> Java provides **floating-point types** to represent numbers with decimals. These are critical when working with measurements, scientific calculations, currency, and real-world quantities that aren't whole numbers.
---
### Overview of Floating-Point Types
| Type | Size | Precision | Range (Approx) | Default Value | Wrapper Class |
|---------------|----------|------------------------|---------------------------------------------|-----------------|----------------|
| `float` | 4 bytes | ~7 decimal digits | ±3.4E−38 to ±3.4E+38 | `0.0f` | `Float` |
| `double` | 8 bytes | ~15 decimal digits | ±1.7E−308 to ±1.7E+308 | `0.0d` | `Double` |
| `BigDecimal` | Varies | Arbitrary precision | Limited only by memory | `BigDecimal.ZERO` | `BigDecimal` |
---
## `float`: Lightweight Decimal Type
```java
float temperature = 36.6f; // must use 'f' suffix
- Memory-efficient (4 bytes)
- Limited precision; use cautiously in financial calculations
double: Standard for Decimals
double pi = 3.141592653589793;
- Higher precision than
float(15 decimal digits) - Default type for decimal literals
- Still susceptible to rounding errors
BigDecimal: Precise Decimal Type
import java.math.BigDecimal;
BigDecimal price = new BigDecimal("19.99");
BigDecimal total = price.multiply(new BigDecimal("2"));
- Arbitrary precision — perfect for finance, billing, tax
- Slower, more verbose; avoid primitive operators (
+,*)
Floating-Point Precision Warning
Floating-point types (float, double) may introduce rounding errors:
System.out.println(0.1 + 0.2); // Output: 0.30000000000000004
Use BigDecimal for accurate monetary operations.
Comparing Floating-Point Values
Avoid direct equality:
double a = 0.1 + 0.2;
if (Math.abs(a - 0.3) < 0.000001) {
System.out.println("Equal enough!");
}
Summary Table
| Type | Use Case | Pros | Cons |
|---|---|---|---|
float |
Graphics, games | Memory efficient | Low precision |
double |
Scientific, engineering | Good precision & speed | Not accurate for currency |
BigDecimal |
Finance, tax, e-commerce | Accurate, arbitrary size | Verbose, slower performance |
Best Practices
- Use
doublefor general-purpose decimal values. - Use
BigDecimalfor financial or high-precision needs. - Avoid
floatunless working in performance-sensitive environments like game development or embedded systems.
Example: Choosing the Right Floating Type
double distanceKm = 42.195;
BigDecimal accountBalance = new BigDecimal("1050.75");
float speed = 24.5f;