org.opensha.sra.riskmaps.func
Class DiscreteInterpExterpFunc

java.lang.Object
  org.opensha.sra.riskmaps.func.DiscreteFunction
      org.opensha.sra.riskmaps.func.DiscreteInterpExterpFunc

All Implemented Interfaces:

FunctionAPI

public class DiscreteInterpExterpFunc
extends DiscreteFunction

This class behaves much like the DiscreteFunction class, however it allows for values between the discrete domain values to be interpolated, and values outside the domain altogether to be extrapolated. This interpolation/extrapolation can be done in different arithmetic spaces to allow for a “best fit” regression curve. Extrapolation follows one of four different approaches for each side of the curve. See the predefined constants for a list of options and their meanings.

 -=* CHANGE LOG *=-
 07/07/08 -- EMM: Original implementation.

Version:

0.0.1

Author:

Eric Martinez

Field Summary

static int	EXTRAP_EXTEND Extrapolation should continue a linear regression of the final points.
static int	EXTRAP_FLATTEN Extrapolation should assume the constant value of the endpoint.
static int	EXTRAP_INFINITY Extrapolation should jump to +/- infinity (slope dependent).
static int	EXTRAP_ZERO Extrapolation should just drop values to zero.
protected int	extrapAbove
protected int	extrapBelow
protected int	interpType
static int	LINEAR_LINEAR Identifies interpolation in linear-x, linear-y domains.
static int	LINEAR_LOG Identifies interpolation in linear-x, logarithmic-y domains.
static int	LOG_LINEAR Identifies interpolation in logarithmic-x, linear-y domains.
static int	LOG_LOG Identifies interpolation in logarithmic-x, logarithmic-y domains.

Fields inherited from class org.opensha.sra.riskmaps.func.DiscreteFunction

xvals, yvals

Constructor Summary

DiscreteInterpExterpFunc(double[] xvals, double[] yvals)
Creates a discrete function with interpolation and extrapolation methods based on the given xvals and yvals.

DiscreteInterpExterpFunc(double[] xvals, double[] yvals, String name)
Creates a discrete function with interpolation and extrapolation methods based on the given xvals, yvals and name.

DiscreteInterpExterpFunc(double[] xvals, double[] yvals, String name, int interpType, int extrapBelow, int extrapAbove)
Creates a discrete function with interpolation and extrapolation methods based on the given xvals, yvals and name.

Method Summary

static DiscreteInterpExterpFunc	fromArbDistFunc(ArbitrarilyDiscretizedFunc func)
int	getExtrapAbove()
int	getExtrapBelow()
int	getInterp()
void	setExtrapMethod(int extrapAbove, int extrapBelow) Attempts to set the extrapolation methods used for this discrete function.
void	setInterpMethod(int interpType) Attempts to set the interpolataion method used for this discrete function.
double	valueOf(double x) For the given value x returns the function value f(x) for this function.
double[]	valueOf(double[] x) For the given array of x values, a corresponding array of f(x) values is created and returned according to the computation methods described in valueOf(double x).

Methods inherited from class org.opensha.sra.riskmaps.func.DiscreteFunction

equals, getDomain, getName, getRange, getXVals, getYVals, toString

Methods inherited from class java.lang.Object

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

Field Detail

LINEAR_LINEAR

public static final int LINEAR_LINEAR

Identifies interpolation in linear-x, linear-y domains.

See Also:

Constant Field Values

LINEAR_LOG

public static final int LINEAR_LOG

Identifies interpolation in linear-x, logarithmic-y domains.

See Also:

Constant Field Values

LOG_LINEAR

public static final int LOG_LINEAR

Identifies interpolation in logarithmic-x, linear-y domains.

See Also:

Constant Field Values

LOG_LOG

public static final int LOG_LOG

Identifies interpolation in logarithmic-x, logarithmic-y domains.

See Also:

Constant Field Values

EXTRAP_ZERO

public static final int EXTRAP_ZERO

Extrapolation should just drop values to zero.

See Also:

Constant Field Values

EXTRAP_FLATTEN

public static final int EXTRAP_FLATTEN

Extrapolation should assume the constant value of the endpoint.

See Also:

Constant Field Values

EXTRAP_EXTEND

public static final int EXTRAP_EXTEND

Extrapolation should continue a linear regression of the final points.

See Also:

Constant Field Values

EXTRAP_INFINITY

public static final int EXTRAP_INFINITY

Extrapolation should jump to +/- infinity (slope dependent).

See Also:

Constant Field Values

interpType

protected int interpType

extrapBelow

protected int extrapBelow

extrapAbove

protected int extrapAbove

Constructor Detail

DiscreteInterpExterpFunc

public DiscreteInterpExterpFunc(double[] xvals,
                                double[] yvals)

Creates a discrete function with interpolation and extrapolation methods based on the given xvals and yvals. The name of this function will be the default name, and the interpolation and extrapoltion methods will be the defaults as well.

Parameters:

xvals - The x-values of this discrete function.

yvals - The y-values of this descrete function.

DiscreteInterpExterpFunc

public DiscreteInterpExterpFunc(double[] xvals,
                                double[] yvals,
                                String name)

Creates a discrete function with interpolation and extrapolation methods based on the given xvals, yvals and name. The interpolation and extrapolation methods will be the defaults.

Parameters:

xvals - The x-values of this discrete function.

yvals - The y-values of this discrete function.

name - The name of this discrete function.

DiscreteInterpExterpFunc

public DiscreteInterpExterpFunc(double[] xvals,
                                double[] yvals,
                                String name,
                                int interpType,
                                int extrapBelow,
                                int extrapAbove)

Creates a discrete function with interpolation and extrapolation methods based on the given xvals, yvals and name. Assuming the interpType, extrapBelow and/or extrapAbove correspond to a well-known method, they will be used as the respective interpolation/extrapolation methods. Otherwise the defaults will be used.

Parameters:

xvals - The x-values of this discrete function.

yvals - The y-values of this discrete function.

name - The name of this discrete function.

interpType - The arithmetic space in which to interpolate.

extrapBelow - The method of extrapolation for values below the lower bound of the function domain.

extrapAbove - The method of extrapolation for the values above the upper bound of the function domain.

Method Detail

setInterpMethod

public void setInterpMethod(int interpType)

Attempts to set the interpolataion method used for this discrete function. If the given interpType does not correspond to a known interpolation type, then nothing is done.

Parameters:

interpType - The new interpolation type to use.

setExtrapMethod

public void setExtrapMethod(int extrapAbove,
                            int extrapBelow)

Attempts to set the extrapolation methods used for this discrete function. If the given extrapBelow or extrapAbove does not correspond to a known extrapolation method, then the respective value is not set.

Parameters:

extrapBelow - The extrapolation method used for values below the lower bound of the domain of this discrete function.

extrapAbove - The extrapolation method used for values above the upper bound of the domain of this discrete function.

getInterp

public int getInterp()

Returns:

The current interpolation scheme used for this function.

getExtrapBelow

public int getExtrapBelow()

Returns:

The current extrapolation method used for values below the lower bound of the domain of this function.

getExtrapAbove

public int getExtrapAbove()

Returns:

The current extrapolation method used for values above the upper bound of the domain of this function.

valueOf

public double valueOf(double x)

For the given value x returns the function value f(x) for this function. If x is one of the discrete defining values of this function, its corresponding discrete y-value is returned. If x is somehwere within the bounds of the domain of this discrete function, then an interpolation is performed according to the current interpolation type and that interpolated value is returned. If x is above or below the upper or lower bounds of this discrete function's domain (respectively), then the funcntion value of x is extrapolated according to the current extrapolation method for values above/below the domain (respectively).

Specified by:

valueOf in interface FunctionAPI

Overrides:

valueOf in class DiscreteFunction

Parameters:

x - The value for which we want to know f(x).

Returns:

The function value f(x), interpolated or extrapolated as needed.

valueOf

public double[] valueOf(double[] x)

For the given array of x values, a corresponding array of f(x) values is created and returned according to the computation methods described in valueOf(double x).

Parameters:

x - The array of x-values for which you want to know y-values.

Returns:

An array of function values for the input array of x.

fromArbDistFunc

public static DiscreteInterpExterpFunc fromArbDistFunc(ArbitrarilyDiscretizedFunc func)