in

void TCFitV0::Test(int mode)

constraints can be added to the fit :

if (mode==1) dat.FixPar(kCNRJ);

As far as I know,  2 constraints can be added

1. a constraint a kcnrj :

   dat.FixPar(kCNRJ)

      2. a constraint on klen2

          dat.FixPar(kLEN_2)

Here the name are misleading since kCNRJ could refer to a constraint based on the mass, whereas kLEN_2 on the decay length

test 1 : kCNRJ fixed

kLEN_2 fixed

parameters used in TCFIT

TkPars &operator+=(const TkPars &a);
  /// point 
  double  dca,z;
  /// angle between track direction and X-axis in xy plane
  double phi;
  /// signed invert pt [sign = sign(-qB)]
  double ptin;  
  /// tangent of the track momentum dip angle
  double tanl;
  /// signed curvature [sign = sign(-qB)]
  double curv;  
  /// Z component magnetic field in units Pt(Gev) = Hz * RCurv(cm)
  double hz;  
  /// Mass of track
  double mass;
};


class TCFitV0 : public TCFitData {
public:
enum eTCFitV0 { kDCA_0= 0,kZ_0  = 1,kPHI_0= 2,kPTIN_0= 3,kTANL_0= 4
              , kDCA_1=10,kZ_1  =11,kPHI_1=12,kPTIN_1=13,kTANL_1=14
              , kLEN_0=20,kLEN_1=21,kLEN_2=22
              , kCX_0 =30,kCY_0 =31,kCZ_0 =32
              , kCX_1 =33,kCY_1 =34,kCZ_1 =35
              , kCNRJ =36 };

It looks from the names of the parameters that they're the parameters tracks (5)

KPTIN_0 : 1/pT of the first track

1/KPTIN_0

When I compare to the pT of the same daughter from the TLorentzVector, it could match (modulo a factor 10)

KPHI_0

When I compare to the phi angle of the same daughter from the TLorentzVector :

10/25 : UPDATE : the current version of my DNP talk is here 

 version with Gang's comments : here