Tracker Recon Report
Analsysis Group Meeting
Aug 2, 2004
- Finishing the work to integrate the standalone single tower alignment into
the Gaudi environment
- Some "annoyances" encountered in converting Hiro's original code...
- Most problems now dealt with resulting in a few simplifications
- e.g. use TkrRecon clusters and track fitting utilities
- Attempting to connect directly with Leon's TkrAlignmentSvc
- This has taken me longer than I thought it would... but about ready for
- Re-structuring TkrRecon
- Current focus on implementing new TDS classes in the Event package
- Plan is to implement "new" (restructured?) TkrCluster and TkrKalFitTrack
- TkrCluster highlights:
- TkrClusterCol to become solely a container
- Use relational tables (or something similar) to provide access to clusters
in "standard" groups
- Tower, x-y pairs, plane, etc.
- Highlights of considerations for a new TkrCluster class can be found
- Johann/Leon now developing the new version
- TkrKalFitTrack highlights:
- Content mostly unchanged
- Key changes include keeping a smart pointer to cluster instead of index,
keeping volume identifier instead of tower, layer, view, ...
- Primarily changing access methods for hit information
- TDS object now contains data, methods to operate on the data moved to
- In particular, chi-square and energy loss calculations moved out of
- Restructure inheritances to fix problem (?) current structure creates for
- Also, dropping our previous goal of a single track fit base class for any
future imagined global reconstruction algorithm
- When we got one we'll deal with it
- Big change: dropping TkrPatCand TDS class, output of pattern
recognition will be a TkrKalFitTrack to be sent to the fitter for final fit
- Highlights of considerations for a new TkrKalFitTrack class can be found
- New version under development
- TkrVertex is also under consideration, longer time scale here
- Bill is looking into this now
- First step (post redesigned code): Make the existing code work with the
new TDS classes
- Since TkrRecon is undergoing modification, have been asked by the core
group to consider also modifying the package structure.
- See Traudl's guidelines here
- I&T (Leon)
- Ntuple documentation: I've added the seven new AnaTup variables that have
appeared since DC1. We still need a strategy for keeping track of (the not
necessarily synchronized) versions of AnaTup and merit.
- Bad Strips calibration: some refinements. Configuration (EM, EM2, 2-tower
LAT, etc.) is now set from options file.
- Working on revamped TkrCluster class, including new information for ToT.
- Alignment: working on code to apply alignment constants at during recon.
- Geometry: a new engineer, Steve Lundgren, will be helping with the
- Tracker JIRA items
- See Confluence page
- Alignment (see above)
- Restructuring TkrRecon (see above)
- Event Filter (in progress, expect report from Johann at tracker meeting
- Improving TkrRecon Website
- Full Monte Carlo energy accounting in the tracker
- Continuing improvements to Tracker display for FRED
T.Usher Last Modified: