com.pervasive.datarush.operators.assertion

Class RelativeErrorBound

java.lang.Object

com.pervasive.datarush.operators.assertion.RelativeErrorBound

All Implemented Interfaces:

FloatingPointTolerance

public class RelativeErrorBound
extends Object
implements FloatingPointTolerance

Checks equivalence based on the relative error between two floating point numbers. Two values are considered equal if this error is less than a fixed bound.

Specifically, for two values v1 and v2 where v1 > v2 and a non-negative bound e, it is checked whether Math.abs((v1 - v2)/v1) <= e

Field Summary

Fields inherited from interface com.pervasive.datarush.operators.assertion.FloatingPointTolerance

EXACT

Constructor Summary

Constructors

Constructor and Description
RelativeErrorBound(double epsilon) Define a fixed bound on the relative error of two floating point numbers.

Method Summary

Modifier and Type	Method and Description
double	getEpsilon() Gets the maximum percentage difference allowed between two values.
boolean	isWithinTolerance(BigDecimal value1, BigDecimal value2) Indicates whether the specified values are equal under the tolerance.
boolean	isWithinTolerance(double value1, double value2) Indicates whether the specified values are equal under the tolerance.
boolean	isWithinTolerance(float value1, float value2) Indicates whether the specified values are equal under the tolerance.

Methods inherited from class java.lang.Object

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

Constructor Detail

RelativeErrorBound

public RelativeErrorBound(double epsilon)

Define a fixed bound on the relative error of two floating point numbers.

Parameters:

epsilon - the fixed bound on the error

Throws:

com.pervasive.datarush.graphs.physical.InvalidPropertyValueException - if epsilon is negative.

Method Detail

isWithinTolerance

public boolean isWithinTolerance(float value1,
                                 float value2)

Description copied from interface: FloatingPointTolerance

Indicates whether the specified values are equal under the tolerance.

Specified by:

isWithinTolerance in interface FloatingPointTolerance

Parameters:

value1 - the first value to compare

value2 - the second value to compare

Returns:

true if the two values lie with the error bounds, otherwise false

isWithinTolerance

public boolean isWithinTolerance(double value1,
                                 double value2)

Description copied from interface: FloatingPointTolerance

Indicates whether the specified values are equal under the tolerance.

Specified by:

isWithinTolerance in interface FloatingPointTolerance

Parameters:

value1 - the first value to compare

value2 - the second value to compare

Returns:

true if the two values lie with the error bounds, otherwise false

isWithinTolerance

public boolean isWithinTolerance(BigDecimal value1,
                                 BigDecimal value2)

Description copied from interface: FloatingPointTolerance

Indicates whether the specified values are equal under the tolerance.

Specified by:

isWithinTolerance in interface FloatingPointTolerance

Parameters:

value1 - the first value to compare

value2 - the second value to compare

Returns:

true if the two values lie with the error bounds, otherwise false

getEpsilon

public double getEpsilon()

Gets the maximum percentage difference allowed between two values.

Returns:

the relative error bound