System Tests Report Summary
Minor differences in the comparison with GR v15r25.
MIP filter status and state have changed - this is expected from the updates to the OnboardFilter.
Changes
- OnboardFilter - Fix to calibration file names. Change default MC filter behavior to match data - do not leak all MC events
- OnboardFilterTDS - Methods to provide enumerated state information added (as opposed to bits).
- AcdUtil - alignment added
MIP Status Bits
Update July 23: Tracy informed me that the new behavior of the MIP filter comes from Fred noting that events were being improperly vetoed for MC simulations. This GR version updates the MC interpretation of the results reported by the filter to handle changes in the flight software between B1-0-6 and B1-0-8. Veto line 26 should be disabled in the filter configuration, but was not previously. This is consistent with the behavior shown below.
The ObfMipStatus and ObfMipState (merit variables) distributions show more events passing the MIP filter in GRv15r26. Protons are shown below, but this occurs at varying levels for all the system tests, more noticeably for the proton, muon, and 100 MeV gamma tests. The change to leaked events does not impact anything here because the job options still specify that events should be leaked.
OnboardFilter.GammaFilterTool.LeakAllEvents=true;
OnboardFilter.HIPFilterTool.LeakAllEvents=true;
OnboardFilter.MIPFilterTool.LeakAllEvents=true;
OnboardFilter.DGNFilterTool.LeakAllEvents=true;
The MIP filter status plot from the system tests for the VerticalProton1GeV test (below) shows less vetos (bit 31) being set (note that this distribution is per bit, not event). The other bits are unchanged.
The MIP filter state shows a corresponding increase in passing events (state 0 = pass, state 4 = veto). The trend is strongest for low E gamma, proton, and muon tests, but appears at some level in all.
The MIP filter bit counts from the GLEAM output suggest that veto line 26 is excluded in r26. (The last listing is bit 31, which indicates any veto.)
GRV15r25 veto bit summary
MFC_STATUS_M_NO_TKR_ADJ 36682
MFC_STATUS_M_LYR_COUNTS 8328
MFC_STATUS_M_NO_ACD_TKR 2
MFC_STATUS_M_GEM_NOTKR 58
MFC_STATUS_M_GEM_CNO 438
MFC_STATUS_M_VETOED 39015
GRv15r26 veto bit summary
MFC_STATUS_M_NO_TKR_ADJ 36681
MFC_STATUS_M_LYR_COUNTS 8324
MFC_STATUS_M_NO_ACD_TKR 2
MFC_STATUS_M_GEM_NOTKR 58
MFC_STATUS_M_GEM_CNO 439
MFC_STATUS_M_VETOED 8823
This plot shows that events with status bit 26 set (and not bits 27-30) used to be vetoed are now passed. GRv15r25 is black, GRv15r26 is red.
