Trigger Workshop material CERN Anton Osika

L1 Trigger selections for Fat JetsContribution to the Hadronic Calibration 2013 Workshop

(Jet Substructure and tagging)

Anton Osika, Anna Sfyrla, Zachary Marshall

CERN

[email protected]

September 16, 2013

Anton Osika, Anna Sfyrla, Zachary Marshall (CERN)L1 Trigger selections for Fat Jets September 16, 2013 1 / 51

Overview

1 L1 Trigger selections for Fat JetsIntroductionResultsSummary

2 Detailed MaterialIntroduction - more detailsOnline/Offline CorrelationsTrigger EfficienciesEvent properties per sampleSummary


Table of Contents

1 L1 Trigger selections for Fat JetsIntroductionResultsSummary

2 Detailed Material


Introduction & Motivation

Events with fat jets and no leptons are typically triggered using fat jettriggers.

These require a fat jet at the HLTi.e. significant energy deposits in cones of ∆R = 1.0.They are seeded at L1 by a ‘standard’ (narrow) single jet item, as L1uses exclusively cones of 0.8× 0.8 in the η − φ space.

We are seeking answers in the two following questions:

How is the L1 seed affecting the fat jet trigger efficiencies?What is the best alternative to a single jet L1 seed?


Trigger Selections

Investigated variables to cut at:Selection Threshold example AcronymET of leading single jet - Default L1 selection 100 [ GeV ] J1004 jets w/ ET above theshold - L1 multijet selection 20 [ GeV ] 4J202/3 jets w/ ET > 20, all closer than ∆R 1.0 2J20DR12Sum of ET for jets w/ ET > 20 200 [ GeV ] HT200As HT, for jets w/ |η| < 2.5 200 [ GeV ] HTC200Sum of ET for (up to) 2 jets closer than ∆R = 1.0 100 [ GeV ]

PET(2)100

Sum of ET for (up to) 3 jets closer than ∆R = 1.0 100 [ GeV ]P

ET(3)100

We have also considered additional variables, proven not that interesting inthe end: Invariant mass of two closeby jets and requirements close-by taus,in combination to single jet, multijet or HT selections.These will not be shown in the following slides.

Reminders:

run1 lowest unprescaled single jet L1 item: L1 J75

lowest unprescaled single jet L1 item planned for run2: L1 J100


Results 1 - efficiency(fat jet pT)

The baseline L1 J100 is compared to L1 HTC200 and two other selections thatinclude close-by jet requirements; OR-ing this selection to the baseline recoversinefficiencies of the individual selections.

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 close jet AND HT selection ORed with J100, zprime1000 (20000)

Efficiency for close jet selection

Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 close jet AND HT selection ORed with J100, Ttbar (50000)


Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 close jet AND HT selection ORed with J100, GttL900 (10000)


Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets



Results 2 - efficiency(fat jet pT)

The baseline L1 J100 is compared to L1 4J20, HT200, HTC200 and the ET sumof up to two close-by jets. The 4-jet selection leads to large inefficiencies inevents without a large jet multiplicity.

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 Summarizing Selections , zprime1000 (20000)

Efficency for L1 multijet selection

Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 Summarizing Selections , Ttbar (50000)


Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 Summarizing Selections , GttL900 (10000)


Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets



Summary

Summary:

We have looked at L1 seeds for fat jet triggers; more specifically, atalternatives to the baseline L1 single jet trigger items:

a selection of L1 jet close-by pairs with high∑

ET , ORed to thebaseline selection;HT constructed from 20GeV L1 jets with |η| < 3.2 or < 2.5,multijet triggers,the two above, including requirements on the distance between jets.

HT(C)200 seems to have the best overall performance in the models wehave considered and for the assumed threshold of 360GeV for the EFselection.

Other selections can recover inefficiencies if lower EF thresholds can beallowed; e.g. if EF fat jet selections are made more robust to pile-up.

Further possible steps:

Investigate the pile-up robustness for the various selections;

Investigate L1 processing limitations when searching for combinations ofclose-by jets.


Table of Contents

1 L1 Trigger selections for Fat Jets

2 Detailed MaterialIntroduction - more detailsOnline/Offline CorrelationsTrigger EfficienciesEvent properties per sampleSummary


Introduction - more details


L1(Topo) after LS1

After LS1, at L1 there will be provided new topological selection capability;

Selections will be possible on angles and kinematic compbinations ofobjects found in L1Calo, and with limited information from L1Muon.

This will be critical for physics channels with multiple objects in final statethat so far relied on inclusive (high rate) L1 triggers.

Proposed (hadronic) selections include HT, MHT, ∆η, ∆φ, ∆R (e.g.between jets or jets and MET), dijet invariant mass, ...


Processes

For the studies presented in these slides we used the following three samplecategories (resulting in four samples)

tt̄ production;

Z ′ → tt̄, with the Z ′ massed fixed at 1TeV;

g̃ → tt̄χ01, pair produced. Two samples are used from this model; both have

the g̃ mass fixed at 1.4TeV; one has the χ01 mass fixed at 1GeV and the

other has the χ01 mass fixed at 900GeV.

These four samples give a variety of fat jet multiplicity and pT spectrum in thefinal state, thus ensuring a good coverage.


Rates

The threshold for a possible variable is primarily decided by its estimatedrate;

The efficiency decides if the trigger is desirable or not;

The table below shows rates for various triggers at 14TeV, pile-up 54 for25ns and the 12 first BCIDs vetoes. Lumi considered here: 2e34. Source:

https://twiki.cern.ch/twiki/bin/viewauth/Atlas/RateEstimator

Trigger Rate Unique rate w.r.t. J100

J100 5.8 ± 0.7 0HT200 4.8 ± 0.7 1.2 ± 0.3HTC200 3.8 ± 1.0 1.0 ± 0.4HT250 2.1 ± 0.4 0.7 ± 0.34J20 4.5 ±0.7 4.0 ± 0.7∆R < 1.0

PET (2) > 120 0.5 ± 0.1 0.1 ± 0.1

∆R < 1.0P

ET (2) > 100 0.8 ± 0.3 0.3 ± 0.3∆R < 1.2

PET (2) > 120 0.5 ± 0.2 0.2 ± 0.2

∆R < 1.2P

ET (2) > 100 1.1 ± 0.4 0.5 ± 0.3


Event Filter and Offline selection

“EF selection” always imposes a fat jet requirement (R = 1.0) of ET > 360GeV and |η| < 3.2; EF jets are ‘AntiKt10 lctopo’;

“Offline fat jet selection” pre-requires pT> 50 GeV and |η| < 2.0


Online/Offline Correlations


L1 Jet ET vs Fat Jet pT

0

20

40

60

80

100

120

140

160

180

200

Offline fat jet pT0 100 200 300 400 500 600 700 800

Onl

ine

lead

ing

jet E

T

0

100

200

300

400

500

600

700

800

Leading online ET vs Offline leading pT, a10, zprime1000 (20000)

Correlation Factor: 0.86

Leading online ET vs Offline leading pT, a10, zprime1000 (20000)

0

500

1000

1500

2000

2500

3000

3500

4000


Onl

ine

lead

ing

jet E

T

0

100

200

300

400

500

600

700

800

Leading online ET vs Offline leading pT, a10, Ttbar (50000)


Leading online ET vs Offline leading pT, a10, Ttbar (50000)

0

20

40

60

80

100

120

140

160

180

200


Onl

ine

lead

ing

jet E

T

0

100

200

300

400

500

600

700

800

Leading online ET vs Offline leading pT, a10, GttL900 (10000)



0

20

40

60

80

100

120

140


Onl

ine

lead

ing

jet E

T

0

100

200

300

400

500

600

700

800





L1∑

ET (2) vs Fat Jet pT

0

50

100

150

200

250

300

350

400

450


sum

clo

se je

tsT

Onl

ine

E

0

100

200

300

400

500

600

700

800

Online sum of ET for, up to 3, close jets vs Offline leading pT, a10, zprime1000 (20000)


Online sum of ET for, up to 3, close jets vs Offline leading pT, a10, zprime1000 (20000)

0

1000

2000

3000

4000

5000

6000

7000

8000


sum

clo

se je

tsT

Onl

ine

E

0

100

200

300

400

500

600

700

800

Online sum of ET for, up to 3, close jets vs Offline leading pT, a10, Ttbar (50000)


Online sum of ET for, up to 3, close jets vs Offline leading pT, a10, Ttbar (50000)

0

50

100

150

200

250

300

350


sum

clo

se je

tsT

Onl

ine

E

0

100

200

300

400

500

600

700

800

Online sum of ET for, up to 3, close jets vs Offline leading pT, a10, GttL900 (10000)



0

50

100

150

200

250

300

350

400


sum

clo

se je

tsT

Onl

ine

E

0

100

200

300

400

500

600

700

800





L1 HT vs Fat Jet pT

0

10

20

30

40

50

60

70

80

90


Onl

ine

HT

0

100

200

300

400

500

600

700

800

Online HT vs Offline leading pT, a10, zprime1000 (20000)


Online HT vs Offline leading pT, a10, zprime1000 (20000)

0

200

400

600

800

1000

1200


Onl

ine

HT

0

100

200

300

400

500

600

700

800

Online HT vs Offline leading pT, a10, Ttbar (50000)


Online HT vs Offline leading pT, a10, Ttbar (50000)

0

10

20

30

40

50

60

70


Onl

ine

HT

0

100

200

300

400

500

600

700

800

Online HT vs Offline leading pT, a10, GttL900 (10000)



0

10

20

30

40

50

60

70


Onl

ine

HT

0

100

200

300

400

500

600

700

800





Trigger Efficiencies


L1 HT triggers

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

HTC200 HT250 J100 comparison, zprime1000 (20000)


Signal distribution

J100

HT250

HTC200

HT200


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

HTC200 HT250 J100 comparison, Ttbar (50000)


Signal distribution

J100

HT250

HTC200

HT200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

HTC200 HT250 J100 comparison, GttL900 (10000)


Signal distribution

J100

HT250

HTC200

HT200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

HT250

HTC200

HT200



L1 HT triggers & EF Fat Jet selection

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Applying EF

J100 + EF

HT250 + EF

HTC200 + EF

HT200 + EF


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Applying EF

J100 + EF

HT250 + EF

HTC200 + EF

HT200 + EF


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Applying EF

J100 + EF

HT250 + EF

HTC200 + EF

HT200 + EF


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Applying EF

J100 + EF

HT250 + EF

HTC200 + EF

HT200 + EF



L1∑

ET > 100, J100

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

sum ET 2 close jets DR < 1.0, zprime1000 (20000)

Efficiency for ET sum cut for close jets

Signal distribution

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets

sum ET 2 close jets DR < 1.0, zprime1000 (20000)

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

sum ET 2 close jets DR < 1.0, Ttbar (50000)


Signal distribution

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets

sum ET 2 close jets DR < 1.0, Ttbar (50000)

jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

sum ET 2 close jets DR < 1.0, GttL900 (10000)


Signal distribution

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets



L1∑

ET > 100, J100 & EF Fat Jet selection

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

sum E_T 2 close jets DR < 1.0, zprime1000 (20000)

Efficiency for ET sum cut for close jets + EF

(EF only)

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets

sum E_T 2 close jets DR < 1.0, zprime1000 (20000)

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

sum E_T 2 close jets DR < 1.0, Ttbar (50000)


(EF only)

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets

sum E_T 2 close jets DR < 1.0, Ttbar (50000)

jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

sum E_T 2 close jets DR < 1.0, GttL900 (10000)


(EF only)

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only)

HTC200

J100

> 100T

E∑2 close jets

> 100T

E∑3 close jets



L1∑

ET > 100,∑

ET > 120

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 cuts on max ET sum of 1, 2 DR<1.0 and 3 DR<1.0 jets, ET > 20, a10, zprime1000 (20000)

Efficiency for selecting close jets, cutting on summed ET + EF

Signal distribution > 120

T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 cuts on max ET sum of 1, 2 DR<1.0 and 3 DR<1.0 jets, ET > 20, a10, Ttbar (50000)



T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

L1 cuts on max ET sum of 1, 2 DR<1.0 and 3 DR<1.0 jets, ET > 20, a10, GttL900 (10000)



T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200



L1∑

ET > 100,∑

ET > 120 & EF Fat Jet selection

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only) > 120

T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only) > 120

T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only) > 120

T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only) > 120

T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200



Summary 1 - efficiency(fat jet pT)

The baseline L1 J100 is compared to L1 HTC200 and two other selections thatinclude close-by jet requirements; OR-ing this selection to the baseline recoversinefficiencies of the individual selections.

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets



Summary 1 - including EF Fat Jet selection

“Dominant” inefficiencies after a EF j360 a10tclcw selection.

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1


Efficiency for close jet selection + EF

(EF Only)

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF Only)

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF Only)

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF Only)

J100

HT200

2 close & HT200

2 close & HT200 OR J100 > 100

T E∑2 close jets



Summary 2 - efficiency(fat jet pT)

The baseline L1 J100 is compared to L1 4J20, HT200, HTC200 and the ET sumof up to two close-by jets. The 4-jet selection leads to large inefficiencies inevents without a large jet multiplicity.

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

J100

4J20

HT200

HTC200 > 100

T E∑2 close jets



Summary 2 - including EF Fat Jet selection

“Dominant” inefficiencies after a EF j360 a10tclcw selection.

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only)

J100 + EF

4J20 + EF

HT200 + EF

HTC200 + EF > 100

T E∑2 close jets


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only)

J100 + EF

4J20 + EF

HT200 + EF

HTC200 + EF > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only)

J100 + EF

4J20 + EF

HT200 + EF

HTC200 + EF > 100

T E∑2 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only)

J100 + EF

4J20 + EF

HT200 + EF

HTC200 + EF > 100

T E∑2 close jets



Event properties per sample


Total Offline HT histogram

0 200 400 600 800 1000 1200 1400 1600 1800 20000

200

400

600

800

1000

1200

distribution of Offline HT, zprime1000distribution of Offline HT, zprime1000

0 200 400 600 800 1000 1200 1400 1600 1800 20000

1000

2000

3000

4000

5000

6000

7000

distribution of Offline HT, Ttbardistribution of Offline HT, Ttbar

0 200 400 600 800 1000 1200 1400 1600 1800 20000

100

200

300

400

500

600

700

800

distribution of Offline HT, GttL900distribution of Offline HT, GttL900

0 200 400 600 800 1000 1200 1400 1600 1800 20000

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

distribution of Offline HT, GttL1distribution of Offline HT, GttL1


offline number of jets

0 2 4 6 8 10 12 14 16 18 200

1000

2000

3000

4000

5000

Number of jets on Offline, zprime1000Number of jets on Offline, zprime1000

0 2 4 6 8 10 12 14 16 18 200

1000

2000

3000

4000

5000

6000

Number of jets on Offline, TtbarNumber of jets on Offline, Ttbar

0 2 4 6 8 10 12 14 16 18 200

1000

2000

3000

4000

5000

6000

7000

Number of jets on Offline, GttL900Number of jets on Offline, GttL900

0 2 4 6 8 10 12 14 16 18 200

1000

2000

3000

4000

5000

6000

7000

8000

9000

Number of jets on Offline, GttL1Number of jets on Offline, GttL1


offline number of fat jets

0 2 4 6 8 10 12 14 16 18 200

2000

4000

6000

8000

10000

12000

Number of fat jets on Offline, zprime1000Number of fat jets on Offline, zprime1000

0 2 4 6 8 10 12 14 16 18 200

1000

2000

3000

4000

5000

6000

7000

8000

Number of fat jets on Offline, TtbarNumber of fat jets on Offline, Ttbar

0 2 4 6 8 10 12 14 16 18 200

1000

2000

3000

4000

5000

6000

7000

8000

Number of fat jets on Offline, GttL900Number of fat jets on Offline, GttL900

0 2 4 6 8 10 12 14 16 18 200

2000

4000

6000

8000

10000

Number of fat jets on Offline, GttL1Number of fat jets on Offline, GttL1


Eta histograms

-5 -4 -3 -2 -1 0 1 2 3 4 50

200

400

600

800

1000

1200

1400

jet eta, zprime1000jet eta, zprime1000

-5 -4 -3 -2 -1 0 1 2 3 4 50

500

1000

1500

2000

2500

3000

jet eta, Ttbarjet eta, Ttbar

-5 -4 -3 -2 -1 0 1 2 3 4 50

200

400

600

800

1000

1200

jet eta, GttL900jet eta, GttL900

-5 -4 -3 -2 -1 0 1 2 3 4 50

1000

2000

3000

4000

5000

6000

jet eta, GttL1jet eta, GttL1


Jet orientations on different trigger levels, Gtt;L900

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

eta-3 -2 -1 0 1 2 3

phi

-3

-2

-1

0

1

2

3

eta vs phi for an event, 3R10: 0, 2R06: 0eta-phi1.pdf

Entries 8Mean x -0.1Mean y -0.1473RMS x 0.5916RMS y 2.047

eta phi for jets in 1 event

L1

Offline jets

Offline Fat jets

eta vs phi for an event, 3R10: 0, 2R06: 0

Triggerbit for 2/3 close jets in title



0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

eta-3 -2 -1 0 1 2 3

phi

-3

-2

-1

0

1

2

3


Entries 5Mean x -0.48Mean y -0.4712RMS x 0.6765RMS y 1.65


L1

Offline jets

Offline Fat jets





eta-3 -2 -1 0 1 2 3

phi

-3

-2

-1

0

1

2

3eta-phi3.pdf

Entries 7Mean x -0.6571Mean y 0.3085RMS x 0.798RMS y 1.886

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1eta-phi3.pdfEntries 7Mean x -0.6571Mean y 0.3085RMS x 0.798RMS y 1.886



L1

Offline jets

Offline Fat jets




0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

eta-3 -2 -1 0 1 2 3

phi

-3

-2

-1

0

1

2

3


Entries 6Mean x -0.12Mean y 0.1571RMS x 0.5154RMS y 2.216


L1

Offline jets

Offline Fat jets





eta-3 -2 -1 0 1 2 3

phi

-3

-2

-1

0

1

2

3eta-phi5.pdf

Entries 6Mean x 0.6333Mean y -0.589RMS x 0.725RMS y 2.109

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1eta-phi5.pdfEntries 6Mean x 0.6333Mean y -0.589RMS x 0.725RMS y 2.109



L1

Offline jets

Offline Fat jets



Summary

Summary:

We have looked at L1 seeds for fat jet triggers; more specifically, atalternatives to the baseline L1 single jet trigger items:

a selection of L1 jet close-by pairs with high∑

ET , ORed to thebaseline selection;HT constructed from 20GeV L1 jets with |η| < 3.2 or < 2.5,multijet triggers,the two above, including requirements on the distance between jets.

HT(C)200 seems to have the best overall performance in the models wehave considered and for the assumed threshold of 360GeV for the EFselection.

Other selections can recover inefficiencies if lower EF thresholds can beallowed; e.g. if EF fat jet selections are made more robust to pile-up.

Further possible steps:

Investigate the pile-up robustness for the various selections;

Investigate L1 processing limitations when searching for combinations ofclose-by jets.


Backup


L1∑

ET > 100,∑

ET > 120∆R = 1.0

Comparison of cutting 1.2 or 1.0

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

HTC200



L1∑

ET > 100,∑

ET > 120∆R = 1.2

Comparison of cutting 1.2 or 1.0

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

applying HTC200


jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

applying HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

applying HTC200


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1




T E∑3 close jets

> 100T

E∑3 close jets

> 120T

E∑2 close jets

> 100T

E∑2 close jets

applying HTC200



Correlation between L1 L2 EF, GttL900

0

20

40

60

80

100

pT names L20 100 200 300 400 500 600 700 800

pT n

ames

L1

0

100

200

300

400

500

600

700

800

Scatter of leading ET no selection for 5000 events, L1 vs L2

0.97523591288

Scatter of leading ET no selection for 5000 events, L1 vs L2

0

2

4

6

8

10

12

14

16

18

pT names EF0 100 200 300 400 500 600 700 800

pT n

ames

L2

0

100

200

300

400

500

600

700

800

Scatter of leading ET no selection for 5000 events, L2 vs EF

0.666858029371

Scatter of leading ET no selection for 5000 events, L2 vs EF

0

10

20

30

40

50

pT names OL0 100 200 300 400 500 600 700 800

pT n

ames

EF

0

100

200

300

400

500

600

700

800

Scatter of leading ET no selection for 5000 events, EF vs OL

0.983869758058

Scatter of leading ET no selection for 5000 events, EF vs OL


Online/Offline Comparison; L1 Inv. mass vs Fat Jet pT

0

50

100

150

200

250

300

350


Onl

ine

Inva

riant

Mas

s

0

100

200

300

400

500

600

700

800

Online invariant mass of 2 close by jets vs Offline leading pT, a10, zprime1000 (20000)


Online invariant mass of 2 close by jets vs Offline leading pT, a10, zprime1000 (20000)

0

1000

2000

3000

4000

5000

6000


Onl

ine

Inva

riant

Mas

s

0

100

200

300

400

500

600

700

800

Online invariant mass of 2 close by jets vs Offline leading pT, a10, Ttbar (50000)


Online invariant mass of 2 close by jets vs Offline leading pT, a10, Ttbar (50000)

0

50

100

150

200

250


Onl

ine

Inva

riant

Mas

s

0

100

200

300

400

500

600

700

800

Online invariant mass of 2 close by jets vs Offline leading pT, a10, GttL900 (10000)



0

50

100

150

200

250

300


Onl

ine

Inva

riant

Mas

s

0

100

200

300

400

500

600

700

800





Online/Offline Comparison; L1 HT vs HT

0

10

20

30

40

50

60

70

Offline HT0 100 200 300 400 500 600 700 800

Onl

ine

HT

0

100

200

300

400

500

600

700

800

Online HT vs Offline HT, zprime1000 (20000)


Online HT vs Offline HT, zprime1000 (20000)

0

100

200

300

400

500

Offline HT0 100 200 300 400 500 600 700 800

Onl

ine

HT

0

100

200

300

400

500

600

700

800

Online HT vs Offline HT, Ttbar (50000)


Online HT vs Offline HT, Ttbar (50000)

0

5

10

15

20

25

30

35

40

Offline HT0 100 200 300 400 500 600 700 800

Onl

ine

HT

0

100

200

300

400

500

600

700

800

Online HT vs Offline HT, GttL900 (10000)



0

5

10

15

20

25

Offline HT0 100 200 300 400 500 600 700 800

Onl

ine

HT

0

100

200

300

400

500

600

700

800





L1∑

ET > 120, J120

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

cluster 2 close jets DR < 1.0, zprime1000 (20000)


Signal distribution

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets

cluster 2 close jets DR < 1.0, zprime1000 (20000)

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

cluster 2 close jets DR < 1.0, Ttbar (50000)


Signal distribution

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets

cluster 2 close jets DR < 1.0, Ttbar (50000)

jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

cluster 2 close jets DR < 1.0, GttL900 (10000)


Signal distribution

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



Signal distribution

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets



L1∑

ET > 120, J120 & EF Fat Jet selection

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

2 close jets DR < 1.0, zprime1000 (20000)


(EF only)

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets

2 close jets DR < 1.0, zprime1000 (20000)

jet pT (GeV)0 100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

2 close jets DR < 1.0, Ttbar (50000)


(EF only)

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets

2 close jets DR < 1.0, Ttbar (50000)

jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1

2 close jets DR < 1.0, GttL900 (10000)


(EF only)

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets


jet pT (GeV)100 200 300 400 500 600

Effi

cenc

y

0

0.2

0.4

0.6

0.8

1



(EF only)

HTC200

J120

> 120T

E∑2 close jets

> 120T

E∑3 close jets



Multijet Rates

Trigger 8 TeV Data 8 TeV MC 14 TeV MC MC Scaling Data scaled to 14 TeV

4J20 0.8+- 0.0 0.7+- 0.1 4.2+- 0.3 (5.7+-0.9) 4.5+- 0.7

Unique 1.5+- 0.0 0.4+- 0.1 3.0+- 0.3 (7.4+-1.6) 3.8+- 0.8

HTC200 1.2+- 0.0 0.9+- 0.1 2.8+- 0.3 (3.3+-0.5) 3.8+- 0.6

Unique 0.0+- 0.0 0.0+- 0.0 0.0+- 0.0 (0.0+-0.0) 0.0+- 0.0

HT200 1.5+- 0.0 1.2+- 0.1 4.0+- 0.3 (3.3+-0.5) 4.8+- 0.7

Unique 0.2+- 0.0 0.2+- 0.1 1.0+- 0.2 (3.9+-1.2) 0.9+- 0.3

========================================================================================

Total 2.0+- 0.0 1.6+- 0.2 7.1+- 0.4 (4.3+-0.5) 8.6+- 1.0


1J100 1.7+- 0.0 1.4+- 0.1 4.6+- 0.3 (3.3+-0.4) 5.8+- 0.7

Unique 0.6+- 0.0 0.4+- 0.1 1.7+- 0.2 (3.9+-0.9) 2.2+- 0.5

4J20 0.8+- 0.0 0.7+- 0.1 4.2+- 0.3 (5.7+-0.9) 4.5+- 0.7

Unique 0.5+- 0.0 0.4+- 0.1 3.0+- 0.3 (7.4+-1.6) 3.7+- 0.8

HT200 1.5+- 0.0 1.2+- 0.1 4.0+- 0.3 (3.3+-0.5) 4.8+- 0.7

Unique 0.2+- 0.0 0.1+- 0.0 0.6+- 0.1 (5.0+-2.1) 0.8+- 0.3

HT300 0.3+- 0.0 0.3+- 0.1 1.1+- 0.2 (3.6+-1.0) 1.0+- 0.3

Unique 0.0+- 0.0 0.0+- 0.0 0.0+- 0.0 (0.0+-0.0) 0.0+- 0.0

=======================================================================================

Total 2.5+- 0.0 2.1+- 0.2 8.8+- 0.5 (4.3+-0.4) 10.7+- 1.1


Estimation for∑

ET rate


2J55:DR10-J55-J55 0.0+- 0.0 0.0+- 0.0 0.1+- 0.1 (8.7+-9.5) 0.4+- 0.4

Unique 0.0+- 0.0 0.0+- 0.0 0.0+- 0.0 (0.0+-0.0) 0.0+- 0.0

J65_2J45:DR10-J65-J45 0.1+- 0.0 0.0+- 0.0 0.2+- 0.1 (6.1+-4.9) 0.5+- 0.4

Unique 0.0+- 0.0 0.0+- 0.0 0.0+- 0.0 (0.0+-0.0) 0.0+- 0.0

J75_2J35:DR10-J75-J35 0.1+- 0.0 0.1+- 0.0 0.2+- 0.1 (3.1+-1.9) 0.3+- 0.2

Unique 0.0+- 0.0 0.0+- 0.0 0.0+- 0.0 (0.0+-0.0) 0.0+- 0.0

J85_2J25:DR10-J85-J25 0.1+- 0.0 0.1+- 0.0 0.2+- 0.1 (2.6+-1.4) 0.3+- 0.1

Unique 0.0+- 0.0 0.0+- 0.0 0.0+- 0.0 (0.0+-0.0) 0.0+- 0.0

J95_2J20:DR10-J95-J20 0.1+- 0.0 0.1+- 0.0 0.2+- 0.1 (2.8+-1.6) 0.2+- 0.1

Unique 0.0+- 0.0 0.0+- 0.0 0.0+- 0.0 (0.0+-0.0) 0.0+- 0.0

=======================================================================================================

Total 0.2+- 0.0 0.1+- 0.0 0.3+- 0.1 (2.9+-1.5) 0.5+- 0.2


The End


Education

Trigger Workshop material CERN Anton Osika