Module ieee754java

Package at.syntaxerror.ieee754.rounding

Class Rounding

java.lang.Object

at.syntaxerror.ieee754.rounding.Rounding

public abstract class Rounding
extends Object

This class is used for rounding IEEE 754 floating-point numbers.

There are 5 predefined rounding modes:

• round to nearest, ties to even
• round to nearest, ties away from zero
• round toward 0
• round toward +∞
• round toward -∞

DEFAULT_ROUNDING controls the default rounding mode used by FloatingCodec.encode(Floating) and its descendents.

Author:

Thomas Kasper

Field Summary

Fields

Modifier and Type	Field	Description
static @NonNull Rounding	DEFAULT_ROUNDING	The rounding method used by FloatingCodec.encode(at.syntaxerror.ieee754.Floating) and its descendents.
static final Rounding	TIES_AWAY	Round to nearest, ties away from zero
static final Rounding	TIES_EVEN	Round to nearest, ties to even.
static final Rounding	TOWARD_NEGATIVE	Round toward -∞
static final Rounding	TOWARD_POSITIVE	Round toward +∞
static final Rounding	TOWARD_ZERO	Round toward 0

Constructor Summary

Constructors

Constructor	Description
Rounding()

Method Summary

Modifier and Type	Method	Description
abstract boolean	roundBinary(boolean sign, boolean guard, boolean round, boolean sticky)	Rounds a binary floating-point number.
abstract BigDecimal	roundDecimal(BigDecimal value)	Rounds the BigDecimal (to 0 decimal places) according to the rounding mode's rules

Methods inherited from class java.lang.Object

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

Field Details

TIES_EVEN

public static final Rounding TIES_EVEN

Round to nearest, ties to even.

Rounds to the nearest value. If the number falls midway, it is rounded to the nearest even value.

TIES_AWAY

public static final Rounding TIES_AWAY

Round to nearest, ties away from zero

Rounds to the nearest value. If the number falls midway, it is rounded to the nearest value above (positive numbers) or below (negative numbers).

TOWARD_ZERO

public static final Rounding TOWARD_ZERO

Round toward 0

Rounds toward zero. Effectively discards any decimal places.

TOWARD_POSITIVE

public static final Rounding TOWARD_POSITIVE

Round toward +∞

Rounds toward positive infinity, also known as rounding up or ceiling.

TOWARD_NEGATIVE

public static final Rounding TOWARD_NEGATIVE

Round toward -∞

Rounds toward negative infinity, also known as rounding down or floor.

DEFAULT_ROUNDING

@NonNull
public static @NonNull Rounding DEFAULT_ROUNDING

The rounding method used by FloatingCodec.encode(at.syntaxerror.ieee754.Floating) and its descendents.

Constructor Details

Rounding

public Rounding()

Method Details

roundBinary

public abstract boolean roundBinary(boolean sign,
 boolean guard,
 boolean round,
 boolean sticky)

Rounds a binary floating-point number. The method returns true when rounding up and false when rounding down.

The sign bit specifies whether the number is negative. The guard bit is the least significant bit than can be stored (the ULP). The round bit is the bit following the guard bit. The sticky bit is the bitwise OR of all other following bits.

Parameters:

sign - the sign bit

guard - the guard bit

round - the round bit

sticky - the sticky bit

Returns:

whether the number is rounded up

roundDecimal

public abstract BigDecimal roundDecimal(BigDecimal value)

Rounds the BigDecimal (to 0 decimal places) according to the rounding mode's rules

Parameters:

value - the input value

Returns:

the rounded value