Ignoring the problems with the acceptance efficiency, I can still calculate trigger efficiencies.  The L0 trigger requirement is: a 3 tower cluster > 4 GeV.  In my code, I use the L2 simulator for this.

I modified the Quarkonia class so that it saves the tower soft IDs for all the L0 Triggers in a given Event.  Then, if either of the MC daughter electrons point to a L0 triggered soft ID, the event is considered to have passed the L0 Trigger requirement.

There are 727 Events which pass the acceptance criteria.

There are 392 Events which pass the L0 acceptance criteria.

Efficiency = 53.9% +/- 3.4%

Figure 1: MC Upsilon Rapidity of upsilons that pass the acceptance criteria of eleEsum>0.1&&posEsum>0.1

Figure 2: MC Upsilon pT distribution of accepted upsilons

Figure 3: MC upsilon rapidity distribution of upsilons that passed L0 trigger requirement

Figure 4: MC upsilon pT distribution of upsilons that have passed the L0 trigger

Figure 5: L0 trigger efficiency vs rapidity.  Error bars are binomial.

Figure 6: L0 trigger efficiency vs pT

There are 684 Events with upsilon |y|< 1 which pass the acceptance criteria

There are 382 events with upsilon |y| < 1 which pass L0 trigger requirement

Efficiency = 382/684 = 55.8% +/- 3.6%

Figures are not all that different, so I'll exclude them for now.

There are 323 events with trigger ID 117602 which pass the acceptance criteria

There are 313 events with trigger ID 117602 which pass the L0 criteria

Efficiency = 313/323 = 96.9% +/- 7.7%  Hmm....can't just do root N here....

There are 317 events with trigger ID 117602 which pass the acceptance criteria

There are 307 events with trigger ID 117602 + |y| < 1 which pass the L0 criteria

Eff = 96.8% +/-7.7%

There are 161 events with trigger ID 137602 which pass the acceptance criteria

There are 160 events with trigger ID 137602 which pass the L0 criteria

Efficiency = 160/161 = 99.4% +/- 11.1% 

There are 358 events with trigger ID 137602 which pass the acceptance criteria

There are 157 events with trigger ID 137602 + |y| < 1 which pass the L0 criteria

Eff = 99.4% +/-11.2%

Figure 7: Summary of L0 trigger efficiency, with wrong efficiency errors