| Class 1, Part 2 |
Intro to Java 30-IT-396 |
|
Floating Point Types
-
Numbers with fractional (decimal) values.
-
Two types:
| Primitive Type |
Size |
Range |
| float |
4 bytes |
+/- 3.40282347E + 38F |
| double |
8 bytes |
+/- 1.79769313486231570E+308 |
-
float is probably the most common.
-
double is sometimes called double-precision.
-
Consider yourself warned:
-
These two types use floating point arithmetic (FPA). FPA values speed
and efficiency over 100% accuracy. Thus, occasionally the results
of operations may be off by a fraction. This isn't a big deal for
some applications, but when dealing with currency, it can creep up and
cause a headache. I speak from experience!
-
If you require 100% accuracy, use the BigDecimal object. It is slower,
but it is much more accurate.
-
There are three special types:
-
NaN: not a number. Where have we seen this before? You can
test with Double.isNaN();
-
POSITIVE_INFINITY: a positive number divided by 0.
-
NEGATIVE_INFINITY: a negative number divided by 0.
char, boolean
Created by: Brandan
Jones January 4, 2002
