net.sourceforge.webcompmath.functions
Class TableFunction

java.lang.Object
  net.sourceforge.webcompmath.functions.FunctionParserExtension
      net.sourceforge.webcompmath.functions.TableFunction

All Implemented Interfaces:

java.io.Serializable, ExpressionCommand, Function, MathObject, ParserExtension

Direct Known Subclasses:

WCMTableFunctionBean

public class TableFunction
extends FunctionParserExtension

A TableFunction is a function that is specified by a table of (x,y)-points. Values are interpolated between the specified x-values. This can be done in several differnt ways; the method that is used is controlled by the "Style" property. Since a TableFunction extends FunctionParserExtension, a TableFunction can be added to a Parser (provided it has a name), and it can then be used in expressions parsed by that parser. Note that this class is meant to be used for functions that are defined by a fairly small number of points, since each function evaluation involves a linear search through the list of x-values of the defining points.

See Also:

Serialized Form

Field Summary

static int	PIECEWISE_LINEAR If the style of the function is set to PIECEWISE_LINEAR, then linear interpolation is used to find the value of the functions for x-values between the points that define the function.
static int	SMOOTH If the style of the function is set to SMOOTH, then cubic interpolation is used to find the value of the functions for x-values between the points that define the function.
static int	STEP If the style of the function is set to STEP, then the function is piecewise constant, and the value of the function at x is taken from the nearest point in the list of points that define the function.
static int	STEP_LEFT If the style of the function is set to STEP_LEFT, then the function is piecewise constant, and the value of the function at x is taken from the nearest point to the left in the list of points that define the function.
static int	STEP_RIGHT If the style of the function is set to STEP_RIGHT, then the function is piecewise constant, and the value of the function at x is taken from the nearest point to the right in the list of points that define the function.

Fields inherited from class net.sourceforge.webcompmath.functions.FunctionParserExtension

name

Constructor Summary

TableFunction()
Create a TableFunction with SMOOTH style and no points.

TableFunction(int style)
Create a TableFunction with specified style and no points.

Method Summary

void	addIntervals(int intervals, double xmin, double xmax) Add points to the table.
int	addPoint(double x, double y) Add a point with the specified x and y coordinates.
void	addPoints(double[] xCoords, double[] yCoords) Add points to the table.
void	apply(StackOfDouble stack, Cases cases) Override method apply() from interface FunctionParserExtension, to handle cases properly.
void	copyDataFrom(TableFunction source) Copy data from another TableFunction, except that the name of the funcion is not duplicated.
boolean	dependsOn(Variable wrt) Returns false.
Function	derivative(int wrt) Compute the derivative of this function.
Function	derivative(Variable wrt) Returns null.
int	findPoint(double x) If there is a point in the list with x-coordinate x, then this function returns the index of that point in the list (where the index of the first point is zero).
int	getArity() Returns the arity of the function, which is 1.
int	getPointCount() Gets the number of points in the table.
int	getStyle() Get the style of this TableFunction, which specifies how values are interpolated between points on the curve.
double	getVal(double x) Get the value of the function at x, using interpolation if x lies between two x-coordinates in the list of points that define the function.
double	getVal(double[] params) Get the value of the function at the specified parameter value.
double	getValueWithCases(double[] params, Cases cases) Get the value of the function at the specified parameter value.
double	getX(int i) Get the x-coordinate in the i-th point, where the first point is number zero.
double	getY(int i) Get the y-coordinate in the i-th point, where the first point is number zero.
void	removeAllPoints() Remove all points.
void	removePointAt(int i) Removes the i-th point from the list of points.
void	setStyle(int style) Set the style of this TableFunction, to specify how values are interpolated between points on the curve.
void	setY(int i, double y) Set the y-coordinate in the i-th point to y, where the first point is number zero.

Methods inherited from class net.sourceforge.webcompmath.functions.FunctionParserExtension

appendOutputString, compileDerivative, doParse, extent, getName, setName, setParensCanBeOptional

Methods inherited from class java.lang.Object

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

Field Detail

SMOOTH

public static final int SMOOTH

If the style of the function is set to SMOOTH, then cubic interpolation is used to find the value of the functions for x-values between the points that define the function.

See Also:

Constant Field Values

PIECEWISE_LINEAR

public static final int PIECEWISE_LINEAR

If the style of the function is set to PIECEWISE_LINEAR, then linear interpolation is used to find the value of the functions for x-values between the points that define the function.

See Also:

Constant Field Values

STEP

public static final int STEP

If the style of the function is set to STEP, then the function is piecewise constant, and the value of the function at x is taken from the nearest point in the list of points that define the function.

See Also:

Constant Field Values

STEP_LEFT

public static final int STEP_LEFT

If the style of the function is set to STEP_LEFT, then the function is piecewise constant, and the value of the function at x is taken from the nearest point to the left in the list of points that define the function.

See Also:

Constant Field Values

STEP_RIGHT

public static final int STEP_RIGHT

If the style of the function is set to STEP_RIGHT, then the function is piecewise constant, and the value of the function at x is taken from the nearest point to the right in the list of points that define the function.

See Also:

Constant Field Values

Constructor Detail

TableFunction

public TableFunction()

Create a TableFunction with SMOOTH style and no points.

TableFunction

public TableFunction(int style)

Create a TableFunction with specified style and no points.

Parameters:

style - The style for the function: SMOOTH, PIECEWISE_LINEAR, STEP, STEP_LEFT, or STEP_RIGHT.

Method Detail

copyDataFrom

public void copyDataFrom(TableFunction source)

Copy data from another TableFunction, except that the name of the funcion is not duplicated. The new TableFunction is nameless.

Parameters:

source - table to copy from

setStyle

public void setStyle(int style)

Set the style of this TableFunction, to specify how values are interpolated between points on the curve.

Parameters:

style - One of the style constants SMOOTH, PIECEWISE_LINEAR, STEP, STEP_LEFT, STEP_RIGHT. Other values are ignored.

getStyle

public int getStyle()

Get the style of this TableFunction, which specifies how values are interpolated between points on the curve.

Returns:

The style of this TableFunction. This is one of the constants SMOOTH, PIECEWISE_LINEAR, STEP, STEP_LEFT, or STEP_RIGHT.

addPoints

public void addPoints(double[] xCoords,
                      double[] yCoords)

Add points to the table. The x-coordinates of the points are taken from the xCoords array. The y-coordinate for the i-th point is yCoords[i], if an i-th position exists in this array. Otherwise, the y-coordinate is is zero. (Note that if xCoords[i] duplicates an x-value already in the table, then no new point is added but the corresponging y-value is changed.)

Parameters:

xCoords - A list of x-coordinates to be added to the table. If this is null, then nothing is done.

yCoords - The value of yCoords[i], if it exists, is the y-coordinate corresponding to xCoords[i]. Otherwise, the y-coordinate is undefined. This can be null, in which case all y-coordinates are zero.

addIntervals

public void addIntervals(int intervals,
                         double xmin,
                         double xmax)

Add points to the table. The number of points added is intervals + 1. The x-coordinates are evenly spaced between xmin and xmax. The y-coordinates are zero.

Parameters:

intervals - The number of intervals. The number of points added is intervals + 1. The value should be at least 1. If not, nothing is done.

xmin - The minimim x-coordinate for added points.

xmax - The maximum x-coodinate for added points. Should be greater than xmin, for efficiency, but no error occurs if it is not.

addPoint

public int addPoint(double x,
                    double y)

Add a point with the specified x and y coordinates. If a point with the given x coordinate already exists in the table, then no new point is added, but the associated y-value is changed. (If x is Double.NaN, then no change is made and the return value is -1.)

Parameters:

x - The x-coordinate of the point to be added or modified.

y - The y-coordinate of the point.

Returns:

the position of the point in the list of points, where the first point is at position zero.

getPointCount

public int getPointCount()

Gets the number of points in the table.

Returns:

number of points

getX

public double getX(int i)

Get the x-coordinate in the i-th point, where the first point is number zero. Throws an IllegalArgumentException if i is less than zero or greater than or equal to the number of points.

Parameters:

i - index of point

Returns:

it's x coordinate

getY

public double getY(int i)

Get the y-coordinate in the i-th point, where the first point is number zero. Throws an IllegalArgumentException if i is less than zero or greater than or equal to the number of points.

Parameters:

i - index of point

Returns:

it's y coordinate

setY

public void setY(int i,
                 double y)

Set the y-coordinate in the i-th point to y, where the first point is number zero. Throws an IllegalArgumentException if i is less than zero or greater than or equal to the number of points.

Parameters:

i - index of point

y - y coordinate to set

findPoint

public int findPoint(double x)

If there is a point in the list with x-coordinate x, then this function returns the index of that point in the list (where the index of the first point is zero). If there is no such point, then -1 is returned.

Parameters:

x - x coordinate to check

Returns:

index of the point

removePointAt

public void removePointAt(int i)

Removes the i-th point from the list of points. Throws an IllegalArgumentException if i is less than zero or greater than or equal to the number of points.

Parameters:

i - index of point to remove

removeAllPoints

public void removeAllPoints()

Remove all points. The resulting function is undefined everywhere.

getVal

public double getVal(double x)

Get the value of the function at x, using interpolation if x lies between two x-coordinates in the list of points that define the function. If x is outside the range of x-coords in the table, the value of the function is Double.NaN.

Parameters:

x - x coordinate

Returns:

value of function

getValueWithCases

public double getValueWithCases(double[] params,
                                Cases cases)

Get the value of the function at the specified parameter value.

Parameters:

params - should be an array of length 1 holding the argument of the function. However if the length is greater than one, the extra arguments are simply ignored.

cases - if cases is non-null, a case value is stored here, for help in continuity computations.

Returns:

value of function

getVal

public double getVal(double[] params)

Get the value of the function at the specified parameter value.

Parameters:

params - should be an array of length 1 holding the argument of the function. However if the length is greater than one, the extra arguments are simply ignored.

Returns:

value of function

derivative

public Function derivative(int wrt)

Compute the derivative of this function. The value of the parameter, wrt, must be 1 or an IllegalArguemntException will be thrown.

Parameters:

wrt - arg. number with respect to

Returns:

the derivative

derivative

public Function derivative(Variable wrt)

Returns null. It really should be the constant function zero, but I don't expect this ever to be called. Since dependsOn(wrt) returns false, it will never be called within the JCM system.

Parameters:

wrt - variable with respect to

Returns:

the derivative

dependsOn

public boolean dependsOn(Variable wrt)

Returns false.

Parameters:

wrt - variable with respect to

Returns:

always false in this case

getArity

public int getArity()

Returns the arity of the function, which is 1.

Returns:

arity of function

apply

public void apply(StackOfDouble stack,
                  Cases cases)

Override method apply() from interface FunctionParserExtension, to handle cases properly. Not meant to be called directly.

Specified by:

apply in interface ExpressionCommand

Overrides:

apply in class FunctionParserExtension

Parameters:

stack - stack to use

cases - cases to use