What does less than 0. (0 dot) mean?

Question

I just saw this for the first time. The source code I'm looking at is in C

 if( rate < 0.){
   ...
 }
 else{
   ...
 }

What happens if rate=0 ?

0 is not less than 0, so if rate = 0 it would result in false. — ChrisCM, Aug 20 '13 at 20:40
check this: http://codepad.org/GhfrQ9YO What I mean to say in first comment — Grijesh Chauhan, Aug 20 '13 at 20:49
If `rate = 0` then `rate` will be assigned the value of `0`. :) — lurker, Aug 20 '13 at 22:20
@GrijeshChauhan: The size of `0.` vs. `0.0` doesn't really make the relevant point, which is that `0.0`, `.0`, and `0.` are all valid syntax for a floating-point constant of type `double` with the value zero. — Keith Thompson, Aug 20 '13 at 23:28
@KeithThompson Hmm got it! I was not aware of this that all three are valid syntax. Thanks! — Grijesh Chauhan, Aug 21 '13 at 03:43
This was marked as a duplicate of two different questions. It's not a duplicate of the first one, "C++ difference between 0 and 0.0". It is a duplicate of the second, "C++ difference between 0. and 0.0". — Keith Thompson, Aug 21 '13 at 06:48

score 16 · Accepted Answer · answered Aug 20 '13 at 20:41

16

0. is a literal of type double (and value zero). By contrast, 0 is a literal of type int.

answered Aug 20 '13 at 20:41

Kerrek SB

447,451
88
851
1,056

3

BTW: `0` is an octal-constant. ;-) Like _that_ makes any difference.(C11 6.4.4.1 1) – chux - Reinstate Monica Aug 20 '13 at 22:15
3

@chux: Well of course - who'd think anything else? :-) – Kerrek SB Aug 20 '13 at 23:04

Jiminion · Answer 2 · 2013-08-21T14:17:16.273

6

It is interpreting 0. as a double (0.0) instead of an integer (0).

Check the link: of "working code", showing the different sizes of various types of zero constants:

edited Aug 21 '13 at 14:17

answered Aug 20 '13 at 20:38

Jiminion

5,184
1
25
52

6

It certainly isn't: if it's doing anything it's making it a double rather than an int. – Phil Willoughby Aug 20 '13 at 20:40
3

Yes, @Phil is correct. You need 0.0f to get a float value, double is the default. – ChrisCM Aug 20 '13 at 20:41
Fixed. (Thanks for noting that.) – Jiminion Aug 20 '13 at 20:43
@Jim add this link of [codepade](http://codepad.org/GhfrQ9YO) – Grijesh Chauhan Aug 20 '13 at 20:48
@GrijeshChauhan good link, but should have sizeof(0) as well, no? – Jiminion Aug 20 '13 at 20:51
`foo = 1; if (foo == 1.0)…` works perfectly well; there is no floating-point error, and the test evaluates to true. – Eric Postpischil Aug 20 '13 at 20:51
@EricPostpischil I wasn't necessarily referring to that particular comparision, but comparing ints with floating-point in general. Most of the time they are fine, but sometimes not. – Jiminion Aug 20 '13 at 20:55
@Jim remove that its confusing, and not needed – Grijesh Chauhan Aug 20 '13 at 20:57
@GrijeshChauhan can you edit the codepad? I don't have an account there. – Jiminion Aug 20 '13 at 20:58
@Jim I have edited and updated link in your answer already check. I suggest you to post that link with a small explanation. – Grijesh Chauhan Aug 20 '13 at 20:59
@StephenCanon Fixed. Removed cast reference. – Jiminion Aug 21 '13 at 14:17
@Jim: Thanks for fixing that! – Stephen Canon Aug 21 '13 at 15:18

Shafik Yaghmour · Answer 3 · 2013-08-21T09:22:19.173

0. is a floating constant and since it does not have a suffix it is a double, from the draft C99 standard section 6.4.4.2 Floating constants we have the following grammar:

floating-constant:
   decimal-floating-constant
   hexadecimal-floating-constant
decimal-floating-constant:
   fractional-constant exponent-part_opt floating-suffix_opt
   digit-sequence exponent-part floating-suffix_opt
[...]
fractional-constant:
    digit-sequence_opt . digit-sequence
    digit-sequence .                                   &lt ---- This covers 0.
[...]

We then have in paragraph 4:

An unsuffixed floating constant has type double. If suffixed by the letter f or F, it has type float. If suffixed by the letter l or L, it has type long double.

What does less than 0. (0 dot) mean?

3 Answers3