double or float

Question

Possible Duplicate:
Should I use double or float ?

When would I rather use double and when should I use float?

score 2 · Answer 1 · answered Feb 16 '11 at 13:43

2

The only time you need to use float is when you are storing large arrays of numbers. There is generally little difference in speed between the two and natively most things are double anyway.

answered Feb 16 '11 at 13:43

Gabe

79,868
10
131
226

2

The speed difference is extremely dependent on the HW you are using. – Šimon Tóth Feb 16 '11 at 13:45

score 2 · Accepted Answer · answered Feb 16 '11 at 13:43

2

It's all about precision.

If you need to store very precise numbers then use a double.

If you need to store less precise numbers and are worried about the size of memory you're using then use a float.

answered Feb 16 '11 at 13:43

Nick

22,728
6
44
79

score 2 · Answer 3 · answered Feb 16 '11 at 13:45

2

Use double when you require the range it supports. Refer to Range of floating-point numbers. You should also typically use the native type, so if you're doing graphics or GPU programming, probably better stick to floats.

But whatever you do, please, do not use either to represent currency or money.

answered Feb 16 '11 at 13:45

yan

19,660
3
33
47

1

@lital: Because it cannot represent most decimal fractions. Read http://floating-point-gui.de/ to understand why and what that means. – Michael Borgwardt Feb 16 '11 at 13:51
1

Unless your currency has coins (and prices) for base-2 fractions. :) – Matteo Italia Feb 16 '11 at 13:55
even then it isn't necessarily appropriate you may need to track cents at the same time as tracking trillions of dollars – jk. Feb 16 '11 at 16:25

double or float

3 Answers3