RiemannsMethod.cc

/**************************************************************************
 * BASF2 (Belle Analysis Framework 2)                                     *
 * Copyright(C) 2014 - Belle II Collaboration                             *
 *                                                                        *
 * Author: The Belle II Collaboration                                     *
 * Contributors: Oliver Frost                                             *
 *                                                                        *
 * This software is provided "as is" without any warranty.                *
 **************************************************************************/
#include <tracking/trackFindingCDC/fitting/RiemannsMethod.h>
 
#include <tracking/trackFindingCDC/fitting/EigenObservationMatrix.h>
#include <tracking/trackFindingCDC/fitting/CDCObservations2D.h>
 
#include <tracking/trackFindingCDC/eventdata/trajectories/CDCTrajectory2D.h>
 
#include <framework/logging/Logger.h>
 
#include <Eigen/Eigen>
#include <Eigen/Core>
 
#include <cassert>
 
using namespace Belle2;
using namespace TrackFindingCDC;
 
RiemannsMethod::RiemannsMethod()
  : m_lineConstrained(false)
  , m_originConstrained(false)
{
}
 
void RiemannsMethod::update(CDCTrajectory2D& trajectory2D,
                            CDCObservations2D& observations2D) const
{
 
  if (observations2D.getNObservationsWithDriftRadius() > 0) {
    updateWithDriftLength(trajectory2D, observations2D);
  } else {
    updateWithoutDriftLength(trajectory2D, observations2D);
  }
 
  EForwardBackward isCoaligned = observations2D.isCoaligned(trajectory2D);
  if (isCoaligned == EForwardBackward::c_Backward) trajectory2D.reverse();
}
 
 
 
void RiemannsMethod::updateWithoutDriftLength(CDCTrajectory2D& trajectory2D,
                                              CDCObservations2D& observations2D) const
{
  EigenObservationMatrix eigenObservation = getEigenObservationMatrix(&observations2D);
  assert(eigenObservation.cols() == 4);
  size_t nObservations = observations2D.size();
 
  if (isLineConstrained()) {
 
    //do a normal line fit
    Eigen::Matrix<double, Eigen::Dynamic, 2> points = eigenObservation.leftCols<2>();
 
    Eigen::Matrix<double, 1, 2> pointMean;
    //RowVector2d pointMean;
    // cppcheck-suppress constStatement
    pointMean << 0.0, 0.0;
    if (!(isOriginConstrained())) {
      // subtract the offset from the origin
      pointMean = points.colwise().mean();
 
      points = points.rowwise() - pointMean;
    }
    Eigen::Matrix<double, 2, 2> covarianceMatrix = points.transpose() * points;
 
    Eigen::SelfAdjointEigenSolver< Eigen::Matrix<double, 2, 2> > eigensolver(covarianceMatrix);
 
    if (eigensolver.info() != Eigen::Success)
      B2WARNING("SelfAdjointEigenSolver could not compute the eigen values of the observation matrix");
 
    //the eigenvalues are generated in increasing order
    //we are interested in the lowest one since we want to compute the normal vector of the plane
 
    Eigen::Matrix<double, 2, 1> normalToLine = eigensolver.eigenvectors().col(0);
 
    double offset = -pointMean * normalToLine;
 
    // set the generalized circle parameters
    // last set to zero constrains to a line
    trajectory2D.setGlobalCircle(UncertainPerigeeCircle(PerigeeCircle::fromN(offset, normalToLine(0), normalToLine(1), 0)));
 
  } else {
 
    //lift the points to the projection space
    Eigen::Matrix<double, Eigen::Dynamic, 3> projectedPoints(nObservations, 3);
 
    projectedPoints.col(0) = eigenObservation.col(0);
    projectedPoints.col(1) = eigenObservation.col(1);
    projectedPoints.col(2) = eigenObservation.leftCols<2>().rowwise().squaredNorm();
 
 
    Eigen::Matrix<double, 1, 3> pointMean;
    // cppcheck-suppress constStatement
    pointMean << 0.0, 0.0, 0.0;
    if (!(isOriginConstrained())) {
      // subtract the offset from the origin
      pointMean = projectedPoints.colwise().mean();
 
      projectedPoints = projectedPoints.rowwise() - pointMean;
    }
 
    Eigen::Matrix<double, 3, 3> covarianceMatrix = projectedPoints.transpose() * projectedPoints;
 
    Eigen::SelfAdjointEigenSolver< Eigen::Matrix<double, 3, 3> > eigensolver(covarianceMatrix);
 
    if (eigensolver.info() != Eigen::Success)
      B2WARNING("Eigen::SelfAdjointEigenSolver could not compute the eigen values of the observation matrix");
 
    //the eigenvalues are generated in increasing order
    //we are interested in the lowest one since we want to compute the normal vector of the plane
 
    Eigen::Matrix<double, 3, 1> normalToPlane = eigensolver.eigenvectors().col(0);
 
    double offset = -pointMean * normalToPlane;
 
    trajectory2D.setGlobalCircle(UncertainPerigeeCircle(PerigeeCircle::fromN(offset, normalToPlane(0), normalToPlane(1),
                                                        normalToPlane(2))));
    //fit.setParameters();
 
  }
 
  //check if the orientation is alright
  Vector2D directionAtCenter = trajectory2D.getFlightDirection2D(Vector2D(0.0, 0.0));
 
 
  size_t voteForChangeSign = 0;
  for (size_t iPoint = 0; iPoint < nObservations; ++iPoint) {
    double pointInSameDirection = eigenObservation(iPoint, 0) * directionAtCenter.x() +
                                  eigenObservation(iPoint, 1) * directionAtCenter.y();
    if (pointInSameDirection < 0) ++voteForChangeSign;
  }
 
  if (voteForChangeSign > nObservations / 2.0) trajectory2D.reverse();
 
}
 
 
 
void RiemannsMethod::updateWithDriftLength(CDCTrajectory2D& trajectory2D, CDCObservations2D& observations2D) const
{
  EigenObservationMatrix eigenObservation = getEigenObservationMatrix(&observations2D);
  assert(eigenObservation.cols() == 4);
  size_t nObservations = observations2D.size();
 
  //cout << "updateWithRightLeft : " << endl;
  //observations always have the structure
  /*
  observables[0][0] == x of first point
  observables[0][1] == y of first point
  observables[0][2] == desired distance of first point
  */
 
  Eigen::Matrix< double, Eigen::Dynamic, 1 > distances = eigenObservation.col(2);
 
  //cout << "distances : " << endl << distances << endl;
 
  if ((isLineConstrained()) && (isOriginConstrained())) {
 
 
    Eigen::Matrix< double, Eigen::Dynamic, Eigen::Dynamic > projectedPoints(nObservations, 2);
 
    //all coordiates
    //projectedPoints.col(0) = Eigen::Matrix<double,Eigen::Dynamic,1>::Constant(nObservations,1.0);
    projectedPoints.col(0) = eigenObservation.col(0);
    projectedPoints.col(1) = eigenObservation.col(1);
    //projectedPoints.col(2) = eigenObservation.leftCols<2>().rowwise().squaredNorm() - distances.rowwise().squaredNorm();
 
    Eigen::Matrix<double, 2, 1> parameters =
      projectedPoints.jacobiSvd(Eigen::ComputeThinU | Eigen::ComputeThinV).solve(distances);
 
    trajectory2D.setGlobalCircle(UncertainPerigeeCircle(PerigeeCircle::fromN(0.0, parameters(0), parameters(1), 0.0)));
  }
 
  else if ((! isLineConstrained()) && (isOriginConstrained())) {
 
    Eigen::Matrix< double, Eigen::Dynamic, Eigen::Dynamic > projectedPoints(nObservations, 3);
 
    //all coordiates
    //projectedPoints.col(0) = Eigen::Matrix<double,Eigen::Dynamic,1>::Constant(1.0);
    projectedPoints.col(0) = eigenObservation.col(0);
    projectedPoints.col(1) = eigenObservation.col(1);
    projectedPoints.col(2) = eigenObservation.leftCols<2>().rowwise().squaredNorm() - distances.rowwise().squaredNorm();
 
    Eigen::Matrix<double, 3, 1> parameters =
      projectedPoints.jacobiSvd(Eigen::ComputeThinU | Eigen::ComputeThinV).solve(distances);
 
    trajectory2D.setGlobalCircle(UncertainPerigeeCircle(PerigeeCircle::fromN(0.0, parameters(0), parameters(1), parameters(2))));
  }
 
  else if ((isLineConstrained()) && (! isOriginConstrained())) {
 
    Eigen::Matrix< double, Eigen::Dynamic, Eigen::Dynamic > projectedPoints(nObservations, 3);
 
    //all coordiates
    projectedPoints.col(0) = Eigen::Matrix<double, Eigen::Dynamic, 1>::Constant(nObservations, 1.0);
    projectedPoints.col(1) = eigenObservation.col(0);
    projectedPoints.col(2) = eigenObservation.col(1);
    //projectedPoints.col(3) = eigenObservation.leftCols<2>().rowwise().squaredNorm()- distances.rowwise().squaredNorm();
 
    Eigen::Matrix<double, 3, 1> parameters =
      projectedPoints.jacobiSvd(Eigen::ComputeThinU | Eigen::ComputeThinV).solve(distances);
 
    trajectory2D.setGlobalCircle(UncertainPerigeeCircle(PerigeeCircle::fromN(parameters(0), parameters(1), parameters(2), 0.0)));
    //fit.setParameters(parameters(0),parameters(1),parameters(2),0.0);
 
  }
 
  else if ((! isLineConstrained()) && (! isOriginConstrained())) {
 
    Eigen::Matrix< double, Eigen::Dynamic, Eigen::Dynamic > projectedPoints(nObservations, 4);
 
    //all coordiates
    projectedPoints.col(0) = Eigen::Matrix<double, Eigen::Dynamic, 1>::Constant(nObservations, 1.0);
    projectedPoints.col(1) = eigenObservation.col(0);
    projectedPoints.col(2) = eigenObservation.col(1);
    projectedPoints.col(3) = eigenObservation.leftCols<2>().rowwise().squaredNorm() - distances.rowwise().squaredNorm();
 
    //cout << "points : " << endl << projectedPoints << endl;
 
    Eigen::Matrix<double, 4, 1> parameters =
      projectedPoints.jacobiSvd(Eigen::ComputeThinU | Eigen::ComputeThinV).solve(distances);
 
    trajectory2D.setGlobalCircle(UncertainPerigeeCircle(PerigeeCircle::fromN(parameters(0), parameters(1), parameters(2),
                                                        parameters(3))));
 
  }
}