org.whaka.util

Class DoubleMath

java.lang.Object

org.whaka.util.DoubleMath

public final class DoubleMath
extends java.lang.Object

Class provides static utility methods to perform simple numerical operations on double value. For example, direct double comparison is not safe, so double values need to be compared with a delta. This class provides methods for common check and compare operations.

Note:

• roundTo(double, int)
• compare(double, double, double)
• equals(Double, Double, double)

These methods take argument to configure "accuracy" of an operation. But!

• round(double)
• compare(double, double)
• equals(Double, Double)

These methods select appropriate "accuracy" automatically. Note again: there's no default constant epsilon (delta) to compare two double values. Instead, methods select "last affected" decimal place, based on the exponent of the specified value. See method: getLastAffectedDecimal(double)

Field Summary

Fields

Modifier and Type	Field and Description
static int	MAXIMUM_DEFAULT_DECIMALS Maximum value that can be automatically selected for round(double) method.
static int	MAXIMUM_POSSIBLE_DECIMALS Maximum absolute value that can be used in the roundTo(double, int) method.

Method Summary

Modifier and Type	Method and Description
static java.lang.Double	asDouble(java.lang.Number x) If specified number is null - returns null.
static int	compare(double a, double b)
static int	compare(double a, double b, double accuracy)
static boolean	equals(java.lang.Double a, java.lang.Double b)
static boolean	equals(java.lang.Double a, java.lang.Double b, double accuracy)
static int	getLastAffectedDecimal(double d) Method returns dynamically calculated "index" (deximal position) of the digit that is considered as last for rounding or comparison.
static boolean	isFinite(java.lang.Double d)
static boolean	isNegative(java.lang.Double d)
static boolean	isNumber(java.lang.Double d)
static boolean	isPositive(java.lang.Double d)
static boolean	isZero(java.lang.Double d)
static double	round(double a)
static double	roundTo(double a, int decimals)

Methods inherited from class java.lang.Object

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

Field Detail

MAXIMUM_DEFAULT_DECIMALS

public static final int MAXIMUM_DEFAULT_DECIMALS

Maximum value that can be automatically selected for round(double) method. Also maximum decimal position to be used for compare(double, double) method.

See Also:

Constant Field Values

MAXIMUM_POSSIBLE_DECIMALS

public static final int MAXIMUM_POSSIBLE_DECIMALS

Maximum absolute value that can be used in the roundTo(double, int) method.

See Also:

Constant Field Values

Method Detail

asDouble

public static java.lang.Double asDouble(java.lang.Number x)

If specified number is null - returns null.

If specified number is instance of Double - the same instance is returned.

In any other case - x.doubleValue() is returned.

isZero

public static boolean isZero(java.lang.Double d)

isPositive

public static boolean isPositive(java.lang.Double d)

Returns:

true if specified number is not null, not Double.NaN, and greater than zero.

isNegative

public static boolean isNegative(java.lang.Double d)

Returns:

true if specified number is not null, not Double.NaN, and lower than zero.

isFinite

public static boolean isFinite(java.lang.Double d)

Returns:

true if specified number is not null, not Double.NaN, not Double.POSITIVE_INFINITY, and not Double.NEGATIVE_INFINITY

isNumber

public static boolean isNumber(java.lang.Double d)

Returns:

true if specified number is not null and not Double.NaN

equals

public static boolean equals(java.lang.Double a,
                             java.lang.Double b)

equals

public static boolean equals(java.lang.Double a,
                             java.lang.Double b,
                             double accuracy)

compare

public static int compare(double a,
                          double b)

compare

public static int compare(double a,
                          double b,
                          double accuracy)

round

public static double round(double a)

getLastAffectedDecimal

public static int getLastAffectedDecimal(double d)

Method returns dynamically calculated "index" (deximal position) of the digit that is considered as last for rounding or comparison. Used to dinamically calculate "accuracy" for the equality, comparison, or rounding operations.

Note again: there's no default constant epsilon (delta) to compare two double values. Instead, methods select "last affected" decimal place, based on the exponent of the specified value. Basic formula is:

 (14 + 1) - number of digits before decimal point
 

Note that zero before decimal point is treated as one digit. For example:

• For the number 0.01 selected decimal place will be: 14
• For the number 0.1 selected decimal place will be: 14
• For the number 1.0 selected decimal place will be: 14
• For the number 10.0 selected decimal place will be: 14 - 1
• For the number 10 000.0 selected decimal place will be: 14 - 4

Therefore the bigger the integer part of a number - the lower its "accuracy". Just the way doubles are.
But it doesn't work the other way around - selected decimal place never will be greater than MAXIMUM_DEFAULT_DECIMALS.
Basically for any x where |x| < 10.0 - maximum possible accuracy will be used.
For any double with more than 14 digits before decimal point - 0 is returned.

roundTo

public static double roundTo(double a,
                             int decimals)

See Also:

MAXIMUM_POSSIBLE_DECIMALS