Package org.ddogleg.optimization.math

Interface HessianMath

All Known Subinterfaces:
HessianBFGS, HessianLeastSquares<S>, HessianSchurComplement<S>
All Known Implementing Classes:
HessianBFGS_DDRM, HessianLeastSquares_DDRM, HessianLeastSquares_DSCC, HessianMath_DDRM, HessianMath_DSCC, HessianSchurComplement_Base, HessianSchurComplement_DDRM, HessianSchurComplement_DSCC
public interface HessianMath
Abstraction layer for operations related to hessian matrix. This allows different internal representations of the Hessian matrix to be used by the same code.

  • Method Details

    • init

      void init(int numParameters)
      Initialize Hessian to be in its initial state with the specified dimensions
      Parameters:
      numParameters - Number of optimization parameters. Hessian will be N by N

    • innerVectorHessian

      double innerVectorHessian(DMatrixRMaj v)
      Returns v^T*M*v
      Parameters:
      v - vector

    • extractDiagonals

      void extractDiagonals(DMatrixRMaj diag)
      Extracts diagonal elements from the hessian and stores them in the vector diag
      Parameters:
      diag - vector

    • setDiagonals

      void setDiagonals(DMatrixRMaj diag)
      Sets the diagonal elements in the Hessian to the provided vector
      Parameters:
      diag - vector

    • divideRowsCols

      void divideRowsCols(DMatrixRMaj scaling)
      Applies row and column division using the scaling vector. B = inv(diag(s))*B*inv(diag(s))
      Parameters:
      scaling -

    • initializeSolver

      boolean initializeSolver()
      Initializes the solver
      Returns:
      true if successful

    • solve

      boolean solve(DMatrixRMaj Y, DMatrixRMaj step)
      Solves the system step = inv(B)*Y
      Parameters:
      Y - (Input) vector
      step - (output) vector