Electron Identification Supriya Das Gesellschaft fr Schwerionenforschung mb
Electron Identification Supriya Das Gesellschaft für Schwerionenforschung mb. H (GSI) 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden
Motivation Till the last collaboration meeting we had the concept and results from electron identification. But, the software was standalone i. e. one code for electron identification and pair analysis. Requirement of self sufficient software to provide identified electrons to be used in further analysis, mixed event generation etc. Needs to be flexible/user friendly in terms of Ø changing cuts for different detectors, Ø decision of using different cuts for identification Ø ability to selectrons identified by one or more detectors Requirement to disentangle the real and MC world. 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 2
Software Task : Cbm. Dilepton. Assign. Pid Input : All arrays from reconstucted data (no MC) Output : 1. Collection of identified tracks (Cbm. Dilepton. Track. Real) Global track index, chi 2 primary, momentum, is. Rich, is. Trd, is. Tof 2. Histograms before and after cuts Constructor contains all default cut values and switches User has to provide user defined values in macro (see later) [ no cut on chi 2 primary, can be used for secondary electron analyses] Task : Cbm. Dilepton. Assign. MCid Input : All reco and mc arrays Output : 1. Collection of tracks after MC association (Cbm. Dilepton. Track. Sim) momentum, fake. Track, fake. Ring, MCPdg, MCMother. Id, MCMother. Pdg 2. Histograms for efficiency, purity, pion, suppression etc. This collection has one-to-one correspondence with previous one [ two parameters needed to be supplied, cut on chi – to select primary radial distance – to define RICH acceptance] 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 3
Block diagram of electron id Simulation Real Reconstructed event Cbm. Dilepton. Assign. Pid Cbm. Dilepton. Track. Real 27 th. September, 2007 MC information Cbm. Dilepton. Assign. MCid Cbm. Dilepton. Track. Sim Supriya Das, CBM Collaboration meeting, Dresden 4
Cbm. Dilepton. Assign. Pid – flow chart is. TRD 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden is. TOF 5
Rich. Pid – flow chart 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 6
Software (contd. ) Macro to run the electron id … • g. System. Load->(“lib. Dilepton”); • // Electron id assignment Cbm. Dilepton. Assign. Pid* task. Pid = new Cbm. Dilepton. Assign. Pid(1); • • • • Float_t a. Rich. Cuts[] = {1. , 0. 4, 0. , 130, 21. 85, 4. 35, 6. 17, 0. 14}; // Distance, Selection. NN, selection 2 D, Radial, NHit. Mean, NHit. Sigma, RMean, RSigma task. Pid->Set. Rich. Cuts(true, a. Rich. Cuts); //s. Momentum, s. Selection (true=NN, false=2 D) Float_t a. Trd. Cuts[] = {1. 5, 0. 95, 1. 1, 21. , 0. 9}; // Mom, Pid. Like. Low, Pid. Like. High, Pid. Wkn, Pid. Ann task. Pid->Set. Trd. Cuts(true, false, a. Trd. Cuts); //s. Like, s. Wkn, s. Ann task. Pid->Set. Tof. Cuts(true, 0. 01); //s. Momentum, Mass 2 f. Run->Add. Task(task. Pid); // MC Association, needed for simulated data stream Cbm. Dilepton. Assign. MCid* task. MCid = new Cbm. Dilepton. Assign. MCid(); f. Run->Add. Task(task. MCid); 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 7
Systematics • 5 e- and 5 e+ embedded into central Ur. QMD events for three different beam energies • Standard and low mass dilepton* set up used • 10, 000 events simulated • Standard cuts used for identification • ANN and Wkn methods for TRD are used – Likelihood needs tuning • Software used from aug 07 release Efficiency = identified true electrons / accepted electrons Pion suppression = pions identified as electrons / accepted pions * thin target, extended MVD and STS, 70% magnetic field 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 8
Identification criteria Identification by RICH Identification by TRD based on statistical analyses of total energy loss: Identification by TOF § ANN - e > 0. 8 § Wkn - e > 11 § Likelihood – 0. 95 < e < 1. 1 (any one or combination could be used) 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 9
Impurities in identified electrons Particles/event All : 5. 81 True e : 5. 35 Fake : 0. 09 p : 0. 06 p : 0. 014 Others : 0. 02 Au+Au @ 25 AGe. V Standard set up Combined information of RICH+TRD+TOF 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 10
Falsely matched pions Au+Au @ 25 AGe. V Standard set up Electron ring falsely matched with a nearby pion track 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 11
Pions mis-id as electrons Au+Au @ 25 AGe. V Standard set up True p Falsely matched p 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 12
Efficiency of electron identification Low mass dilepton set up Standard set up 25 AGe. V 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 13
Pion suppression Low mass dilepton set up Standard set up 25 AGe. V 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 14
Efficiency Low mass dilepton set up Standard set up 35 AGe. V 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 15
Pion suppression Low mass dilepton set up Standard set up 35 AGe. V 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 16
Efficiency Low mass dilepton set up Standard set up 15 AGe. V 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 17
Pion suppression Low mass dilepton set up Standard set up 15 AGe. V 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 18
Efficiency - energy dependence Efficiency from RICH only Standard set up Low mass dilepton set up Variation of efficiency with energy < 10 % 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 19
Energy dependence (contd. ) Standard set up Low mass dilepton set up Variation of efficiency with energy < 10 % 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 20
p-suppression – energy dependence Standard set up 27 th. September, 2007 Low mass dilepton set up Supriya Das, CBM Collaboration meeting, Dresden 21
Comparison between two TRD methods § The cuts need to be optimized to get the best efficiency/ pion suppression factor § Likelihood method needs more tuning 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 22
Summary and Outlook • New, user friendly software for electron identification for di-electron analyses is in place. • Results from systematic study of this package have been presented Ø Ø Ø Different set ups (standard, low mass dileptons) Different energies Different methods for TRD identification • Cuts could further be optimized to obtain the best efficiency and pion suppression. • Study of identification with the compact RICH to be done. 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 23
Summary & outlook (contd. ) • All cuts/switches can be modified from macro – user friendly. • The collections could be kept in root memory and accessed by the next task / could be stored in an intermediate file for later use (self sufficient, has all information). • Caveat : Can not provide cut-by-cut efficiencies possible solution : go to subtask level • All codes are in svn (in the last release) Ø Tasks and other classes in cbmroot/dilepton Ø Macro in cbmroot/macro/dilepton • Documentation exists in Wiki page, http: //cbm-wiki. gsi. de/cgi-bin/view/Cbm. Root/Cbm. Reconstruction/#Electron_Identification CBM-Note in preparation. 27 th. September, 2007 Supriya Das, CBM Collaboration meeting, Dresden 24
- Slides: 24