Package org.ddogleg.optimization

Class DerivativeChecker

java.lang.Object

org.ddogleg.optimization.DerivativeChecker

public class DerivativeChecker
extends Object

Used to validate an algebraic Jacobian numerically.

Constructor Summary

Constructors

Constructor	Description
DerivativeChecker()

Method Summary

Modifier and Type	Method	Description
static boolean	gradient(FunctionNtoS func, FunctionNtoN gradient, double[] param, double tol)	Compares the passed in gradient function to a numerical calculation.
static boolean	gradient(FunctionNtoS func, FunctionNtoN gradient, double[] param, double tol, double differenceScale)
static boolean	gradientR(FunctionNtoS func, FunctionNtoN gradient, double[] param, double tol)	Compares the passed in gradient function to a numerical calculation.
static boolean	gradientR(FunctionNtoS func, FunctionNtoN gradient, double[] param, double tol, double differenceScale)
static <S extends DMatrix> boolean	jacobian(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol)
static <S extends DMatrix> boolean	jacobian(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol, double differenceScale)
static <S extends DMatrix> void	jacobianPrint(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol)
static <S extends DMatrix> void	jacobianPrint(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol, double differenceScale)
static <S extends DMatrix> void	jacobianPrintR(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol)	Prints out the difference using a relative error threshold
static <S extends DMatrix> void	jacobianPrintR(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol, double differenceScale)
static <S extends DMatrix> boolean	jacobianR(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol)	Checks the jacobian using a relative error threshold.
static <S extends DMatrix> boolean	jacobianR(FunctionNtoM func, FunctionNtoMxN<S> jacobian, double[] param, double tol, double differenceScale)	Checks the jacobian using a relative error threshold.

Methods inherited from class java.lang.Object

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

Constructor Details

DerivativeChecker

public DerivativeChecker()

Method Details

jacobianPrint

public static <S extends DMatrix> void jacobianPrint(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol)

jacobianPrint

public static <S extends DMatrix> void jacobianPrint(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol,
 double differenceScale)

jacobian

public static <S extends DMatrix> boolean jacobian(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol)

jacobian

public static <S extends DMatrix> boolean jacobian(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol,
 double differenceScale)

jacobianPrintR

public static <S extends DMatrix> void jacobianPrintR(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol)

Prints out the difference using a relative error threshold

Parameters:

tol - fractional difference

jacobianPrintR

public static <S extends DMatrix> void jacobianPrintR(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol,
 double differenceScale)

jacobianR

public static <S extends DMatrix> boolean jacobianR(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol)

Checks the jacobian using a relative error threshold.

Parameters:

tol - fractional difference

jacobianR

public static <S extends DMatrix> boolean jacobianR(FunctionNtoM func,
 FunctionNtoMxN<S> jacobian,
 double[] param,
 double tol,
 double differenceScale)

Checks the jacobian using a relative error threshold.

Parameters:

tol - fractional difference

gradient

public static boolean gradient(FunctionNtoS func,
 FunctionNtoN gradient,
 double[] param,
 double tol)

Compares the passed in gradient function to a numerical calculation. Comparison is done using an absolute value.

Returns:

true for within tolerance and false otherwise

gradient

public static boolean gradient(FunctionNtoS func,
 FunctionNtoN gradient,
 double[] param,
 double tol,
 double differenceScale)

gradientR

public static boolean gradientR(FunctionNtoS func,
 FunctionNtoN gradient,
 double[] param,
 double tol)

Compares the passed in gradient function to a numerical calculation. Comparison is done using an absolute value.

Returns:

true for within tolerance and false otherwise

gradientR

public static boolean gradientR(FunctionNtoS func,
 FunctionNtoN gradient,
 double[] param,
 double tol,
 double differenceScale)