com.opengamma.strata.math.impl.statistics.leastsquare

Class GeneralizedLeastSquare

java.lang.Object

com.opengamma.strata.math.impl.statistics.leastsquare.GeneralizedLeastSquare

public class GeneralizedLeastSquare
extends Object

Generalized least square method.

Constructor Summary

Constructors

Constructor and Description
GeneralizedLeastSquare() Creates an instance.

Method Summary

Modifier and Type	Method and Description
<T> GeneralizedLeastSquareResults<T>	solve(List<T> x, List<Double> y, List<Double> sigma, List<Function<T,Double>> basisFunctions)
<T> GeneralizedLeastSquareResults<T>	solve(List<T> x, List<Double> y, List<Double> sigma, List<Function<T,Double>> basisFunctions, double lambda, int differenceOrder) Generalised least square with penalty on (higher-order) finite differences of weights.
<T> GeneralizedLeastSquareResults<T>	solve(List<T> x, List<Double> y, List<Double> sigma, List<Function<T,Double>> basisFunctions, int[] sizes, double[] lambda, int[] differenceOrder) Specialist method used mainly for solving multidimensional P-spline problems where the basis functions (B-splines) span a N-dimension space, and the weights sit on an N-dimension grid and are treated as a N-order tensor rather than a vector, so k-order differencing is done for each tensor index while varying the other indices.
<T> GeneralizedLeastSquareResults<T>	solve(T[] x, double[] y, double[] sigma, List<Function<T,Double>> basisFunctions)
<T> GeneralizedLeastSquareResults<T>	solve(T[] x, double[] y, double[] sigma, List<Function<T,Double>> basisFunctions, double lambda, int differenceOrder) Generalised least square with penalty on (higher-order) finite differences of weights.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

GeneralizedLeastSquare

public GeneralizedLeastSquare()

Creates an instance.

Method Detail

solve

public <T> GeneralizedLeastSquareResults<T> solve(T[] x,
                                                  double[] y,
                                                  double[] sigma,
                                                  List<Function<T,Double>> basisFunctions)

Type Parameters:

T - The type of the independent variables (e.g. Double, double[], DoubleArray etc)

Parameters:

x - independent variables

y - dependent (scalar) variables

sigma - (Gaussian) measurement error on dependent variables

basisFunctions - set of basis functions - the fitting function is formed by these basis functions times a set of weights

Returns:

the results of the least square

solve

public <T> GeneralizedLeastSquareResults<T> solve(T[] x,
                                                  double[] y,
                                                  double[] sigma,
                                                  List<Function<T,Double>> basisFunctions,
                                                  double lambda,
                                                  int differenceOrder)

Generalised least square with penalty on (higher-order) finite differences of weights.

Type Parameters:

T - The type of the independent variables (e.g. Double, double[], DoubleArray etc)

Parameters:

x - independent variables

y - dependent (scalar) variables

sigma - (Gaussian) measurement error on dependent variables

basisFunctions - set of basis functions - the fitting function is formed by these basis functions times a set of weights

lambda - strength of penalty function

differenceOrder - difference order between weights used in penalty function

Returns:

the results of the least square

solve

public <T> GeneralizedLeastSquareResults<T> solve(List<T> x,
                                                  List<Double> y,
                                                  List<Double> sigma,
                                                  List<Function<T,Double>> basisFunctions)

Type Parameters:

T - The type of the independent variables (e.g. Double, double[], DoubleArray etc)

Parameters:

x - independent variables

y - dependent (scalar) variables

sigma - (Gaussian) measurement error on dependent variables

basisFunctions - set of basis functions - the fitting function is formed by these basis functions times a set of weights

Returns:

the results of the least square

solve

public <T> GeneralizedLeastSquareResults<T> solve(List<T> x,
                                                  List<Double> y,
                                                  List<Double> sigma,
                                                  List<Function<T,Double>> basisFunctions,
                                                  double lambda,
                                                  int differenceOrder)

Generalised least square with penalty on (higher-order) finite differences of weights.

Type Parameters:

T - The type of the independent variables (e.g. Double, double[], DoubleArray etc)

Parameters:

x - independent variables

y - dependent (scalar) variables

sigma - (Gaussian) measurement error on dependent variables

basisFunctions - set of basis functions - the fitting function is formed by these basis functions times a set of weights

lambda - strength of penalty function

differenceOrder - difference order between weights used in penalty function

Returns:

the results of the least square

solve

public <T> GeneralizedLeastSquareResults<T> solve(List<T> x,
                                                  List<Double> y,
                                                  List<Double> sigma,
                                                  List<Function<T,Double>> basisFunctions,
                                                  int[] sizes,
                                                  double[] lambda,
                                                  int[] differenceOrder)

Specialist method used mainly for solving multidimensional P-spline problems where the basis functions (B-splines) span a N-dimension space, and the weights sit on an N-dimension grid and are treated as a N-order tensor rather than a vector, so k-order differencing is done for each tensor index while varying the other indices.

Type Parameters:

T - The type of the independent variables (e.g. Double, double[], DoubleArray etc)

Parameters:

x - independent variables

y - dependent (scalar) variables

sigma - (Gaussian) measurement error on dependent variables

basisFunctions - set of basis functions - the fitting function is formed by these basis functions times a set of weights

sizes - The size the weights tensor in each dimension (the product of this must equal the number of basis functions)

lambda - strength of penalty function in each dimension

differenceOrder - difference order between weights used in penalty function for each dimension

Returns:

the results of the least square