
GLAST LAT Project IA Workshop 5 • August 29, 2005
VtxStatus DistributionsVtxStatus Distributions
MC AllGamma 2 Towers
no cuts all cuts
0 and 1
01
162128
43 %28 %12 % 7 %
16234
12816
55 %14 %11 %8 %
162
2.7 % of initial triggers!
Page 26
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 26
GLAST LAT Project IA Workshop 5 • August 29, 2005
...and with DATA...and with DATA
DATA Cosmic rays 2 Towers
no cuts all cuts
1162
0
86 % 6 % 3 %
16234
128
59 %13 %7 %
0 and 1162
0.6 % of initial triggers!
Page 27
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 27
GLAST LAT Project IA Workshop 5 • August 29, 2005
W
SI
SI
VtxStatus = 1622 tracks vertex, vertex tracks share first hit and DOCA pointlies inside track hits
tracks
dir
My friend 162My friend 162
tracks
dir
hit
hit
Page 28
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 28
GLAST LAT Project IA Workshop 5 • August 29, 2005
Energy distributionEnergy distribution
DATA 2 Towers
all cutsVtxStatus=162
all cutsVtxStatus=162
MC AllGamma 2 Towers
TkrNumTracks>0CalEnergySum>10. VtxAngle>0. Tkr1TotFirst>1Tkr1SSDVeto>1Tkr1ToTTrAve>1.3VtxStatus=162
Page 29
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 29
GLAST LAT Project IA Workshop 5 • August 29, 2005
SummarySummary
• Developed a simple algorithm to select in cosmic ray data using Bill’s analysis as a starting point
• Learned about how different the topologies are from the simple picture from pair conversion
• Future Work: Possibility to use the selected sample to study the
behaviour of the detector with different register configurations Depending on the physics analysis, we may be more
interested in selected vertex topologies Explore this further in my thesis Analysis of reprocessed 4, 6 and 8 towers in progress...
Page 30
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 30
GLAST LAT Project IA Workshop 5 • August 29, 2005
GRAZIE...GRAZIE...
• Eduardo• Bill
• Francesco• Anders• Leon• Eric• Luis
...all other italian summer students and people I met at SLAC!
Page 31
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 31
GLAST LAT Project IA Workshop 5 • August 29, 2005
Backup slidesBackup slides
Page 32
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 32
GLAST LAT Project IA Workshop 5 • August 29, 2005
6 Towers Results6 Towers Results
AllGammaSurface muons
Cosmic Data 75
Cosmic Data 77
STEPSN°
events%
N° events
%N°
events%
N° events
%
Recon 89638 100 558757 100 390033 100 389690 100
1 57 98 92 92
2 31 51 45 45
3 15 2 5 5
4 13 0.3 2 2
5 10335 11.5 983 0.18 7588 1.9 7499 1.9
Page 33
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 33
GLAST LAT Project IA Workshop 5 • August 29, 2005
6 Towers Results6 Towers Results
Cosmic Data RUN 135004075
Cosmic Data RUN 135004077
Applied cuts N° events % N° events %
Recon 390033 100 389690 100
1. 92.4 92.4
2. 45.5 45.4
3. 10.6 10.7
4. 28.2 28.1
5. 62.9 63.1
1.+2. 44.7 44.6
1.+2.+3. 4.6 4.6
1.+2.+3.+4. 2.4 2.3
1.+2.+3.+4.+5. 7588 1.9 7499 1.9
Page 34
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 34
GLAST LAT Project IA Workshop 5 • August 29, 2005
4 Towers Results4 Towers Results
AllGamma Surface muons
Applied cuts N° events % N° events %
Recon 62070 100 391538 100
1. 54.2 97.2
2. 71.2 48.6
3. 28.1 4.9
4. 43.9 6.9
5. 46.6 63.8
1.+2. 27.2 45.9
1.+2.+3. 14.0 2.1
1.+2.+3.+4. 10.1 0.2
1.+2.+3.+4.+5. 6171 9.9 608 0.1
Page 35
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 35
GLAST LAT Project IA Workshop 5 • August 29, 2005
4 Towers Results4 Towers Results
Cosmic Data RUN 135002778
Applied cuts N° events %
Recon 61994 100
1. 97.0
2. 43.4
3. 14.7
4. 31.2
5. 67.6
1.+2. 42.7
1.+2.+3. 6.3
1.+2.+3.+4. 3.1
1.+2.+3.+4.+5. 1533 2.5
Page 36
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 36
GLAST LAT Project IA Workshop 5 • August 29, 2005
6 Towers Results6 Towers Results
STEPS 1 2 3 4 5
100 %
10 %
1 %
0.1 %
AllGamma
Data 75Data 77
Muons11.5
1.9
0.18
Page 37
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 37
GLAST LAT Project IA Workshop 5 • August 29, 2005
4 Towers Results4 Towers Results
100 %
10 %
1 %
0.1 %
STEPS 1 2 3 4 5
9.9
2.5
0.1
AllGamma
Data
Muons
Page 38
E. Bissaldi – INFN Trieste, Italy / SLAC, USA 38
GLAST LAT Project IA Workshop 5 • August 29, 2005
MC Surface Muons
no cuts
0 and 1
10
92 %3 %
Same with MC Muons...Same with MC Muons...