Skip to content

MASSWITHHYPOTHESES mixing pre- and post-fit momenta

I use the MASSWITHHYPOTHESES functor in an HLT2 line to do a veto. I tupled the results of a line and wanted to see the distribution of events around the veto region and encountered an interesting discrepancy.

Checking the code, I see that the functor combines the pre-vertex-fit momenta of the daughters with the post-vertex-fit momentum of the mother. This means that the MASSWITHHYPOTHESES distribution, when given the correct hypotheses, differs significantly from the mass distribution given by MASS. Much more than just using the pre-vertex-fit momenta for both mother and daughters.

Mind that I use the MASSWITHHYPOTHESES in the CompositeCut. Not sure what happens when it is used in the CombinationCut.

D_MASS D_MASS_DMISID D_MASS_DMISID2

  • D_MASS: Mass distribution from the MASS functor
  • D_MASS_DMISID: Mass distribution calculated from the PX, PY, PZ of the daughters (pre-vertex-fit, I assume) using the correct hypotheses:
  • D_MASS_DMISID2: Mass distribution calculated using the MASSWITHHYPOTHESES algorithm using the correct hypotheses (using the pre-vertex-fit 3-momenta to calculate the energy of the mother. Combined with the post-vertex-fit 3-momentum of the mother to calculate its mass).

MASSWITHHYPOTHESES cannot be used in FunTuple, so I had to recreate the algorithm. I checked that it matches the functor by looking at a vetoed region (grey band) using the alternate hypotheses.

D_MASS_DMISID_veto D_MASS_DMISID2_veto

Is this a desired behaviour? Is it consistent with Run 2?