Si D Software Status Framework Overview DETECTOR DESCR

Si. D Software Status

Framework Overview DETECTOR DESCR Event Gen Simulation LCDD Reconstruction Analysis Visualization 3/20/2011 Compact APPLICATIONS DATA Pythia, WHIZARD, etc. Std. Hep SLIC LCIO Events Tracking, PFA, etc. LCIO PFOs LCSim AIDA JAS 3 (or any AIDA tool) Jeremy Mc. Cormick, SLAC – ALCPG 2011 2

Event Generation Ø Physics event generation is performed externally to general framework (e. g. slic & lcsim). Ø WHIZARD /w Guinea Pig, Pythia, or other legacy event generators (HERWIG, ISAJET). Ø Std. Hep binary format used for persistency. Ø Extensive sets of physics events and diagnostic events available via FTP. Ø Physics events at 100 Ge. V – 1 Te. V Ø Single particles at wide range of energies, angular distributions Ø ftp: //ftp-lcd. slac. stanford. edu/lcd/ILC/single. Particle/stdhep/ 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 3

LCIO Ø Common Data Model and API Ø Persistency using SIO (Simple Input Output) Ø C++, Java, Fortran, and Python Bindings Ø Generic data with LCIOGeneric. Objects Ø Supported in all ILC software frameworks (ILD, Si. D) Ø Allows “mix and match” of software components from different frameworks Ø Facilitates exchange of data between users (Diagrams from Frank Gaede, DESY) http: //lcio. desy. de 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 4

LCDD Ø Detailed geometry and detector root element <lcdd/> description format (XML) meta data <header/> Ø Full description of detector ID definitions Ø No additional information <iddict/> required at runtime detectors <sensitive_detectors/> Ø Embedded GDML section physics limits <limits/> Ø Binds XML elements to Geant 4 data detector regions <regions/> structures (G 4 Logical. Volume, etc. ) visualization settings <display/> Ø Fast load using SAX parser Ø Takes < 1 second to load complex <gdml/> GDML root element ILC detector geometry variable definitions <define/> Ø Possible to write by hand but not material definitions <materials/> recommended (verbose/redundant) geometric solids Ø Easiest way to generate is converting <solids/> from compact to LCDD using nested volumes <structure/> Geom. Converter. <fields/> magnetic fields http: //www. lcsim. org/software/lcdd/ 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 5

Ø Flexible command-line simulation tool with Geant 4 backend Ø Std. Hep event input Ø LCIO event output Ø Grid ready (no ext. DB or internet connection required) Ø Fully descriptive runtime geometry with no “magic numbers” (LCDD) Ø Extended particle set from Hep. PDT (SUSY, resonances, etc. ) with generic tracking + edep Ø Maintained by SLAC Ø Also used by non-ILC experiments SLIC LCDD Geometry <xml/> Std. Hep Events LCIO Events SLIC Frontend Geant 4 LCDD LCIO Hep. PDT Example Command Line Usage: slic –g geometry. lcdd –m stuff. mac –i events. stdhep –o output. slcio –l QGSP_BERT –r 1 slic –h # print help slic –m run. mac # run with macro only slic –n # run interactively 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 6

Geom. Converter Compact Detector Hep. Rep Wired GDML / LCDD SLIC, ROOT, Geant 4 <xml/> Geom. Converter Pandora XML Slic. Pandora (Pandora. PFANe w) Ø Compact subdetector types added as needed/requested. Java Runtime LCSim Convertor Example: java –jar Geom. Converter-bin. jar –o lcdd compact. xml mydet. lcdd 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 7

Compact Description Ø High level detector description format geared towards end users Ø Can be converted into various useful data formats and Java runtime objects Ø New detector types can be supported by adding the appropriate classes to Geom. Converter. Ø LCDetectors is data repository for compact detectors and their conditions. Example Detector: <lccdd/> <info/> <define/> <materials/> <limits/> <regions/> <display/> root element detector meta data variable definitions custom materials physics limits detector regions visualization settings <readouts/> detector description segmentation and IDs <fields/> magnetic fields <detectors/> <detector id=“ 1” name=“ECB” type=“Ecal. Barrel” readout=“ECBHits” vis=“ECBVis” type=“EM_BARREL”> <dimensions numsides=“ 12” rmin=“Ecal. Barrel_rmin” z=“Ecal. Barrel_zmin”/> <layer repeat=“ 30”> <slice material=“Silicon” thickness=“ 0. 25*mm” /> <slice material=“Tungsten” thickness=“ 2. 5*mm”/> </layer> </detector> 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 8

LCDetectors Ø CVS module containing detector data for Geom. Converter/LCSim. Ø At a minimum, each detector must have a compact. xml file. Ø Conditions are stored in a detector’s directory and can be read in using the LCSim Conditions. Manager. Ø Detector data is packaged into zip files and distributed via the WWW. Ø Sampling fractions Ø Tracking strategies Ø etc. Ø Makefile contains helper commands for generating data files, running common Geant 4 commands (such as geometric overlap checking), and for packaging data. Ø Detectors can also be packaged into a single jar file using Maven. Example Detector Data http: //www. lcsim. org/detectors 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 9

Trackers 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 10

Calorimeters 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 11

LCSim Ø Java-based reconstruction and analysis framework Ø Built with Maven Ø Easy build/installation/use Ø One build command gets you up and running. Ø Reads LCIO files Ø UIowa PFA Ø Seedtracker Ø Contrib module for nonessential user code contributions Ø Packaged into single jar file Ø Grid ready Ø XML steering files Ø JAS 3 Plugin Reconstruction tracking clustering event Geom. Conv conditions util lcio https: //confluence. slac. stanford. edu/display/ilc/org. lcsim 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 12

SLIC Pandora Ø Frontend to Pandora PFA that takes SLIC output plus LCSim Tracks Ø Outputs LCIO PFO collections Ø Runs on both digital and analog calorimeters Ø Optional Pandora. Monitoring for visualization Ø Work ongoing to improve performance https: //confluence. slac. stanford. edu/display/ilc/Running+SLIC+Events+in+Pandora+PFA+New Geometry Format <pandora. Setup> <detector> <calorimeters> <calorimeter> <id> <layers> <coil> <tracking> Command Line Syntax Pandora. Frontend geometry. xml settings. xml input. slcio output. slcio nevents nskip 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 13

Wired JAS 3 LCIO Record Source Ø Java-based analysis program Ø AIDA compliant Ø Can use plugins for data visualization and other purposes. Ø See talk by Tony Johnson LCSim Event Browser 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 14

Grid Ø All primary ALCPG software tools can be run on grid. (slic, lcsim, slic. Pandora) Ø Need to get grid certificate first to use and join ILC VO Ø Two main grid sites are accessible. Ø LHC grid (CLi. C) Ø Dirac toolkit with command line tools Ø Online monitoring tool Ø Python scripts for job submission Ø Job chaining Ø Large number of accessible machines (14 k) Ø Use lxplus machines for environment Ø Grid data catalog for file storage. Ø Fermi grid (Si. D) Ø Use Open Science Grid (OSG) Ø Uses Condor commands for job submission with JDF files Ø Good resource allocation (500 concurrent jobs) Ø Doc DB to access output: http: //docdb. fnal. gov/ILC/sid/ Ø Use ilcsim 2 machine for environment and virtual grid file system 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 15

Conclusions Ø Si. D software chain is complete and robust, from event generation through analysis. Ø Used for generation and analysis of ~20 million events for LOI Ø Also used by non-ILC collaborations Ø Work still remains to be done for the DBD. Ø Detailed simulation of RPC response Ø Track fitting Ø Data set generation for DBD using detailed detector models Ø Detector optimization Ø See other talks Ø Contributions welcome Ø Easy to add and integrate your own code or software packages Ø Java code is easily built/integrated/debugged. Ø Analysis & reconstruction code can be added to lcsim-contrib and immediately accessed in lcsim jobs.

Documentation & Links Ø ILC Wiki https: //confluence. slac. stanford. edu/display/ilc/Home Ø LCSim Website http: //www. lcsim. org/ Ø FTP for Event Data ftp: //ftp-lcd. slac. stanford. edu/lcd/ILC/ Ø SLIC FAQ https: //confluence. slac. stanford. edu/display/ilc/SLIC+FAQ Ø LCSim Job XML https: //confluence. slac. stanford. edu/display/ilc/lcsim+xml Ø Detector Data http: //www. lcsim. org/detectors Ø Sim. Dist Instructions https: //confluence. slac. stanford. edu/display/ilc/Simulation+Software+Distribution Ø Forums http: //forum. linearcollider. org/ 3/20/2011 Jeremy Mc. Cormick, SLAC – ALCPG 2011 17