Another look at the dijet response matrices ...

Old correction scheme had 4 bins based on detector eta topology and trigger content.
New test correction scheme also has 4 bins based on physical jet rapidity and trigger content.

The old trigger divisions were:
JP2-JP2, JP2-UnJP2, JP1Hi-JP1Hi, JP1Hi-UnJP1Hi
and
JP2-JP1Lo, JP1Lo-JP1Lo, JP1Lo-JP1Hi, JP1Lo-UnJP1Lo

The new trigger divisions will break things up by which L0 trigger the trigger combination included:
JP2-JP2, JP2-JP1Lo, JP2-UnJP2
and
JP2-JP1Lo, JP1Lo-JP1Lo, JP1Lo-JP1Hi, JP1Lo-UnJP1Lo, JP1Hi-JP1Hi, JP1Hi-UnJP1Hi

Note that JP2-JP1Lo has both L2JetHigh and JP1 contributions so is included in both groups.

Changes were made to make final analysis easier. The new topologies reflect divisions I plan on using in final result. The new trigger combinations seperate the L0 L2JetHigh and JP1 triggers so prescales can be applied.

The new correction bins are:
Bin1 = |jet rapidity| < 1 && L2JetHigh triggers
Bin2 = |jet rapidity| < 1 && JP1 triggers
Bin3 = |jet rapidity| >= 1 && L2JetHigh triggers
Bin4 = |jet rapidity| >= 1 && JP1 triggers

Figure 1: This figure shows the away vs same side jet pt for the different correction bins

Figure 2: This figure shows the away vs same side jet physical (top) and detector (bottom) eta.

Figure 3: This figure shows the particle vs detector level dijet mass. The top figure shows the scatter plot, the middle plot shows the bin averages overlayed, and the bottom plot shows the overlayed averages for the old correction bin scheme.

A few thoughts:

• The trigger dependent upper jet pt cuts sculpt things in an odd way and its not obvious how to apply them to the truth level (jets from pythia.root files) jets. Maybe we could have one loose cut for all jets and imploy a pt balance cut?
• The par vs det mass bin averages show much less seperation in the new scheme. Would it be possible to apply the prescales to the data yields for the 8 unique trigger combinations before the corrections are applied and use the old trigger divisions?