Versions Compared

Old Version 20

changes.mady.by.user Maurik Holtrop

Saved on

compared with

New Version Current

changes.mady.by.user Maurik Holtrop

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
  • Monte Carlo:
    •  Use data for beam background. (Unknown User (btu29))
      • Allow data (pulser trigger, mixed up events, ...) to be mixed in with MC generated tridents+wabs to save processing time.
    •  Improve Monte Carlo - Data agreement.
      • Check/fix resolutions at the low levels (see Readout tasks).
      • Check/fix material budgets, geometry differences.
  • Vertexing:
    •  Update current algorithm for proper calculation of track parameters for detached vertexes. "Vertexing Bug". Fixed: Issue89 (Norman Graf)
    •  Look for/implement improved vertexing algorithms. Started. (Norman Graf)
    •  Vertexing in a varying B-field. OLD task.  (Norman Graf) 
  • Tracking:
    •  Implement a “best track” selection by default to pare down the track list so there aren't so many overlapping tracks (and v0s) given to the generic user. (@Miriam)
    •  New tracking finding algorithm to replace Seed Tracker / Speed up Seed Tracker. (@Miriam)+(Omar Moreno)
      • Investigate more modern tracking algorithms (cellular automata, DAF, etc.)
      • Speed up existing track finding code, particularly in the track confirm/extend steps.
        • Possibly also the search algorithm that looks for next scattering point.
    •  GBL/Alignment - improve the code to make alignment easier (work with @Alessandra Filippi) 
    • Replacement of material management classes.
    • Do not use the detector components for storing links to readout hits; instead use the event to store a transient hit map per event which will improve potential MT compatibility (Jeremy McCormick) (not feasible without major changes to lcsim)
  • Data Quality Monitoring
    • Revisit and cleanup the histograms in DQM
    • Make this a much smoother process (specifics???)
    • More automated, easier to understand, diagnostics on whether data is good quality
    • Follow up with the output of DQM → Good Runs List
  • Monte Carlo:
    • Use data for beam background.
      • Allow data (pulser trigger, mixed up events, ...) to be mixed in with MC generated tridents+wabs to save processing time.
    • Improve Monte Carlo - Data agreement.
      • Check/fix resolutions at the low levels (see Readout tasks).
      • Check/fix material budgets, geometry differences.
  • Readout
      Readout:
      • Improve speed.
        • Thorough review of what is going on in readout
      • Thorough review of what is going on in readout.
        • Run the readout step through a profiler and identify slow parts. (Maurik Holtrop)
        • Check and improve the ECAL readout code. Major Progress (Samantha McCarty)
          • Allow truth information to be propagated.
          • Look for areas where code can be cleaned up/sped up.
          • Improve accuracy, correspondence with data (see Monte Carlo tasks)
        • Check the SVT readout code.
          • Look for areas where code can be cleaned up/sped up.
          • Improve accuracy, correspondence with data (see Monte Carlo tasks)
    • DST
      • Put the "Tuple Maker" functionality in the DST-maker. (Not good to have 2 separate DSTs creation tools.)
        • May need to update information in LCIO file if it is not available.
    • MC Production:
      •  Provide script for wrapping job chain  (Jeremy McCormick, Unknown User (btu29))  – Resolved with hps-mc project.
        • EGS5
        • Madgraph4
        • Madgraph5
        • slic
        • hps-java (readout)
        • hps-java (recon)
        • DST Maker
      •  Setup running at SLAC  – Resolved with hps-mc project.
        • Port job scripts from Auger
      • Catalog existing data files in datacat
        • Provide scripts for crawling and adding MC data to the datacat
      • Implement correct beam parameters (Takashi Maruyama)
    • Event Display
        Improve usability for producing single event displaysButton for saving a quick screenshot
      • Provide better instructions (and or facilities) for selecting events (what does this mean?)
      • Get more feedback from users on what is needed here
    • Conditions System
      • Rewrite to use tabular data model as provided by freehep package (not feasible)
      • Import trigger time offsets to remove dep on run db
    • Data Catalog
      • Add authorization scheme and unlock the "/r/*" URLs
      • Add files
        • MC SLIC and readout
        • Data recon and DST
        • DQM 
    • Run Database
    • SLIC 
      •  Move to github along with LCDD and GDML
      •  Port sim plugin functionality from ldmx-sw (Jeremy McCormick)
      •  Write one or more event biasing plugins for producing specialized samples quickly (Omar Moreno)
        •  bias for WAB conversion in tracker "layers" 1-4
      •  Provide replacement CMake-based build system to ilcinstall (Jeremy McCormick)
    • hps-java
      •  Remove ConditionsDriver which is already replaced by command line arguments
      •  Remove EventMarkerDriver which is already replaced by command line arguments
      •  Test steering files and delete or sandbox non-working ones
      •  Move users module to separate github repo without branch protection on master
      • It could be pulled in as a git sub-module relatively easily if desired.
      •  Normalize the meaning and usage of the same switches across different command line programs
        • In general, if two command line programs accept the same or similar options, they should use the same switches for them.
      •  Import the fieldmaps into a git repo, which could be a Maven project
      • Decide whether to try and handle test run data
        • If not rip out test run code and detectors
        • If test run data should still be supported then many things would need to be fixed up!
    • lcsim
      • Provide capability to run a multi-threaded event loop (not feasible at the moment)
      • Make sure current LCIO implementation is up to date with standalone LCIO project
      • Replace LCIO internal implementation with the standard Java one from the standalone project
    • Geometry
      • Use "new" alignable ECal geometry model in production detector