com.google.common.math

Class DoubleMath

java.lang.Object

com.google.common.math.DoubleMath

public final class DoubleMath
extends Object

A class for arithmetic on doubles that is not covered by Math.

Since:

11.0

Author:

Louis Wasserman

Method Summary

Methods

Modifier and Type	Method and Description
static double	factorial(int n) Returns n!, that is, the product of the first n positive integers, 1 if n == 0, or e n!}, or Double.POSITIVE_INFINITY if n! > Double.MAX_VALUE.
static int	fuzzyCompare(double a, double b, double tolerance) Compares a and b "fuzzily," with a tolerance for nearly-equal values.
static boolean	fuzzyEquals(double a, double b, double tolerance) Returns true if a and b are within tolerance of each other.
static boolean	isMathematicalInteger(double x) Returns true if x represents a mathematical integer.
static boolean	isPowerOfTwo(double x) Returns true if x is exactly equal to 2^k for some finite integer k.
static double	log2(double x) Returns the base 2 logarithm of a double value.
static int	log2(double x, RoundingMode mode) Returns the base 2 logarithm of a double value, rounded with the specified rounding mode to an int.
static double	mean(double... values) Returns the arithmetic mean of the values.
static double	mean(int... values) Returns the arithmetic mean of the values.
static double	mean(Iterable<? extends Number> values) Returns the arithmetic mean of the values.
static double	mean(Iterator<? extends Number> values) Returns the arithmetic mean of the values.
static double	mean(long... values) Returns the arithmetic mean of the values.
static BigInteger	roundToBigInteger(double x, RoundingMode mode) Returns the BigInteger value that is equal to x rounded with the specified rounding mode, if possible.
static int	roundToInt(double x, RoundingMode mode) Returns the int value that is equal to x rounded with the specified rounding mode, if possible.
static long	roundToLong(double x, RoundingMode mode) Returns the long value that is equal to x rounded with the specified rounding mode, if possible.

Methods inherited from class java.lang.Object

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

Method Detail

roundToInt

public static int roundToInt(double x,
             RoundingMode mode)

Returns the int value that is equal to x rounded with the specified rounding mode, if possible.

Throws:

ArithmeticException - if

• x is infinite or NaN
• x, after being rounded to a mathematical integer using the specified rounding mode, is either less than Integer.MIN_VALUE or greater than Integer.MAX_VALUE
• x is not a mathematical integer and mode is RoundingMode.UNNECESSARY

roundToLong

public static long roundToLong(double x,
               RoundingMode mode)

Returns the long value that is equal to x rounded with the specified rounding mode, if possible.

Throws:

ArithmeticException - if

• x is infinite or NaN
• x, after being rounded to a mathematical integer using the specified rounding mode, is either less than Long.MIN_VALUE or greater than Long.MAX_VALUE
• x is not a mathematical integer and mode is RoundingMode.UNNECESSARY

roundToBigInteger

public static BigInteger roundToBigInteger(double x,
                           RoundingMode mode)

Returns the BigInteger value that is equal to x rounded with the specified rounding mode, if possible.

Throws:

ArithmeticException - if

• x is infinite or NaN
• x is not a mathematical integer and mode is RoundingMode.UNNECESSARY

isPowerOfTwo

public static boolean isPowerOfTwo(double x)

Returns true if x is exactly equal to 2^k for some finite integer k.

log2

public static double log2(double x)

Returns the base 2 logarithm of a double value.

Special cases:

• If x is NaN or less than zero, the result is NaN.
• If x is positive infinity, the result is positive infinity.
• If x is positive or negative zero, the result is negative infinity.

The computed result is within 1 ulp of the exact result.

If the result of this method will be immediately rounded to an int, log2(double, RoundingMode) is faster.

log2

public static int log2(double x,
       RoundingMode mode)

Returns the base 2 logarithm of a double value, rounded with the specified rounding mode to an int.

Regardless of the rounding mode, this is faster than (int) log2(x).

Throws:

IllegalArgumentException - if x <= 0.0, x is NaN, or x is infinite

isMathematicalInteger

public static boolean isMathematicalInteger(double x)

Returns true if x represents a mathematical integer.

This is equivalent to, but not necessarily implemented as, the expression !Double.isNaN(x) && !Double.isInfinite(x) && x == Math.rint(x).

factorial

public static double factorial(int n)

Returns n!, that is, the product of the first n positive integers, 1 if n == 0, or e n!}, or Double.POSITIVE_INFINITY if n! > Double.MAX_VALUE.

The result is within 1 ulp of the true value.

Throws:

IllegalArgumentException - if n < 0

fuzzyEquals

public static boolean fuzzyEquals(double a,
                  double b,
                  double tolerance)

Returns true if a and b are within tolerance of each other.

Technically speaking, this is equivalent to Math.abs(a - b) <= tolerance || Double.valueOf(a).equals(Double.valueOf(b)).

Notable special cases include:

• All NaNs are fuzzily equal.
• If a == b, then a and b are always fuzzily equal.
• Positive and negative zero are always fuzzily equal.
• If tolerance is zero, and neither a nor b is NaN, then a and b are fuzzily equal if and only if a == b.
• With Double.POSITIVE_INFINITY tolerance, all non-NaN values are fuzzily equal.
• With finite tolerance, Double.POSITIVE_INFINITY and Double.NEGATIVE_INFINITY are fuzzily equal only to themselves.

This is reflexive and symmetric, but not transitive, so it is not an equivalence relation and not suitable for use in Object.equals(java.lang.Object) implementations.

Throws:

IllegalArgumentException - if tolerance is < 0 or NaN

Since:

13.0

fuzzyCompare

public static int fuzzyCompare(double a,
               double b,
               double tolerance)

Compares a and b "fuzzily," with a tolerance for nearly-equal values.

This method is equivalent to fuzzyEquals(a, b, tolerance) ? 0 : Double.compare(a, b). In particular, like Double.compare(double, double), it treats all NaN values as equal and greater than all other values (including Double.POSITIVE_INFINITY).

This is not a total ordering and is not suitable for use in Comparable.compareTo(T) implementations. In particular, it is not transitive.

Throws:

IllegalArgumentException - if tolerance is < 0 or NaN

Since:

13.0

mean

public static double mean(double... values)

Returns the arithmetic mean of the values. There must be at least one value, and they must all be finite.

mean

public static double mean(int... values)

Returns the arithmetic mean of the values. There must be at least one value. The values will be converted to doubles, which does not cause any loss of precision for ints.

mean

public static double mean(long... values)

Returns the arithmetic mean of the values. There must be at least one value. The values will be converted to doubles, which causes loss of precision for longs of magnitude over 2^53 (slightly over 9e15).

mean

public static double mean(Iterable<? extends Number> values)

Returns the arithmetic mean of the values. There must be at least one value, and they must all be finite. The values will be converted to doubles, which may cause loss of precision for some numeric types.

mean

public static double mean(Iterator<? extends Number> values)

Returns the arithmetic mean of the values. There must be at least one value, and they must all be finite. The values will be converted to doubles, which may cause loss of precision for some numeric types.