Package com.opengamma.strata.math.rootfind

Interface NewtonVectorRootFinder

All Known Implementing Classes:

BaseNewtonVectorRootFinder, BroydenVectorRootFinder, NewtonDefaultVectorRootFinder, ShermanMorrisonVectorRootFinder

public interface NewtonVectorRootFinder

Performs Newton-Raphson style multi-dimensional root finding.

This uses the Jacobian matrix as a basis for some parts of the iterative process.

Method Summary

Modifier and Type	Method	Description
static NewtonVectorRootFinder	broyden()	Obtains an instance of the Broyden root finder.
static NewtonVectorRootFinder	broyden(double absoluteTol, double relativeTol, int maxSteps)	Obtains an instance of the Broyden root finder specifying the tolerances.
static NewtonVectorRootFinder	broyden(double absoluteTol, double relativeTol, int maxSteps, Decomposition<?> decomposition)	Obtains an instance of the Broyden root finder specifying the tolerances.
DoubleArray	findRoot(Function<DoubleArray,DoubleArray> function, DoubleArray startPosition)	Finds the root from the specified start position.
DoubleArray	findRoot(Function<DoubleArray,DoubleArray> function, Function<DoubleArray,DoubleMatrix> jacobianFunction, DoubleArray startPosition)	Finds the root from the specified start position.

Method Detail

broyden

static NewtonVectorRootFinder broyden()

Obtains an instance of the Broyden root finder.

This uses SV decomposition and standard tolerances.

Returns:

the root finder

broyden

static NewtonVectorRootFinder broyden(double absoluteTol,
                                      double relativeTol,
                                      int maxSteps)

Obtains an instance of the Broyden root finder specifying the tolerances.

This uses SV decomposition.

Parameters:

absoluteTol - the absolute tolerance

relativeTol - the relative tolerance

maxSteps - the maximum steps

Returns:

the root finder

broyden

static NewtonVectorRootFinder broyden(double absoluteTol,
                                      double relativeTol,
                                      int maxSteps,
                                      Decomposition<?> decomposition)

Obtains an instance of the Broyden root finder specifying the tolerances.

Parameters:

absoluteTol - the absolute tolerance

relativeTol - the relative tolerance

maxSteps - the maximum steps

decomposition - the decomposition function

Returns:

the root finder

findRoot

DoubleArray findRoot(Function<DoubleArray,DoubleArray> function,
                     DoubleArray startPosition)

Finds the root from the specified start position.

This applies the specified function to find the root. Note if multiple roots exist which one is found will depend on the start position.

Parameters:

function - the vector function

startPosition - the start position of the root finder for

Returns:

the vector root of the collection of functions

Throws:

MathException - if unable to find the root, such as if unable to converge

findRoot

DoubleArray findRoot(Function<DoubleArray,DoubleArray> function,
                     Function<DoubleArray,DoubleMatrix> jacobianFunction,
                     DoubleArray startPosition)

Finds the root from the specified start position.

This applies the specified function and Jacobian function to find the root. Note if multiple roots exist which one is found will depend on the start position.

Parameters:

function - the vector function

jacobianFunction - the function to calculate the Jacobian

startPosition - the start position of the root finder for

Returns:

the vector root of the collection of functions

Throws:

MathException - if unable to find the root, such as if unable to converge