The LCFI Vertex Package releases: ----------- | v00-06-01 | ----------- * fixed cmake scripts for out-of-source installations -------- | v00-06 | -------- * made Ptcalc() a public static function so it can be used in analysis code * fixed logical comparison bug in twotrackpid.cpp * relaxed maximum number of iterations in vertexfuncmaxfinderclassicstepper.cpp * added cmake configuration files in order to compile/link against LCFIVertex * moved initialization of _dataSet variable in FlavourTag.cc to init() in order to fix problem of ProcessRunHeader() being skipped when SkipNEvents parameter is set in a steering file -------- | v00-05 | -------- * simplified cmake scripts (now uses the new ILCUTIL package ) * bug fix for building documentation in SL5.5 ------------ | v00-04-pre | ------------ 2010-10-04 15:43 tomohiko * for gcc-4.4 compatibility: the boost library was updated to v1.44; other minor patches were added. ----------- | v00-03-01 | ----------- * same as v00-03 - made compatible for MacOS -------- | v00-03 | -------- 2008-12-16 10:56 walsh * src/RPCutProcessor.cc: Remove ReconstuctedParticle V0s from selected Jets. 2008-10-17 00:01 grimes * src/RPCutProcessor.cc: Modified the Monte Carlo particle ID cut so that it can use a ReconstructedParticle to MCParticle relation collection as well as the current Track to MCParticle. Also added a check on the type of objects the relation collection refers to and from. 2008-10-14 14:21 engels * CMakeLists.txt, LCFIVertexConfig.cmake.in: added 32 bit compatibility option 2008-08-12 14:44 walsh * src/NeuralNetTrainer.cc: bug fix: using D0Significance2 instead of Z0Significance in parts of the code. 2008-07-25 12:21 harderk * src/TrueAngularJetFlavourProcessor.cc: prevent true jet flavour confusion due to anomalous PDG codes with more than five digits 2008-07-22 11:38 walsh * vertex_lcfi/src/lciointerface.cpp: line 165: cerr -> cout 2008-07-21 23:11 walsh * vertex_lcfi/src/lciointerface.cpp: Bug fix: only RPs with a track are used to make a Jet in jetFromLCIORP. 2008-07-17 16:56 jeffery * doc/zvtop.pdf, vertex_lcfi/zvtop/include/doc.h: Put zvtop docs in pdf 2008-07-17 02:18 devetak * src/TrueAngularJetFlavourProcessor.cc: Bug... repetition is bad but errors are worse... 2008-07-16 11:36 harderk * src/ConversionTagger.cc: remove LCCollection::setSubset() call that seems to cause problems with re-reading the V0Veto collection in separate Marlin jobs 2008-07-09 16:24 harderk * diagnostics/: include/V0Performance.h, src/V0Performance.cc: improve treatment of SimTrackerHit count 2008-07-08 18:09 harderk * src/ConversionTagger.cc: adjust default values and histogram ranges according to cut tuning results so far 2008-07-08 16:51 harderk * include/ConversionTagger.h, src/ConversionTagger.cc: replace entire vertex reconstruction by new MarlinUtil HelixClass code 2008-07-08 13:29 harderk * src/ConversionTagger.cc: improved (3d) calculation of helix distances -------------- | v00-02-06-dev | -------------- 2008-07-07 15:57 harderk * src/ConversionTagger.cc: clean up hardcoded numbers such as particle masses 2008-07-07 15:56 harderk * diagnostics/src/V0Performance.cc: fix bug that would cause segmentation fault when dealing with corrupted input collections 2008-07-07 15:04 lastovic * vertex_lcfi/src/TState.cpp: Magnetic field taken from Gear. 2008-07-05 10:52 gaede -------------- | v00-02-05-dev | -------------- 2008-07-04 18:31 walsh * src/DSTCollectionProcessor.cc: - Put try-catch blocks in the correct order; - Define pNames.size() and mcNames.size() as integers in the initialisation loops of fv and mcv; - FTSelectedJets -> Jets as the default for _JetCollectionName 2008-07-03 17:07 martin * src/RPCutProcessor.cc: Comment out warning line: " MCParticles related to this track! No cuts performed on MC PDG code. which appears too much to act as any warning! 2008-07-03 15:32 martin * src/LCFIAIDAPlotProcessor.cc: Bug fix in purity printout for BC tag. Thanks to Sonja for spotting this! 2008-06-30 13:37 harderk * src/ConversionTagger.cc: exploit momentum asymmetry between Lambda decay proton and pion 2008-06-27 18:00 harderk * diagnostics/src/V0Performance.cc: another Lambda plot 2008-06-27 11:54 harderk * diagnostics/src/V0Performance.cc: add histogram with momentum distributions of Lambda decay products -------------- | v00-02-04-dev | -------------- 2008-06-25 18:20 lynch * include/DSTAIDAPlotProcessor.h, include/DSTCollectionProcessor.h, src/DSTAIDAPlotProcessor.cc, src/DSTCollectionProcessor.cc, src/DSTPlotProcessor.cc, steering/ftfordst.xml: changed DST Processors so that PID info is added to the existing collection added DSTAIDAPlotProcessor 2008-06-24 17:06 harderk * include/ConversionTagger.h, src/ConversionTagger.cc: add capability to enable or disable tagging of conv,K0,Lambda0 separately 2008-06-19 16:04 harderk * include/ConversionTagger.h, src/ConversionTagger.cc: reject candidates coming from vicinity of the IP 2008-06-17 16:05 harderk * src/ConversionTagger.cc: bug fix: existing conv/V0 candidates (e.g. found by PandoraPFA prior to running ConversionTagger) were not added to the LCCollection containing all conv/V0 candidates. 2008-06-13 13:59 harderk * diagnostics/: include/V0Performance.h, src/V0Performance.cc: count and print total number of conv/V0 candidates separately for conv, K0, Lambda. this improves diagnostics to tune ConversionTagger cuts separately for the different candidate classes 2008-06-13 11:15 harderk * diagnostics/: include/V0Performance.h, src/V0Performance.cc: add capability to look for conversions only in selected LCCollections. remove some debug output 2008-06-13 11:13 harderk * src/ConversionTagger.cc: reduce verbosity a bit -------------- | v00-02-03-dev | -------------- 2nd release for full-chain tests for ILD mass reconstruction. Changes to the way the flavour tag collections are written out for DST - TrueAngularJetProcessor + NeuralNetTrainer + PlotProcessor + AIDAPlotProcessor: changed so that the MC Truth info is stored in one LCFloatVec, in the order TrueJetFlavour, TruePDGCode, TrueHadronCharge, TruePartonCharge (Erik Devetak, Victoria Martin) - 2 new processors created in order to add FT and MC Truth info to the jets using ParticleID objects This will reduce the number of collections for the mass reconstruction DSTs: DSTCollectionProcessor : adds 2 ParticleID algorithms to each jet: "LCFIFlavourTag": contains the flavour tagging info, "MCTruth": contains the Monte Carlo truth info DSTPlotProcessor : a rewrite of PlotProcessor, which uses the DST collections from the ParticleID objects instead of the full collections. also contains a function that will compare the parameters from both collection types and print out a message if they are different. (Clare Lynch) -------------- | v00-02-03-pre| -------------- release for full-chain tests for ILD mass reconstruction, fixing a number of small issues found since last release and adding further functionality: - steering files updated to include steering file running FullLDCTracking, PandoraPFA (instead of cheattracks+jetfind); based on current version of steering file for mass reconstruction from ILD mailing list (Clare Lynch) - LCFIAIDAPlotProcessor: now compatible with RAIDA; (Victoria Martin) corrected: "ptmasscorrection" --> "ptcorrectedmass" (Roberval Walsh) - RPCutProcessor: small bug fix: this processor should not require GEAR unless hadronic interaction removal is selected (Kristian Harder) - ZVRES: small bug fix in LCFIVertex/vertex_lcfi/zvtop/src/vertexfinderclassic.cpp to ensure consistency with ZVTOP paper (Ben Jeffery) - ZVKIN: return IP if no vertices found; sort found vertices wrt decay length from IP, in ascending order (Ben Jeffery) - flavour tag inputs: code added to obtain new parameters for calculation of joint probability from fit; currently only available if the package is run with JAIDA (Erik Devetak) - neural network code: minor fixes: ensured code compiles with gcc 4.2.1; (Kristian Harder, Dave Bailey) missing "const" added; at request of CALICE added class to determine the importance of the input variables for a given data set and trained network based on the approach used in the TMVA package, see http://tmva.sourceforge.net/docu/TMVAUsersGuide.pdf p66 (Dave Bailey) - NOTE: with this version, a new dependence on MarlinUtil is introduced; this became necessary for the ConversionTagger processor, currently under development - don't use this processor yet! (Kristian Harder) (Sonja Hillert, S.Hillert1@physics.ox.ac.uk) ------------------------------------------------------------- ---------- | v00-02-02| ---------- - LCFIVertex/vertex_lcfi/algo/src/pereventipfitter.cpp changed to use Kalman filter for vertex fitting (reduces run time for PerEventIPFitter.cc by about a factor 100) (Kalman filter developed by S. Gorbunov, I. Kisel, Vertex Package interface: Tomas Lastovicka) - RPCutProcessor: for suppression of hadronic interactions using MC information this processor is now updated to support arbitrary detector geometries (Kristian Harder) - LCFIAIDAPlotProcessor: new processor providing diagnostics for flavour tag (Victoria Martin) Please note -BUILD_WITH="AIDAJNI" must be specified to cmake to build LCFIAIDAPlotProcessor - LCFIVertex/macro/FlavourTagInputsOverlay.py: python script to create plots from output of LCFIAIDAPlotProcessor (Victoria Martin) - LCFIVertex/macro/MakePurityVsEfficiencyRootPlot.C: root macro to plot purity vs efficiency graphs that can be written out by PlotProcessor, if compiling with root option; permits results from different runs (e.g. with different settings) to be compared (Kristian Harder) - vertex charge calculation: correction of algorithm to agree with the procedure developed with fast MC (Erik Devetak) - separate VertexChargeProcessor, storing results in separate LCFloatVec collections; this used to be done in FlavourTagInputsProcessor (Erik Devetak) For further details on these changes, please see the documentation. (Sonja Hillert, S.Hillert1@physics.ox.ac.uk) ------------------------------------------------------------- ---------- | v00-01-01| ---------- - same code as v00-01 the first official release except that cmake build is supported Installation: (using cmake - new build tool for Marlin v00-09-08) ------------------------------------------------------------- 1st: edit the file BuildSetup.cmake according to your system setup. 2nd: type the following commands: $ tar zxvf boost.tgz # -- if not yet done -- $ mkdir build $ cd build $ cmake -C ../BuildSetup.cmake .. $ make install The documentation gets automatically built when you type "make install". If you don't want this, set the variable INSTALL_DOC to OFF in your BuildSetup.cmake Afterwards you can still build it with: $ make doc -> all documentation (HTML+PDF) $ make doc_html -> for the HTML documentation $ make doc_latex -> for the PDF documentation (refman.pdf) For more information on CMake for the ILCSoftware check: http://ilcsoft.desy.de/portal/general_documentation/index_eng.html CMake's official website: http://www.cmake.org (Jan Engels, jan.engels@desy.de) ------------------------------------------------------------- Installation: If you already have MarlinReco installed and working then installing this package should only require placing the directory containing this file in $MARLIN/packages eg. $MARLIN/packages/LCFIVertex Documentation To build the documentation, which includes examples of how to run the processors in the package you will need doxygen (http://www.doxygen.org/) and pdflatex for the pdf documentation. They are usually included with most linux distributions. Run the following commands from the LCFIVertex directory: For the HTML documentation: cd doc doxygen firefox html/index.html (or other browser) For the PDF documentation (a subset of the HTML which does not include all classes in the package): cd doc doxygen Doxyfile.latex cd latex make (a file refman.pdf is produced) Ben Jeffery b.jeffery1@physics.ox.ac.uk