In order to properly ensure our embedded simulation request works properly, we examine the distributions of various quantities. This is to confirm the odd eta behavior seen in my previous blog doesn't continue to linger in our analysis. We believe we have found the underlying cause to the problem and is explained later this document. 

First of all, we wanted to examine the quantities after the MuDst files are created, which should verify our confidence in the embedded bfc macro. We show in the following figure the eta distributions of the EMC tower and the TPC track eta distributions. While all quantities examined (pT, phi, etc) behave as expected, we only display the eta distribution to show it is now consistent with our expectations.  Here we randomly selected relatively large zerobias data files (~10-20 events) and each file was embedded in to an individual fzd file (100 evts) created from Pythia. The previous eta distribution issue was due to the fact I was embedding zerobias events into the same fzd file, which meant that the same ~10-20 events were being repeatedly used.
 

                                        FIG 1: Plot of the EMC tower and TPC track eta distributions

 

Now for sanity purposes we re-examine the several different jet distributions. See the following plots:

                 FIG 2: Jet pT distribution

                FIG 3: Jet Eta Distribution

              FIG 4: Jet phi Distribution

             FIG 5: Tracks within jets distribution

                FIG 6: Jet track eta distribution

                 FIG 7: Jet Track phi distribution

             FIG 8: Jet Tower engery distribution

              FIG 9: Jet Tower Eta distribution

                FIG 9: Jet Tower Eta distribution

 

 

The Embedding process seems to be working properly and all distributions appear to behave as expected. Since we have examined the performance of the dijet pythia filter and trigger filter on straight simulation (see link), we now want to examine the performance of the filters using the embedded scenario. We examined the pT bin with the largest filter bias (pT bin 25-35 GeV) from the previous study. The results can be seen in the following table:

 

Table 1: Performance of the Dijet Filter Test

pT Bin	Events Thrown	Events Passing Pythia Filter	Events Passing Pythia Filter & Trigger	Events Failing Pythia Filter but had dijets after BFC	Events Failing Pythia Filter but had dijets after BFC & Passed the Trigger	Events Passing Trigger Filter	Events needed for 1pb-1
25-35	11216	6853	5304	40	25	7826