double calculation

[Ronnie Ho]

Hi all,

Why does

 double d = 1.00 -0.42;
System.out.println(d); 

gives 0.5800000000000001 ? where does the .0000...001 come from ? Thanks.
[ October 17, 2005: Message edited by: Ronnie Ho ]

[Grant Gainey]

The short answer is, computers store their data in binary, and "5.8" is not precisely representable.

The slightly-longer answer is that Java uses IEEE-754 to store/represent floating-point numbers, and the number you're seeing is the closest one can come to 5.8 in that format.

The best answer is that it's important to understand that a programming language's "float" and "double" are not the same as the real number line. Check out this link for a great explanation:

What Every Computer Scientist Should Know About Floating Point

Grant

[Edited to fix broken tags]
[ October 19, 2005: Message edited by: G Gainey ]

[Ronnie Ho]

Thanks. I'll just take the short answer. The link contains too much details.

[Stan James]

I always jump in when I see two decimal points. If you're dealing with money, floating types will probably make you very unhappy. You've already started to see why. The PowerBall is at $340 million so I'm thinking about money again today.

[Jesper de Jong]

So if you're dealing with amounts of money in your application, use java.math.BigDecimal instead of float or double.