namespace smear {…}

Post on 22-Feb-2016

71 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

DESCRIPTION

namespace Smear {…}. Michael Savastio. Tree Format. 1 to 1. 1 to 1. 1 to 1. Currently ParticleS stores E, p, theta, phi, pz , pT ONLY !. The Device Class. Smears exactly 1 particle-wise variable ( E,p,theta,phi,pz,pT ) using up to 2 of these variables to parametrize the smearing. - PowerPoint PPT Presentation

TRANSCRIPT

namespace Smear {…}Michael Savastio

Tree Format

Original TreeEvent•KS=1 Particles•E,p,theta,phi•KS!=1 Particles

1 to 1

1 to 1

1 to 1

Currently ParticleS stores E, p, theta, phi, pz, pT ONLY!

The Device ClassSmears exactly 1 particle-wise variable (E,p,theta,phi,pz,pT) using up to 2 of these variables to parametrize the smearing.

Each device has it’s own “acceptance” in (E,p,theta,phi) space.

ThetaP

E

pp

E

phi

IP

Particles passing through here will

have smeared

p and theta

Particles passing through here will have smeared E only

These particles will be smeared

only by “innermost” (last added)

device EventSWill output only 1 EventS with only 1 instance of each particle.

Building a “Detector” (1)Device EMCal_front;EMCal_front.SetGenre(1); //this detects photons/leptonsEMCal_front.SetSmearedKinematics(kE); //set kinematics to be smeared by deviceEMCal_front.SetParametrization("0.18*sqrt(E)"); //set parametrization EMCal_front.Accept.SetTheta(0.,pi/4.); //set acceptance in theta//similarly, you can do SetAcceptPhi(min,max), SetAcceptE(min,max) and

SetAcceptP(min,max)

Device EMCal_back;EMCal_back.SetGenre(1); EMCal_back.SetSmearedKinematics(kE);EMCal_back.SetParametrization("0.25*sqrt(E)");EMCal_back.Accept.SetTheta(3.*pi/4.,pi); //set acceptance in theta

Device HCal;HCal.SetGenre(2); //this detects hadronsHCal.SetSmearedKinematics(kE);HCal.SetParametrization("0.35*sqrt(E)");

Building a “Detector” (2)Device ThetaD; //by default devices don't care whether hadron/lepton/gauge bosonThetaD.SetSmearedKinematics(kTheta);ThetaD.SetParametrization(”sqrt(9.e-8+pow(9.e-4/P,2))/sqrt(sin(theta))");//can use 2d parametrizations

Device Tracker; Tracker.SetSmearedKinematics(kP);Tracker.SetParametrization("0.0085*P+0.0025*P*P");

Detector Test;Test.AddDevice(EMCal_front); //add the EMCal to the detectorTest.AddDevice(EMCal_back);Test.AddDevice(HCal);Test.AddDevice(ThetaD);Test.AddDevice(Tracker);Test.SetPID(true); //turn on PID (still limited but works!)Test.SetEventKinematicsCalculator("NM JB DA"); //set how to calculate event kinematcs. Can use scattered electron (null momentum approximation NM), Jacquet-Blondel or Double Angle.

Now in root:gSystem->Load(“BuildTree.so”);.L MyDetector.cxxSmearTree(MyDetector(),”filename”)

To view in interpreter, for example:TTree t;t.AddFriend(“EICTree”,”originaltree.root”)t.AddFriend(“Smeared”,”smearedtree.smear.root”)t.Draw(“EICTree.x:Smeared.x”,”EICTree.QSquared>10”)

Currently Available Smear::Detector Setups

ZEUS

• EMCal "0.18*sqrt(E)"• HCal "0.35*sqrt(E)"• Tracking "0.0085*p+0.0025*p*p"• Polar Angle "0.0005*p+0.003"

STAR

• EMCal "0.015*E+0.14*sqrt(E)"• Tracking "0.005*p+0.004*p*p"• Polar Angle "sqrt(0.0003*0.0003+(0.0009/p,2)^2)/sqrt(sin(θ))"

All with default 4*pi acceptance.

EZEUS ZEUS

STAR STAR

Pythia 20x250

pZEUS ZEUS

STAR STAR

Pythia 20x250

θZEUS ZEUS

STAR STAR

Pythia 20x250

Event-Wise Kinematics

ZEUS STAR

Methods Available:NM (using scattered electron in the Null Momentum approximation)JB (Jacquet-Blondel, using hadronic system)DA (Double Angle, using hadronic system)

Pythia 20x250

Particle ID

p+K+K-

unidentified

π+π-

PID not Generated:Momentum outside range of validity

Based on HERMES RICH P-matrices (shown on grid above)

TO DO• Thoroughly test event-wise kinematics (NM, JB and DA).

Determine appropriate phase space cuts.• Fix anomalous (but very minor) stability issues. These may be

CINT memory management issues.• Overhaul PID to be far more flexible. Currently based on HERMES

RICH format. Software-wise, this is the most labor intensive remaining task.

• Improvements and testing for exception handling (especially with limited acceptance). Make more educated decisions about default conventions.

• Gather parametrizations (help welcome!).• Electron bremsstrahlung

UPDATE• Event Kinematics are fixed, all methods now work reliably (using

scattered electron, or hadronic system).• Improved exception handling, especially for event kinematics.• Particle ID has been completely rewritten and is now much more

versatile.• Specialized “Devices” (tracking, calorimetry).• Arbitrary acceptance cuts.• Acceptance of specific particles.• Smearing of arbitrary functions of particle kinematics.• Many improvements to make scripting simpler.

Building a “Detector” (Update,1)EMCalorimeter EMCal(0.18);

EMCal.Accept.SetTheta(0.,pi/4.); //set acceptance in thetaEMCal.Accept.AddZone();EMCal.Accept.SetTheta(3.*pi/4.,pi,1);//similarly, you can do Accept.SetPhi(min,max), Accept.SetE(min,max) and

Accept.SetP(min,max)

HCalorimeter HCal(0.35); //this has parametrization “0.35*sqrt(E)”

Device ThetaD(kTheta,”sqrt(9.e-8+pow(9.e-4/P,2))/sqrt(sin(theta))”);

Device Tracking(kP,”0.001*P+0.001*P*P”);

ParticleID Ident(“Pmatrixfile.dat”);Ident.GetAcceptanceFromDevice(EMCal); //ParticleID is now declared like a device,

and has its own acceptance. You can use arbitrarily many.

Detector Example;Example = Example << EMCal << HCal << ThetaD << Tracking << Ident;Example = Example << “NM JB DA”; //use all 3 event methods

Some New FeaturesDevious Arbitrary(“1./P”,”0.0085*P+0.0025*P*P”); //the Devious class allows you to smear a function of kinematics. This smears 1/pArbitrary.Accept.SetPt(0.,50.); //this cut requires p_{T}\in[0,50] GeVArbitrary.Accept.AddCustomAcceptance(“P*sin(theta)”,0.,50.); //equivalent to the above

Arbitrary.Accept.AddParticle(321); //now this accepts K^+ ONLYArbitrary.Accept.AddParticle(-321); //now it accepts K^+ and K^- only

Tracker Track;Track.SetRadii(0.01,3.); //this is a tracker with inner radius 1cm, outer radius 3mTrack.SetLength(6.); //and length 6mTrack.SetDimensions(0.01,2.,6.); //this is equivalent to the above two lines

Device CrazyExample(kE,”0.01*E”);CrazyExample.SetDistribution("pow(sin([1]*x+[0]),2)"); //smears according to

arbitrary distribution. [0] is original value of E and [1] is given by parametrization.

Example = Example << Arbitrary << Track;

Now with extensive documentation:https://wiki.bnl.gov/eic/index.php/Namespace_Smear

Event Kinematics: Proof of Concept

x STAR20x100

Q2

STAR20x100

Why it’s so screwed upSTAR20x100Smearing goes as p2

What Now?• Where does Smearing end and

GEANT begin?• What can Smear be used for, what

can it not be used for?• From now on, code will be

maintained by Thomas Burton.

top related