// David Eberly, Geometric Tools, Redmond WA 98052 // Copyright (c) 1998-2020 // Distributed under the Boost Software License, Version 1.0. // https://www.boost.org/LICENSE_1_0.txt // https://www.geometrictools.com/License/Boost/LICENSE_1_0.txt // Version: 4.0.2019.08.13 #pragma once #include <Mathematics/ApprQuery.h> #include <Mathematics/Line.h> #include <Mathematics/SymmetricEigensolver2x2.h> #include <Mathematics/Vector2.h> // Least-squares fit of a line to (x,y) data by using distance measurements // orthogonal to the proposed line. The return value is 'true' if and only // if the fit is unique (always successful, 'true' when a minimum eigenvalue // is unique). The mParameters value is a line with (P,D) = // (origin,direction). The error for S = (x0,y0) is (S-P)^T*(I - D*D^T)*(S-P). namespace WwiseGTE { template <typename Real> class ApprOrthogonalLine2 : public ApprQuery<Real, Vector2<Real>> { public: // Initialize the model parameters to zero. ApprOrthogonalLine2() : mParameters(Vector2<Real>::Zero(), Vector2<Real>::Zero()) { } // Basic fitting algorithm. See ApprQuery.h for the various Fit(...) // functions that you can call. virtual bool FitIndexed( size_t numPoints, Vector2<Real> const* points, size_t numIndices, int const* indices) override { if (this->ValidIndices(numPoints, points, numIndices, indices)) { // Compute the mean of the points. Vector2<Real> mean = Vector2<Real>::Zero(); int const* currentIndex = indices; for (size_t i = 0; i < numIndices; ++i) { mean += points[*currentIndex++]; } mean /= (Real)numIndices; if (std::isfinite(mean[0]) && std::isfinite(mean[1])) { // Compute the covariance matrix of the points. Real covar00 = (Real)0, covar01 = (Real)0, covar11 = (Real)0; currentIndex = indices; for (size_t i = 0; i < numIndices; ++i) { Vector2<Real> diff = points[*currentIndex++] - mean; covar00 += diff[0] * diff[0]; covar01 += diff[0] * diff[1]; covar11 += diff[1] * diff[1]; } // Solve the eigensystem. SymmetricEigensolver2x2<Real> es; std::array<Real, 2> eval; std::array<std::array<Real, 2>, 2> evec; es(covar00, covar01, covar11, +1, eval, evec); // The line direction is the eigenvector in the direction // of largest variance of the points. mParameters.origin = mean; mParameters.direction = evec[1]; // The fitted line is unique when the maximum eigenvalue // has multiplicity 1. return eval[0] < eval[1]; } } mParameters = Line2<Real>(Vector2<Real>::Zero(), Vector2<Real>::Zero()); return false; } // Get the parameters for the best fit. Line2<Real> const& GetParameters() const { return mParameters; } virtual size_t GetMinimumRequired() const override { return 2; } virtual Real Error(Vector2<Real> const& point) const override { Vector2<Real> diff = point - mParameters.origin; Real sqrlen = Dot(diff, diff); Real dot = Dot(diff, mParameters.direction); Real error = std::fabs(sqrlen - dot * dot); return error; } virtual void CopyParameters(ApprQuery<Real, Vector2<Real>> const* input) override { auto source = dynamic_cast<ApprOrthogonalLine2<Real> const*>(input); if (source) { *this = *source; } } private: Line2<Real> mParameters; }; }