First step was to run the statusCheckfill2.py code, which lists all towers that have their status change twice or more in the run.  The towers were listed for all fills I chose the following nine fills spaced throughout the run: 17250, 17281, 17329, 17376, 17415, 17461, 17512, 17534, 17593 (see emc_output.xls)

• Towers of interest are highlighted in read
• If the status of the tower was wrong, I made a correction in the "My Status" column.  For example if the status given was 1, but the tower looked 18, 18 is reported in the "My Status" column
• Propose if the status is changed in 6 of 9 fill, the tower be hard-coded bad
• Tower 282, 916 were bad in a few fills, but I don't believe are worth hard-coding as bad
• Some are are usually marked as bad but sometimes were given a good status.  Want to hard-code them just to be safe.
• Hard-code list: 410, 533, 671, 1567, 1612, 1705, 2415, 2969, 3005, 3739
• Some towers that were masked were sometimes given a status other than zero, would like to hard-code status 0 again being conservative.
• Mask list: 790, 993, 1061, 1232, 1306, 2107, 4464

Then 8 fills were hand-checked in order to determine if towers with bad statuses are missed.  For a tower to be considered "bad", I decided it had to me marked in 6 of the 8 fills.  The following fills were chosed: 17253, 17293, 17347, 17403, 17433, 17488, 17529, 17570. (see missed_towers.xls)

• Tower 360 marked 8 times

• Tower 3668 marked 8 times

• Tower 4223 marked 5 times
• Tower 4684 marked 4 times