During my event QA, I removed 138 runs from the runlist due to apparent hot towers. In an effort to save these runs I wanted to create the barrel and endcp spectra. I binned a histogram with 4800 bins (barrel) and 720 bins (endcap) and filled it each time it registered a hit with Et >2.0 GeV. I then looped through each bin and divided the number of counts in that bin by the number of counts in the histogram to get the proportion of the spectrum in that bin. If this proportion was greater than 0.2, I output that tower ID.

41 of the runs had the same tower that was hot in the barrel -- TowID: 743 (see HotTowerExample.pdf below). There weren't any issues found in the endcap for any of the runs. The runs with true hot towers are (also attached in bemcHotTowers.txt):