The Carnegie Mellon Partial Wave Analysis Project
Status Report on June 30, 2000

We are now very close to performing the simplest fit possible. We are starting with the unpolarized data for the reaction below. We have about 150,000 data events --- events which have been chosen according to Adam's weighting routine.

p n a2+(1320) n ( 0 +) n + - +.

From these 150,000 events, we have selected a 20MeV/c2 wide bin around the center at the a2, this bin contains about 16,000 events. The mass and angular distributions for these events are displayed under our June 29 plots. We have next taken 43,000 phase space Monte Carlo events, and passed them through the initial parts of the pwa2000 code. ("gamp" to do "gamplitudes", and then "camp" to combine the "gamplitudes" into "camplituded" weights.) Using these weighted events we have generated the same distributions as before. They do look satisfyingly similar!

Next, we have compared the event weight as computed by Adam's code with that as generated by the pwa2000. We already know that there is a rather large scale factor between these events, and Adam's weight does contain a Breit-Wigner weight for the a2. However, since we are at the peak of the a2, there should only be a small variation. We include a plot of these two weights against each other here. The vertical axis is Adam's weight, while the horizontal axis is that as computed by "camp" from "gamp".

Weight Picture
At the moment, we do not understand why these look so different, but it is likely that a fit will not converge until there is some stronger correlations between these. The most likely explanation is that the pwa2000 reflectivity amplitude is not completely correct(or appropriate for unpolarized events). It may still be missing some interference effects, but we are not yet sure. The other oddity is the sharp edge in the plot. If there were no correlation, you would not expect this, and if there were good correlations, you would expect all the events along the line. One possible explanation is that the line corresponds to events where both weights are computed more or less correctly, and the region below the line corresponds to something not being computed correctly.

Last Updated on the 30'th of June, 2000 by Curtis A. Meyer.