Package com.aquima.interactions.expressions.runtime

Class FloatMath

java.lang.Object

com.aquima.interactions.expressions.runtime.FloatMath

public final class FloatMath
extends Object

Floating point fixing for enhanced precision.

Since:

5.0

Author:

F. vd Meer, E. de Lang

Method Summary

Modifier and Type	Method	Description
static double	floatFix(double doubleValue)	Round a number after 15 digits, including the digits before the decimal separator.
static double	getFactorFor15DigitShift(double doubleValue)	The method calculates the factor need, for a number to move it's first digit 15 positions before the decimal-point.

Methods inherited from class java.lang.Object

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

Method Detail

floatFix

public static double floatFix(double doubleValue)

Round a number after 15 digits, including the digits before the decimal separator. This will remove most cumulative floating-point errors due to floating-point calculations. The size of floating point error can increase into the 14th digit after more than 6 floating-point operations. For minimal error, it is recommended to use this function after each floating-point calculation. Minimally, it to be call before any compare or modulus operation to have any effect.

Parameters:

doubleValue - Value to be rounded

Returns:

double Rounded value

getFactorFor15DigitShift

public static double getFactorFor15DigitShift(double doubleValue)

The method calculates the factor need, for a number to move it's first digit 15 positions before the decimal-point. The maximum factor return is a 15 digit left-shift. This factor equals: Math.pow(10d, 14 - Math.floor(Math.log(Math.abs(value)) / Math.log(10.0))) For performance reasons and to minimize floating point errors, this method is implemented without the use of Math.log and Math.pow.

Parameters:

doubleValue - to be shifted double

Returns:

double the factor