ATC: ATLAS Center at ANL

User Tools

Site Tools

Trace:
community:refhepsim_analysis

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
community:refhepsim_analysis [2015/05/10 20:53] – asccommunity:refhepsim_analysis [2016/04/28 15:24] (current) – [HepSim Analysis Primer] asc
Line 2: Line 2:
    
  
-This analysis tutorial covers Java/Jython codding. Look at the [[http://atlaswww.hep.anl.gov/hepsim/description.php|main description]] for other topics.+This analysis tutorial covers Python codding on the Java platform (Jython). Look at the [[http://atlaswww.hep.anl.gov/hepsim/description.php|main description]] for other topics.
 For C++/ROOT, please refer [[asc:promc| ProMC web page]]. For C++/ROOT, please refer [[asc:promc| ProMC web page]].
  
-To ensure platform independence and a possibility to run using web browsers, validation programs are written in Jython, which is an implementation of the Python language on the Java platform.+To ensure platform independence and a possibility to run programs using web browsers [(Cross-platform validation and analysis environment for particle physics, S.V. Chekanov, I. Pogrebnyak, D. Wilbern, arXiv:1510.06638 http://arxiv.org/abs/1510.06638)], validation programs are written in Jython, which is an implementation of the Python language on the Java platform.
 Look at several links: Look at several links:
  
   * [[http://www.jython.org/jythonbook/en/1.0/| Jython book]]   * [[http://www.jython.org/jythonbook/en/1.0/| Jython book]]
-  * [[http://jwork.org/dmelt/wikidoc/doku.php?id=start| DataMelt Jython manual]]+  * [[http://jwork.org/dmelt/wikidoc/doku.php?id=start| DMelt manual (community version)]]
  
  
Line 16: Line 16:
   *  Using Java web start from the browser.   *  Using Java web start from the browser.
   * "hs-ide file.py" command to launch the editor (from the hs-tools package)   * "hs-ide file.py" command to launch the editor (from the hs-tools package)
-  * "hs-run file.py" command to run using command-line batch mode (from the hs-tools package) +  * "hs-run file.py" command to run using a command-line ("batch mode", from the hs-tools package) 
-  * [[http://jwork.org/dmelt/ |DataMelt  IDE]] +  * [[http://atlaswww.hep.anl.gov/asc/jas4pp/| Jas4pp]] 
 +  * [[http://jwork.org/dmelt/ |DMelt  IDE]] 
  
  
Line 64: Line 65:
 from proto import FileMC         from proto import FileMC        
 f=FileMC("data.promc")              # get input file from current directory (or URL) f=FileMC("data.promc")              # get input file from current directory (or URL)
 +des=f.getDescription()              # get description
 header = f.getHeader()              # to access information on this file header = f.getHeader()              # to access information on this file
 un=float(header.getMomentumUnit())  # conversion energy units un=float(header.getMomentumUnit())  # conversion energy units
Line 71: Line 73:
 The latter  2 values are needed to convert from integer presentation of 4-momenta, typically used for variable byte encoding to save space, The latter  2 values are needed to convert from integer presentation of 4-momenta, typically used for variable byte encoding to save space,
 to double values. The values are typically around 1000-10000, depending on importance of representing low-momentum with a better precision.  to double values. The values are typically around 1000-10000, depending on importance of representing low-momentum with a better precision. 
-The header file also stores some additional information. Look at Java API:+The header file also stores some additional information. Look at Java API
  
 +  * [[http://atlaswww.hep.anl.gov/asc/promc/hepsim/doc/api/ | HepSim API]] for FileMC
   * [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/promc/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for MC records   * [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/promc/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for MC records
   * [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/pronlo/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for NLO records   * [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/pronlo/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for NLO records
Line 118: Line 121:
            e= pa.getEnergy(j)/un     # energy            e= pa.getEnergy(j)/un     # energy
            m= pa.getMass(j)/un       # mass            m= pa.getMass(j)/un       # mass
-           p=LParticle(pa.getPx(j)/un,pa.getPy(j)/un,pa.getPz(j)/un,pa.getEnergy(j)/un,pa.getMass(j)/un)+           p=LParticle(px,py,pz,e,m) # fill it
 </code> </code>
 To access the PDG status code, us "pa.getPdgId(j) method. For stable particles, use the HEPEVT status code (the method "pa.getStatus(j)"). To access the PDG status code, us "pa.getPdgId(j) method. For stable particles, use the HEPEVT status code (the method "pa.getStatus(j)").
Line 201: Line 204:
                  print "pT of a leading jet =",jets[0].perp()," GeV"                   print "pT of a leading jet =",jets[0].perp()," GeV" 
 </code> </code>
-To build anti-kT jets, we use [[http://jwork.org/dmelt/api/doc.php/hephysics/jet/JetN2|JetN2 clustering algorithm]] implemented in Java. It is a similar to FastJet N*N algorithm and typically shows +To build anti-kT jets, we use [[http://jwork.org/dmelt/api/doc.php/hephysics/jet/JetN2|JetN2 clustering algorithm]] implemented in Java. 
 +The description of this algorithm is given in [[http://arxiv.org/abs/1510.06638]]. 
 + It is a similar to FastJet N*N algorithm and typically shows 
 a performance similar to the C++ analogue. Tests indicate that the JetN2 clustering algorithm is about twice slower that C++, but by a factor 20 faster than the traditional N^3 jet clustering algorithms. You can find benchmarking results in [[https://github.com/chekanov/hepjet/|HepJet]] github repository. a performance similar to the C++ analogue. Tests indicate that the JetN2 clustering algorithm is about twice slower that C++, but by a factor 20 faster than the traditional N^3 jet clustering algorithms. You can find benchmarking results in [[https://github.com/chekanov/hepjet/|HepJet]] github repository.
 Look at a typical example of jet clustering in the [[http://atlaswww.hep.anl.gov/hepsim/code.php?item=88&code=qcd_pythia8_ptbins_jets.py| qcd_pythia8_ptbins_jets.py code]]. Look at a typical example of jet clustering in the [[http://atlaswww.hep.anl.gov/hepsim/code.php?item=88&code=qcd_pythia8_ptbins_jets.py| qcd_pythia8_ptbins_jets.py code]].
Line 260: Line 265:
 stored inside ProMC data files: stored inside ProMC data files:
  
-  * [[http://jwork.org/dmelt/api/doc.php/jhplot/HPlot | HPlot]] - simple canvas to show graphs +===== Histograms and plots ===== 
-  * [[http://jwork.org/dmelt/api/doc.php/jhplot/H1D | H1D]] - 1D histogram + 
-  * [[http://jwork.org/dmelt/api/doc.php/jhplot/H1D | H2D]] - 2D histogram +  * [[http://jwork.org/dmelt/api/doc.php/jhplot/HPlot|HPlot]] - canvas to show X-Y data and histograms in 2D 
-  * [[http://jwork.org/dmelt/api/doc.php/hephysics/hepsim/PromcUtil|PromcUtil]] convenient method to fill arrays with particles from ProMc files+  * [[http://jwork.org/dmelt/api/doc.php/jhplot/HPlot|HPlot3D]] - canvas to show X-Y-Z data and histograms in 3D 
 +  * [[http://jwork.org/dmelt/api/doc.php/jhplot/H1D|H1D]] - 1D histogram 
 +  * [[http://jwork.org/dmelt/api/doc.php/jhplot/H1D|H2D]] - 2D histogram 
 +  * [[http://jwork.org/dmelt/api/doc.php/jhplot/P1D|P1D]] - X-Y container with support of 2-level errors 
 + 
 +See also other math and graphics classes in the [[http://jwork.org/dmelt/api/doc.php/jhplot/package-summary | jhplot package]].  
 +===== Lorentz particles and Jets ===== 
 + 
 +  * [[http://jwork.org/dmelt/api/doc.php/hephysics/hepsim/PromcUtil|PromcUtil]] convenient method to fill arrays with particles from ProMC files
   * [[http://jwork.org/dmelt/api/doc.php/hephysics/particle/LParticle|LParticle]] a HEP particle with the Lorentz transformations   * [[http://jwork.org/dmelt/api/doc.php/hephysics/particle/LParticle|LParticle]] a HEP particle with the Lorentz transformations
   * [[http://jwork.org/dmelt/api/doc.php/hephysics/jet/FastParticle|FastParticle]] a HEP particles with precomputed Et2,Eta,Phi for jet algorithms.    * [[http://jwork.org/dmelt/api/doc.php/hephysics/jet/FastParticle|FastParticle]] a HEP particles with precomputed Et2,Eta,Phi for jet algorithms. 
-  * [[http://jwork.org/dmlt/api/doc.php/hephysics/jet/JetN2|JetN2]] recommended kt-type jet clustering algorithms (kT, anti-kT) implemented in Java using N^2 approach, similar to FastJet +  * [[http://java.freehep.org/freehep-physics/apidocs/hep/physics/vec/package-summary.html|Physics vectors]] typical HEP physics vectors the Lorentz transformations 
-  * [[http://jwork.org/dmelt/api/doc.php/hephysics/jet/SCJet|SCJet]] traditional kt-type jet clustering algorithms (kT, anti-kT) implemented in Java using N^3 approach.+  * [[http://jwork.org/dmelt/api/doc.php/hephysics/jet/JetN2|JetN2]] recommended kt-type jet clustering ng algorithms (kT, anti-kT, CA) implemented in Java using N^2 approach, similar to the FastJet algorithm. Recommended for hadron collisions 
 +  * [[http://jwork.org/dmelt/api/doc.php/hephysics/jet/SCJet|SCJet]] traditional kt-type jet clustering algorithms (kT, anti-kT) for pp implemented in Java using N^3 approach (slow)  
 +  * [[http://java.freehep.org/freehep-physics/apidocs/hep/physics/jet/package-summary.html|Jets and event shapes]] traditional jet algorithms and event shapes for e+e- (Geneva, Jade, Durham jets) 
      
      
 Java implementation of the longitudinally invariant kT and anti-kT clustering algorithms uses the E-scheme to combine particles (p1+p2) and Eta-Phi space (not Rapidity-Phi). Also, Cambridge/Aachen jets are supported.   [[https://github.com/chekanov/hepjet/|HepJet on github]] shows  benchmarks of the Java implementation of the anti-KT with the original FastJet package. Java implementation of the longitudinally invariant kT and anti-kT clustering algorithms uses the E-scheme to combine particles (p1+p2) and Eta-Phi space (not Rapidity-Phi). Also, Cambridge/Aachen jets are supported.   [[https://github.com/chekanov/hepjet/|HepJet on github]] shows  benchmarks of the Java implementation of the anti-KT with the original FastJet package.
  
-See also other classes in [[http://jwork.org/dmelt/api/doc.php/jhplot/package-summary | jhplot package summary]]. +
 You can build the standard kt-jets using [[http://jwork.org/dmelt/api/doc.php/hephysics/hepsim/PromcUtil | PromcUtil]], followed by the  You can build the standard kt-jets using [[http://jwork.org/dmelt/api/doc.php/hephysics/hepsim/PromcUtil | PromcUtil]], followed by the 
-[[http://jwork.org/dmelt/api/doc.php/hephysics/jet/KTjet | KTjet]] class. Also, look at the ProMC Java API that is used to store data:+[[http://jwork.org/dmelt/api/doc.php/hephysics/jet/KTjet | KTjet]] class.  
 + 
 +===== Working with ProMC files ===== 
 + 
 +Also, look at the ProMC Java API that is used to store data:
  
   * [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/promc/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for MC and [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/pronlo/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for NLO   * [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/promc/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for MC and [[http://atlaswww.hep.anl.gov/asc/promc/doc/api/pronlo/io/ProMCHeaderFile.ProMCHeader.html | ProMCHeader]] for NLO
Line 284: Line 303:
 It is simplified compared to full MC simulations, and also includes additional arrays "idata" and "fdata" to store It is simplified compared to full MC simulations, and also includes additional arrays "idata" and "fdata" to store
 information on event weights and uncertainties. information on event weights and uncertainties.
 +
 +The project uses the community edition of [[http://jwork.org/dmelt/|DMelt]] project and  the FreeHEP Java Libraries.
  
  --- //[[[email protected]|Sergei Chekanov]] 2015/03/06 08:49//  --- //[[[email protected]|Sergei Chekanov]] 2015/03/06 08:49//
community/refhepsim_analysis.1431291212.txt.gz · Last modified: 2015/05/10 20:53 by asc