I run my debug code over the minimc files.

A typical event looks llike this. I attached a debug log file at bottom.

-------------------------------------------------------------------------

StMcResonanceMaker::LoopOverMcTracks()
 nMcTracks : 12
##MC##  parent :     0( 0)  daughter : 11151( 1)  trk->pMc(): 0.645 trk->ptMc(): 0.590
##MC##  parent :     0( 0)  daughter : 11151( 2)  trk->pMc():   nan trk->ptMc():   nan
##MC##  parent :     0( 0)  daughter : 11151( 3)  trk->pMc():   nan trk->ptMc():   nan
##MC##  parent :     0( 0)  daughter :     0( 0)  trk->pMc(): 0.000 trk->ptMc(): 0.000
##MC##  parent : 11151( 3)  daughter :     3( 4)  trk->pMc(): 0.509 trk->ptMc(): 0.312
##MC##  parent : 11151( 3)  daughter :     2( 5)  trk->pMc(): 0.509 trk->ptMc(): 0.312
##MC##  parent : 11151( 2)  daughter :     3( 6)  trk->pMc(): 0.507 trk->ptMc(): 0.219
##MC##  parent : 11151( 2)  daughter :     2( 7)  trk->pMc(): 0.507 trk->ptMc(): 0.219
##MC##  parent :     2( 7)  daughter :     3( 8)  trk->pMc(): 0.000 trk->ptMc(): 0.000
##MC##  parent : 11151( 1)  daughter :     3( 9)  trk->pMc(): 0.722 trk->ptMc(): 0.635
##MC##  parent : 11151( 1)  daughter :     2(10)  trk->pMc(): 0.479 trk->ptMc(): 0.472
##MC##  parent :     3( 9)  daughter :     3(11)  trk->pMc(): 0.002 trk->ptMc(): 0.002
MC # of Parents : 3 # of Daughters : 6

Parent GeantId(ParentKey) => daughter GeantId(Key) + ...
11151( 1) =>   3( 9) +   2(10)
11151( 2) =>   3( 6) +   2( 7)
11151( 3) =>   3( 4) +   2( 5)
  2( 7) =>   3( 8)
  3( 9) =>   3(11)

nParents/mult: 0.0333333

StMcResonanceMaker::LoopOverContamPairs()
 nContamPairTracks : 2
##Contam##  parent : 11151( 1)  daughter :   3( 9)  process :   5 commonHits : 32
##Contam##  parent : 11151( 1)  daughter :   2(10)  process :   5 commonHits : 45
Positive Candidates : 1 Negative Candidates : 1
StMcResonanceMaker::FillInvMass()
End StMcResonanceMaker::Make()
---------------------------------------------------------------------------

The first part shows all MC tracks.

The listed items are, from the left, parentgeantId(parentkey), trackgeantId(key), p_Mc, and pt_Mc.

As you can see, only one phi(11151) has meaningful momentum value.

Only daughters decaied from the phi have different momentum from it's partner.

As far as I checked, only such daughter particles from phi with reasonable momentum are included in the contam pairs.

You can see this in the following plot. The Entries are 30661 but integral gives only 4463 counts, ~15%. The 75% of phi have p=nan value.

I used only 3927 events (after event selection) here.

Because those wrong momentum daughters are not included in the contam pairs, it doesn't affect the phi reconstruct efficiency.

However, we loose a lot of statistics.