BigDecimal

BigDecimal is a class in the Java standard library, java.math.BigDecimal, used for representing and performing arithmetic on decimal numbers with arbitrary precision. It provides exact decimal representation and arithmetic, making it suitable for financial calculations where floating point errors are unacceptable.

A BigDecimal stores its value as an unscaled BigInteger and a scale. The numeric value equals unscaledValue × 10^-scale. The instance is immutable; arithmetic operations return new BigDecimal instances rather than modifying the operands.

Operations include add, subtract, multiply, and divide. Division methods may require a scale and rounding mode or a MathContext to specify precision and rounding behavior. RoundingMode and MathContext control how results are rounded. The class also supports setScale to adjust the number of decimal places. Other utility methods include compareTo, equals, toString, precision, scale, and unscaledValue.

Construction: the preferred way is to create a BigDecimal from a ASCII decimal string to avoid binary floating point issues, e.g., new BigDecimal("123.45"). Constructors from long, int, or double exist but can introduce rounding. BigDecimal.valueOf(double) is often a safe alternative because it uses a canonical string representation of the double. BigInteger-based constructors are used for advanced needs.

Performance and scope: BigDecimal provides exact arithmetic at the cost of speed and memory compared to primitive types. It is not intended for scientific computing where wide ranges and performance are critical; it does not provide native transcendental functions. It is Serializable and part of the java.math package.