Difference between revisions of "Brief Kinematic Fitting Example"

Revision as of 14:40, 15 March 2012

Brief summary

Kinematic fitting in the HallD sim-recon code is provided by the DKinFit class.

DKinFit kfit;

The kinematic fitter takes as its input DKinematicData objects, each of which contains both a 4-vector and a kinematic covariance matrix, which are the inputs to the fit. The fitter expects as input two vectors of DKinematicData, containing the initial and final state particles.

vector<DKinematicData> initial;
vector<DKinematicData> final;

DKinematicData beam_photon,target,reconstructed_proton,reconstructed_pi1,...;
//fill these objects appropriately using objects from factories

initial.push_back(beam_photon);
initial.push_back(target);

final.push_back(reconstructed_proton);
final.push_back(reconstructed_pi1);
final.push_back(...);

If there is a missing particle in the final state, use the SetMissingParticle()

const double neutron_mass=.940
kfit.SetMissingParticle(neutron_mass);

If there is a mass constraint (e.g. two photons from a pi0 decay), use SetMassConstraint()

vector<DKinematicData> initial;
vector<DKinematicData> final;
vector<DKinematicData> initial_post;
vector<DKinematicData> final_post;

//Add the beam particle to the initial state
DKinematicData beam_photon;
DVector3 mom(0.0, 0.0, 9.0);
beam_photon.setMomentum(mom);
beam_photon.setPosition(pos);
beam_photon.setMass(0.0);
beam_photon.setCharge(0.0);
beam_photon.setMassFixed();
beam_photon.clearErrorMatrix();

initial.push_back(beam_photon);

//Add the target to the initial state
DKinematicData kd_targ;
kd_targ.setMass(0.93827);
kd_targ.setCharge(1.0);
kd_targ.setMassFixed();
kd_targ.setMomentum( DVector3(0.0, 0.0, 0.0) );
kd_targ.clearErrorMatrix();
initial.push_back(kd_targ);

//Add all our particles to the final state
final.push_back(proton[k]);
final.push_back(pip[l]);
final.push_back(pip[m]);
final.push_back(pim[n]);
final.push_back(pim[o]);

double cl=kfit.Prob();