There were two issues to be dealt with this week.  First, we had trouble extracting the calorimeter information from the MuDst, which did not seem to have an StEmcCollection.  Initially we believed this to be the fault of the reconstruction chain we were using: however, it turned out that the problem was in the StEmcSimulatorMaker.  A temporary fix was obtained by using the StEvent instead, and Adam has checked in a fix to this problem.

The other issue we had was a big spike in the eta distribution for electrons in the endcap at eta=1:

It turns out this spike was closely associated with a large number of very-low-pt tracks.  As the electrons were generated with 5<pt<50 and QA plots from the muon files did not show this spike, it seems likely that the low-pt tracks are from secondary scattering, possibly soft photons.

Imposing a cut of pt>3 on the QA plots cleaned them up considerably:

Having overcome these problems, I could start implementing a crude clustering algorithm which finds the 2x2 block of towers with the most energy and calls that a cluster.  Eta and Phi could then be calculated from the energy-weighted average of the eta and phi values of the towers in the cluster.  This was done for 1000 events of each of the particles generated so far, namely electrons, photons, muons, pi+, pi-, pi0, and protons.  The results are in the following pdfs: the first page shows the results from the clustering, the second and third the QA plots for the file.

electrons
photons
muons
pi-
pi+
pi0
protons