sPHENIX tracking discussion

[David Stewart <0ds.johnny AT gmail.com>]

In the spirit of release early and often, pull request 1760 now passes Jenkins and I think is ready to be merged.

Major notes:

(1) the DST format for matched tracks are trackbase/EmbRecoMatch (and EmbRecoMatchContainer)

(2) only hits in the TPC are matched. I have tried some space for including/evaluating silicon hits, but at least a moderate amount of change will be needed to include them.

(3) current the user can simply set a number of allowed reco tracks to be matched per truth track, and vs. versa (truth tracks per reco track). The default number for both of these is 1 -- such that there is a 1:1 relationship between all matched truth and reco tracks. As per our meeting discussion, that number can be loosened. Currently the logic to save a given match is the same for all matches -- minimum number of matched clusters, and minimum percentage of truth track clusters matched. I assume that we will want something more fine tunable in the future for split tracks, etc...

In order to actually use the code, it must either be run at a point where the node tree has all of the Truth and Reco clusters and tracks already populated in the tree. To test things out, my file /direct/sphenix+u/dstewart/macros/detectors/sPHENIX/g4short.root has saved one such node tree, and my macro g4match.C, located in the same directory, runs the TruthRecoTrackMatching module, 

Best,

Dave

--

David Stewart

Postdoctoral Fellow | Department of Physics, Wayne State University