Apparently the clustering algorithm is working correctly! This morning I modified the program (brought it back closer to what Adam's original code was) to run over all the entries in a given run. The current program can be found in my home directory at STAR: /star/u/aschille/PhotonJetClusterMaker8.C