Package org.apache.commons.math3.optimization.general

Class AbstractLeastSquaresOptimizer

java.lang.Object
org.apache.commons.math3.optimization.direct.BaseAbstractMultivariateVectorOptimizer<DifferentiableMultivariateVectorFunction>
org.apache.commons.math3.optimization.general.AbstractLeastSquaresOptimizer
All Implemented Interfaces:
BaseMultivariateVectorOptimizer<DifferentiableMultivariateVectorFunction>, BaseOptimizer<PointVectorValuePair>, DifferentiableMultivariateVectorOptimizer
Direct Known Subclasses:
GaussNewtonOptimizer, LevenbergMarquardtOptimizer
@Deprecated public abstract class AbstractLeastSquaresOptimizer extends BaseAbstractMultivariateVectorOptimizer<DifferentiableMultivariateVectorFunction> implements DifferentiableMultivariateVectorOptimizer
Deprecated.
As of 3.1 (to be removed in 4.0).
Base class for implementing least squares optimizers. It handles the boilerplate methods associated to thresholds settings, Jacobian and error estimation.
This class constructs the Jacobian matrix of the function argument in method optimize and assumes that the rows of that matrix iterate on the model functions while the columns iterate on the parameters; thus, the numbers of rows is equal to the dimension of the Target while the number of columns is equal to the dimension of the InitialGuess.
Since:
1.2

  • Field Details

    • weightedResidualJacobian

      @Deprecated protected double[][] weightedResidualJacobian
      Deprecated.
      As of 3.1. To be removed in 4.0. Please use computeWeightedJacobian(double[]) instead.
      Jacobian matrix of the weighted residuals. This matrix is in canonical form just after the calls to updateJacobian(), but may be modified by the solver in the derived class (the Levenberg-Marquardt optimizer does this).

    • cols

      @Deprecated protected int cols
      Deprecated.
      As of 3.1.
      Number of columns of the jacobian matrix.

    • rows

      @Deprecated protected int rows
      Deprecated.
      As of 3.1.
      Number of rows of the jacobian matrix.

    • point

      @Deprecated protected double[] point
      Deprecated.
      As of 3.1.
      Current point.

    • objective

      @Deprecated protected double[] objective
      Deprecated.
      As of 3.1.
      Current objective function value.

    • weightedResiduals

      @Deprecated protected double[] weightedResiduals
      Deprecated.
      As of 3.1.
      Weighted residuals

    • cost

      @Deprecated protected double cost
      Deprecated.
      As of 3.1. Field to become "private" in 4.0. Please use setCost(double).
      Cost value (square root of the sum of the residuals).

  • Constructor Details

  • Method Details

    • getJacobianEvaluations

      public int getJacobianEvaluations()
      Deprecated.
      Returns:
      the number of evaluations of the Jacobian function.

    • updateJacobian

      @Deprecated protected void updateJacobian()
      Deprecated.
      As of 3.1. Please use computeWeightedJacobian(double[]) instead.
      Update the jacobian matrix.
      Throws:
      DimensionMismatchException - if the Jacobian dimension does not match problem dimension.

    • computeWeightedJacobian

      protected RealMatrix computeWeightedJacobian(double[] params)
      Deprecated.
      Computes the Jacobian matrix.
      Parameters:
      params - Model parameters at which to compute the Jacobian.
      Returns:
      the weighted Jacobian: W1/2 J.
      Throws:
      DimensionMismatchException - if the Jacobian dimension does not match problem dimension.
      Since:
      3.1

    • updateResidualsAndCost

      @Deprecated protected void updateResidualsAndCost()
      Deprecated.
      As of 3.1. Please use computeResiduals(double[]), BaseAbstractMultivariateVectorOptimizer.computeObjectiveValue(double[]), computeCost(double[]) and setCost(double) instead.
      Update the residuals array and cost function value.
      Throws:
      DimensionMismatchException - if the dimension does not match the problem dimension.
      TooManyEvaluationsException - if the maximal number of evaluations is exceeded.

    • computeCost

      protected double computeCost(double[] residuals)
      Deprecated.
      Computes the cost.
      Parameters:
      residuals - Residuals.
      Returns:
      the cost.
      Since:
      3.1
      See Also:

    • getRMS

      public double getRMS()
      Deprecated.
      Get the Root Mean Square value. Get the Root Mean Square value, i.e. the root of the arithmetic mean of the square of all weighted residuals. This is related to the criterion that is minimized by the optimizer as follows: if c if the criterion, and n is the number of measurements, then the RMS is sqrt (c/n).
      Returns:
      RMS value

    • getChiSquare

      public double getChiSquare()
      Deprecated.
      Get a Chi-Square-like value assuming the N residuals follow N distinct normal distributions centered on 0 and whose variances are the reciprocal of the weights.
      Returns:
      chi-square value

    • getWeightSquareRoot

      public RealMatrix getWeightSquareRoot()
      Deprecated.
      Gets the square-root of the weight matrix.
      Returns:
      the square-root of the weight matrix.
      Since:
      3.1

    • setCost

      protected void setCost(double cost)
      Deprecated.
      Sets the cost.
      Parameters:
      cost - Cost value.
      Since:
      3.1

    • getCovariances

      @Deprecated public double[][] getCovariances()
      Deprecated.
      As of 3.1. Please use computeCovariances(double[],double) instead.
      Get the covariance matrix of the optimized parameters.
      Returns:
      the covariance matrix.
      Throws:
      SingularMatrixException - if the covariance matrix cannot be computed (singular problem).
      See Also:

    • getCovariances

      @Deprecated public double[][] getCovariances(double threshold)
      Deprecated.
      As of 3.1. Please use computeCovariances(double[],double) instead.
      Get the covariance matrix of the optimized parameters.
      Note that this operation involves the inversion of the JTJ matrix, where J is the Jacobian matrix. The threshold parameter is a way for the caller to specify that the result of this computation should be considered meaningless, and thus trigger an exception.
      Parameters:
      threshold - Singularity threshold.
      Returns:
      the covariance matrix.
      Throws:
      SingularMatrixException - if the covariance matrix cannot be computed (singular problem).

    • computeCovariances

      public double[][] computeCovariances(double[] params, double threshold)
      Deprecated.
      Get the covariance matrix of the optimized parameters.
      Note that this operation involves the inversion of the JTJ matrix, where J is the Jacobian matrix. The threshold parameter is a way for the caller to specify that the result of this computation should be considered meaningless, and thus trigger an exception.
      Parameters:
      params - Model parameters.
      threshold - Singularity threshold.
      Returns:
      the covariance matrix.
      Throws:
      SingularMatrixException - if the covariance matrix cannot be computed (singular problem).
      Since:
      3.1

    • guessParametersErrors

      @Deprecated public double[] guessParametersErrors()
      Deprecated.
      as of version 3.1, computeSigma(double[],double) should be used instead. It should be emphasized that guessParametersErrors and computeSigma are not strictly equivalent.

      Returns an estimate of the standard deviation of each parameter. The returned values are the so-called (asymptotic) standard errors on the parameters, defined as sd(a[i]) = sqrt(S / (n - m) * C[i][i]), where a[i] is the optimized value of the i-th parameter, S is the minimized value of the sum of squares objective function (as returned by getChiSquare()), n is the number of observations, m is the number of parameters and C is the covariance matrix.

      See also Wikipedia, or MathWorld, equations (34) and (35) for a particular case.

      Returns:
      an estimate of the standard deviation of the optimized parameters
      Throws:
      SingularMatrixException - if the covariance matrix cannot be computed.
      NumberIsTooSmallException - if the number of degrees of freedom is not positive, i.e. the number of measurements is less or equal to the number of parameters.

    • computeSigma

      public double[] computeSigma(double[] params, double covarianceSingularityThreshold)
      Deprecated.
      Computes an estimate of the standard deviation of the parameters. The returned values are the square root of the diagonal coefficients of the covariance matrix, sd(a[i]) ~= sqrt(C[i][i]), where a[i] is the optimized value of the i-th parameter, and C is the covariance matrix.
      Parameters:
      params - Model parameters.
      covarianceSingularityThreshold - Singularity threshold (see computeCovariances).
      Returns:
      an estimate of the standard deviation of the optimized parameters
      Throws:
      SingularMatrixException - if the covariance matrix cannot be computed.
      Since:
      3.1

    • optimize

      @Deprecated public PointVectorValuePair optimize(int maxEval, DifferentiableMultivariateVectorFunction f, double[] target, double[] weights, double[] startPoint)
      Deprecated.
      As of 3.1. Please use optimize(int,MultivariateDifferentiableVectorFunction,OptimizationData...) instead.
      Optimize an objective function. Optimization is considered to be a weighted least-squares minimization. The cost function to be minimized is ∑weighti(objectivei - targeti)2
      Specified by:
      optimize in interface BaseMultivariateVectorOptimizer<DifferentiableMultivariateVectorFunction>
      Overrides:
      optimize in class BaseAbstractMultivariateVectorOptimizer<DifferentiableMultivariateVectorFunction>
      Parameters:
      maxEval - Maximum number of function evaluations.
      f - Objective function.
      target - Target value for the objective functions at optimum.
      weights - Weights for the least squares cost computation.
      startPoint - Start point for optimization.
      Returns:
      the point/value pair giving the optimal value for objective function.

    • optimize

      @Deprecated public PointVectorValuePair optimize(int maxEval, MultivariateDifferentiableVectorFunction f, double[] target, double[] weights, double[] startPoint)
      Deprecated.
      As of 3.1. Please use optimize(int,MultivariateDifferentiableVectorFunction,OptimizationData...) instead.
      Optimize an objective function. Optimization is considered to be a weighted least-squares minimization. The cost function to be minimized is ∑weighti(objectivei - targeti)2
      Parameters:
      maxEval - Maximum number of function evaluations.
      f - Objective function.
      target - Target value for the objective functions at optimum.
      weights - Weights for the least squares cost computation.
      startPoint - Start point for optimization.
      Returns:
      the point/value pair giving the optimal value for objective function.
      Throws:
      DimensionMismatchException - if the start point dimension is wrong.
      TooManyEvaluationsException - if the maximal number of evaluations is exceeded.
      NullArgumentException - if any argument is null.

    • optimizeInternal

      @Deprecated protected PointVectorValuePair optimizeInternal(int maxEval, MultivariateDifferentiableVectorFunction f, OptimizationData... optData)
      Deprecated.
      As of 3.1. Override is necessary only until this class's generic argument is changed to MultivariateDifferentiableVectorFunction.
      Optimize an objective function. Optimization is considered to be a weighted least-squares minimization. The cost function to be minimized is ∑weighti(objectivei - targeti)2
      Parameters:
      maxEval - Allowed number of evaluations of the objective function.
      f - Objective function.
      optData - Optimization data. The following data will be looked for:
      Returns:
      the point/value pair giving the optimal value of the objective function.
      Throws:
      TooManyEvaluationsException - if the maximal number of evaluations is exceeded.
      DimensionMismatchException - if the target, and weight arguments have inconsistent dimensions.
      Since:
      3.1
      See Also:

    • setUp

      protected void setUp()
      Deprecated.
      Method which a subclass must override whenever its internal state depend on the input parsed by this base class. It will be called after the parsing step performed in the optimize method and just before BaseAbstractMultivariateVectorOptimizer.doOptimize().
      Overrides:
      setUp in class BaseAbstractMultivariateVectorOptimizer<DifferentiableMultivariateVectorFunction>

    • computeResiduals

      protected double[] computeResiduals(double[] objectiveValue)
      Deprecated.
      Computes the residuals. The residual is the difference between the observed (target) values and the model (objective function) value. There is one residual for each element of the vector-valued function.
      Parameters:
      objectiveValue - Value of the the objective function. This is the value returned from a call to computeObjectiveValue (whose array argument contains the model parameters).
      Returns:
      the residuals.
      Throws:
      DimensionMismatchException - if params has a wrong length.
      Since:
      3.1