Slides 23-October-2002

For some reason the difference in time between a hit in an adjacent stack and the same sort of hit if all are in the same stack is not quite what we expect. When I want to calculate the chi-squared I need to correct for the crossing time, and I find that I have to add additional terms depending on whether the stub was entirely within a single stack or not. I do not yet understand all of these correction terms. However, the effect on the chi-squared distributions is gratifying. The bizarre distributions involving two parabolas in last week's plots were due to these two different classes of stubs, with different chi-squares.
Bob's algorithm is more closely based on finding the crossing time (the time when a muon crosses the BMU stacks), and so it is not surprising that his distribution is tighter than mine. The difference in the mean is a little odd, though.
I traced the difference down to one section of code, and re-optimized it, with the result that my fit now looks much more like Bob's. Also note that the number of entries in the plots are much closer than they used to be.

Modified 23-Oct-2002 at 15:17

http://hep.physics.wisc.edu/~jnb/imu/23Oct2002