sPHENIX tracking discussion

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

Hi all,

In the truth to reco track matching code, the algorithm is (in part) as follows:

1. A truth track and a reco track of interest are found to be compared, based on proximity in phi/eta/pT.

2. An array, indexed by layer id, is filled with the cluster keys for both the truth and reco tracks.

3. In each layer, if there is a cluster key in both the reco and truth track, then the clusters are compared based on the {phi,z} and width of {phi,z} value for both the truth and reco track. 

To step 3, I propose to preliminarily add the requirement that the hitsetkey also be identical between truth and reco tracks. Otherwise, the local geometry won't be the same between the two and the comparison of the phi values will be incorrect. True?

thanks--

David Stewart

Postdoctoral Fellow | Department of Physics, Wayne State University