OpenScientist Status (v11) Relationship with AIDA - PowerPoint PPT Presentation

MikeCarlo
openscientist status v11 relationship with aida l.
Skip this Video
Loading SlideShow in 5 Seconds..
OpenScientist Status (v11) Relationship with AIDA PowerPoint Presentation
Download Presentation
OpenScientist Status (v11) Relationship with AIDA

play fullscreen
1 / 15
Download Presentation
OpenScientist Status (v11) Relationship with AIDA
211 Views
Download Presentation

OpenScientist Status (v11) Relationship with AIDA

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. OpenScientist Status (v11)Relationship with AIDA http://www.lal.in2p3.fr/OpenScientist G.Barrand, LAL-Orsay

  2. OpenScientist-11 / AIDA-3.1 • v11r0 : code available now. • OpenScientist / Lab package is now AIDA-3.1 compliant. Big effort to pass from 2.2 to 3.1. • IProfile1D, IProfile2D, IDataPointSet are here. • Fitting migrated to the 3.1 interfaces, but clearly not compleete yet. • Around management : copy for exa are not here yet. G.Barrand, LAL-Orsay

  3. OpenScientist-11 / Rio-4.0 • Rio (v4r0) : a light and clean rewrting of the ROOT IO. 10 klines of code (against 200klines with ROOT/Core) to write/read an histo. • Abstract interfaces (Rio::IObject, IClass, IDirectory, IFile, etc…) • (Able to read LHCb data with it (GaudiRioDb)). • Protect the io (tobuf, IBuffer) without doing carpet hiding (using exception). • Etc… G.Barrand, LAL-Orsay

  4. OpenScientist-11 / Persistency • Read ROOT files with TH[1,2]. But do no write the Lab histos as TH (since AIDA histos are more rich that the TH (handling of entries per bin for example)). • Able to convert hbook files (h1d, h2d, rowwise ntuples) to Lab rio/root files. • A MySQL driver had been introduced : able to read histos stored with FreeHEP. But fare from being compleete. • AIDA / XML : no progress yet. G.Barrand, LAL-Orsay

  5. OpenScientist-11 / GUI, graphic • Improvement of the GUI. • GUI constructed from an XML description. • Tab stack, toolbar, tooltips, icon on buttons, user popup menu on the Inventor viewer. • All that on the four toolkits Windows, OpenMotif, gtk+, Qt. • (Be able to save back the GUI description) • InventorViewer : elastic rectangular draggable box : • Viewing mode : zoom. • Picking mode : collect objects. • gl2ps : plots and 3D graphics. • (GUI organization very similar to jas3 !!!) G.Barrand, LAL-Orsay

  6. OpenScientist-11 / Geant4 • Geant4/examples/extended/analysis/AnaEx01 updated to AIDA-3.x. • OpenScientist / G4Lab, G4LabSimple exa improved. • A G4UIOnX exists now (similar to the other G4UIs : G4UIterminal, G4UIXm, G4UIWin32, etc…) but needs DLLs on Windows for Geant4 ! • A package cloner facility exists to easily create its own “G4Lab”. G.Barrand, LAL-Orsay

  7. OpenScientist-v11 / LHCb • OnX (Coin, HEPVis) used to construct the Panoramix event display. • Lab : available in the same interactive environment. • Gaudi / SoStat package : to plot the Gaudi AIDA::IHistograms with HEPVis/SoPlotter in a GUI constructed with OnX. • Connection done with DaVinci (the project to handle physic analysis software for LHCb). G.Barrand, LAL-Orsay

  8. OpenScientist / Conclusions • Huge progress in all parts. • But fare from completion…(assuming that an analysis system if ever compleete). • (Man power expected in autumn !) G.Barrand, LAL-Orsay

  9. AIDA Answer to people that are fanatic of the project. G.Barrand, LAL-Orsay

  10. To Rene Brun (ALICE) : it does not work • Demo with OpenScientist / Lab implementation. • Sorry Rene, you are no more the only one in this universe, to book, fill, plot and store histos. • And some do it with a fare better architecture and implementation than yours (at least without some TClass::Draw, etc…). G.Barrand, LAL-Orsay

  11. To Pere Mato (LHCb) : it is too naive • Comparison of a “main” to book, fill, plot, store with AIDA and with Gaudi : • AIDA (see tutorial AIDA example). • Gaudi (see GaudiExamples histogramming). • Pere (as plenty of people at CERN) confuses simplicity with naivety. • Is it possible to design a flexible and evolutive framework simple for physicist AND software engineerers ??? G.Barrand, LAL-Orsay

  12. Vincenzo Innocente (CMS) • LCG / PI introduces a new partial implementation. • Why partial ? Because the AIDA::IFactories, ITree are discarded (along the IPlotter ?).(How users will plot, store ?) • Clearly (like for Anaphe choices) LAL and SLAC implementation will not be taken into account. • Is PI::AIDA going to be one more selfish CERNery ? • But Vincenzo (representing the LCG executive ?), at least, had decided to be at the workshop…. G.Barrand, LAL-Orsay

  13. ATLAS ???? G.Barrand, LAL-Orsay

  14. What I expect from this workshop • “all day” work of improving the interfaces. • Some dirty linen washing with CERN. • Avoid CERN demolishing what had been achieved so fare. G.Barrand, LAL-Orsay

  15. My priorities / My not priorities • Priorities : • Have a more compleete implementation. • Integration within LHCb (see Thursday). • Interoperate with other implementations. • Not priorities : • Bear a new partial implementation. G.Barrand, LAL-Orsay