Trigger & Analysis

Trigger & Analysis

Avi YagilUCSD

14-June-2007 HCPSS - Triggers & Analysis Avi Yagil 2

Table of Contents

• Introduction– Rates & cross sections– Beam Crossings– What do we trigger on? – Trigger Table (example)

• Trigger terminology– Trigger paths – Prerequisites– Volunteers– Overlaps

• Primary Data Sets• Online streams

– Why? How many? – Express stream

• Data flow overview

• Analysis– Physics signature– What is my trigger?– Where do I find the data?

• Experiment dependent• Sample location & content

– An aside: Data formats• RAW, RECO, AOD, ESD, TAG…• What do I need?• What can I afford?• Where is it?

– Efficiency measurements• backup triggers, samples

– How fast can I run over my sample?

• Vertical and Horizontal skims

• An Example


Rates…

Cross Section: σ(pp) = 70 mbInteraction Rate, R = 7x108Hz

Bunch Spacing: t= 25 ns

Physics Cross Sections:•Inelastic: 109Hz •W→l ν: 102Hz •t t production: 10 Hz•Higgs (100 GeV/c2): 0.1 Hz •Higgs (600 GeV/c2): 10-2Hz

•250 GeV ET Jets - 1kHz

Rejection needed: 1:1010-11


Time to think?Bunch Spacing…

QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.

Less time to think, MUCH more to think about…


From P. Sphicas, Trigger @LHC talk - i




From P. Sphicas, Trigger @LHC talk - ii




What do we trigger on? General considerations

• Basics:– Hadron colliders produce mostly low momentum hadrons – Most interesting Physics has signatures involving large transverse

energy (ET) particles.

==> Require high ET on reconstructed particles

• Basic objects used for trigger:– Electrons, Photons, Muons, Jets, Missing Et

• Approximate thresholds and associated rates:– Single muon with PT>20 GeV - 10kHz

• Double muon PT>6 GeV - 1kHz

– Single em cluster with ET>30 GeV - 10-20 kHz• Double em cluster with ET>20 GeV - 5 kHz

– Single jet with ET>300 GeV - 0.2-0.4 kHz

The more complexity one adds to the a signature - the lower the rate, and hence the required threshold

The more complex the signature is - the harder it is to understand…


HLT Trigger Table - Example (CMS)


Evolution of Ana goals, implications…

• Detector commissioning– Timing– Cabling– …

• Comish. Low level reco.– Val. hits, segments, clusters– Calibrate– Noise, 0-suppression

• Comish. Object reco.– Establish e, m, jet… IDs– Define basic eff. - Z signal– Measure rejection– Isolation!

• Counting experiments

• Cross section measurements

Raw DataLatch-all (no 0-suppression)Basic detector DQM softwareLocal/global daq

Raw and RecHitsLocal geometry, DB accessLow level reconstruction softwareEstablish 0-suppression asap!

Raw, RecHits and RecoDetector geometry, DB, Ca/Al…Full reconstruction software (HLT)Basic trigger, Access to data

Absolute efficiencies (tracking…), MC scale factors

Luminosity


Trigger Path

• The “name” of the sequence of requirements that resulted in an event being recorded:– L1 accept– HLT confirmation

• Object type(s) and threshold(s), cut(s) (things like: emu_20_20_iso_met_50 or maybe hipt_emu_met)

• There has to be a bit for each one. • Defines how events are classified (immutably!)

– Implication: complete exploration of all L1 accepts


Pre-requisites, Volunteers

• Pre-Requisite: Only muons that have a L1 accept are pursued in the HLT. Moreover, only that region is reconstructed.

• Volunteer: A muon “found” in the HLT, without a corresponding L1 accept cannot result in a trigger– this cannot happen if we do “seeded” (on L1 muon track)

reconstruction in HLT, but can happen if we do global reconstruction.


Overlaps

• A frequently asked question is: – What is the overlap between Primary Data Sets?

• Recall that the overlap between data sets is a result of the overlap between trigger paths.

How big is it? How big should it be?

• Up front overlap - express stream. – An experiment decides how big it is (~10%)– Completely controlled

• Additional duplication occurs due to Physics but is very small in comparison.

• An e-mu-met event will (hopefully) go to three data setsand we want them to!

• But (sadly) there are few of these…– Can have huge overlaps - need good design of trigger table

• One can have as big an overlap as one can afford.


Primary Data Sets

• It is easy to foresee about 50 primary datasets to be identified during the HLT processing, each following a strict trigger path (L1 and the appropriate HLT confirmation of it). – How many will be needed?– There is danger in too many, and also in too few…– Serious Physics Coordination issue!

• As a Data Management artifact in the Online farm, these 50 primary datasets may be grouped into about 10 streams constructed to be roughly of similar sizes and contain more or less related primary datasets.

• These streams have no (physics) significance outside of the Online-to-Offline data management context.


Streams (online, intermediate)

• One can imagine the following example of streams definition: – Jet data stream– electron/photon– Muon, tau/MET – calibration – min bias– …

• Where for example the ele/photon stream will contain the primary datasets for inclusive electron as well as the di-electron triggers of various thresholds (some may be pre-scaled) and the closely related photon triggers.

• These primary datasets should contain back-up triggers (primary datasets) to assist in understanding the efficiency and rejection of the main ones (they will be usually looser and pre-scaled).


End-to-End example

• Basic characteristic:– Low production rate signal– Large background– Many, small channels to assemble

• Search for top quark– Pick a signature: (di-leptons,

lep+jets…)pp-->tt+X and tt-->WbWbWhere: W-->e W-->Final States:

-lepton pair, two b jets- lepton, 2 b-jets, 2 q-jets

• Why limit oneself to only leptonic W decays?

– Why not look at all W and Z decays? (leptonic rates very low)

– …

>> It’s the rates!

• Analysis steps:– Find corresponding

triggers:• Inclusive electron, muon• Di-electron, muon• Maybe others (bb+met),

but much harder (why?)– Define the samples to use:

• W & Z samples (to e and mu)

– Which level of information is needed?

• 4 vectors?• RAW data?• Something in between??

– Where is it to be found?– How big is it? How long

does it take to run over it?– Can it be “reduced”?


Notes from the search for top

Triggers & Data sets used

• Relied primarily on inclusive electron and inclusive muon triggers

• Many backup triggers needed:– Various inclusive jet

samples for jet corrections– Gamma-jet sample for E-

scale– W-notrack trigger for

tracking efficiency– Jet50 sample for b-tagging

MC scale factor– Many many more

A “family” of analyses

• Three main classes– Di-lepton – Lepton+jets– All hadronicAll predicated on the basic

trigger path, or primary data set.

• Finer split within each– Electron, Muon or jets as

“anchor leg” (trigger requirement)

– additional features (e.g. vertex b-tagging, SLT…)


Sample definitions:Inclusive Electron/Muon Samples


W selection


Jets and Met in top events


b jets from top decays


“Soft” leptons in top decays




Documents

Trigger & Analysis