Here I compare a test sample of data jet trees make with the old endcap gains vs a sample made with the new endcap gains ...

In my previous data / simulation comparisons I saw sizable disagreement between the eta spectra. As described here, I believe this discrepency is due to the endcap gains used in making the simulations. The best way to test if this is the case would be to do the data / simulation comparisons using the new endcap gains in both the simulation and the data. Pibero has agreed to rerun the RFF simulation using the new endcap gains to create the simu MuDsts and simu jet trees.

While these new simulations run, I have rerun the 100 RFF data jet trees in the test sample using the new endcap gains. Below, I compare the new data jet trees with the original data jet trees used in the previous data / simu comparisons.

The new endcap gains have been lowered by 7.5% compared to the original gains.

Figure 1: This figure shows the jet Pt spectra for the L2JetHigh, JP1, and Untriggered trigger groupings. Each pannel is normalized seperately. The blue curve is for jets created using the old gains and the red curve is for jets created using the new gains.

Figure 2: This figure has the same format as figure 1 except it shows the high jet eta spectra.

Figure 3: This figure has the same format as figures 1 and 2 but shows the low jet eta spectra.

As expected, there is an increase in the number of jets in the endcap region when using the new gains as compared with the old gains. What I don't understand yet, is the behavior at eta <~ -0.5.

I have also included the comparisons for the other quantities in this pdf.

To try to understand the features seen at the edges of the eta spectra in fig 2 and 3 I have created some new plots. It was thought that the features at the edges could be due to a change in which jet is the high pt jet and which is the low so I have created a histogram showing an inclusive jet comparison. I also thought the bumps could be a result of the jet phase space cuts I impose, so I have created a comparison where I don't impose my usual eta cuts.

The plot below contains all jets returned by the jet finder which have a z-vertex between -90 and 90 cm.

It turns out that the bumps seen in the eta spectra were actually due to a bug in how detector eta was stored for the old jet trees.

Figure 4: This figure shows jet eta spectra (left column) and jet detector eta spectra from the jet trees created with the old endcap gains (blue) and the new endcap gains (red). The jets are subject to my eta/detEta phase space cuts: -0.8<=eta<=1.8 and -0.7<=detEta<=1.7 . This figure used the detector eta which contained the bug.

Figure 5: This figure is the same as figure 4 above except that now I correct the bug in the value of detector eta.