February 13, 2007 Software
Attendees: David L. (chair), Elke A., Simon T. Elliott W., Eugene C., Mark I. On the Phone: Mihajlo K., Richard J.
Mihajlo gave an update to follow on to report he gave last week and to address comments that came up.
Work was started on implementing the generator to throw photons uniformly over solid angle (as opposed to uniformly over theta). The idea was to make the plot of multiple clusters found for single events more representative of the problem via the areas of the histograms. However, after some discussions between Matt and Mihajlo, they were unclear as to how strong the motivation for this was. The primary proponents of this suggestion were not present at this meeting so it was decided Mijhalo should focus on other issues with the reconstruction for now.
Eugene asked how many blocks are included in the clusterizer. Richard said "about 7". Richard also said that the clusterizer from which the current code was derived had an algortihm that could pull in blocks as far as 15 blocks away from the central one. Mihajlo said his implementation has a cut off at 25cm from the central block.
Generation of splash patterns on the FDC for single photon events is still in progress.
There was some discussion about the floor term in the resolutions obtained from the current reconstruction. Mijhalo noted that he is able to manipulate the obtained floor term via the attenuation factor used in the M.C. Richard argued that the floor term really should not be present in the M.C. results since it arises from things like electronics noise which is not currently modeled. He suggested correcting the lead-glass blocks via a map of their response functions wrt to angle and energy.
Put tarballs of external packages on Wiki(David) SVN web interface for browsing change logs(Mark I.)
- Experiment with shared pointers
- 12GeV Software Group meetings
- Place BCAL info on Wiki