error: invalid operands to binary % when taking modulus of float

Question

Its been a fairly long week, so forgive me if I am being thick.

I have some code like:

float someFloat = 0;
//....do some stuff to someFloat
//....

if( someFloat % 1)
{
   //take some action
}

I get a compiler error : error: invalid operands to binary %

Assuming the compiler isnt on drugs, whats wrong with this?

EDIT: As an aside, what I actually wanted to do was detect non-integer value and round up. What I should have been doing was calling roundf (and I guess checking if the return is less than the operand and then incrementing if so, to take care that we have rounded up)

Answer is already there. For some more question like _Why not working on float_? please have a look on [why-does-modulus-division-only-work-with-integers?](http://stackoverflow.com/questions/6102948/why-does-modulus-division-only-work-with-integers) — Dayal rai, Oct 11 '13 at 11:45
To round up if there is **any** non-zero fraction part, use `ceilf`, not `roundf`. Note: `ceilf` rounds toward infinity; if you want negative numbers rounded toward —infinity (e.g., —3.25 to —4), you will need additional code. — Eric Postpischil, Oct 11 '13 at 11:50
@EricPostpischil thats almost exactly what I want. I dont have it on my target though, so i guess its the double version `ceil` for me... — mjs, Oct 11 '13 at 12:28
Look in your C manual under `%`. It will say what the valid operands are. See if `float` is listed as a valid operand. (Spoiler alert: It's not.) — Raymond Chen, Oct 11 '13 at 13:45

score 20 · Accepted Answer · answered Oct 11 '13 at 11:42

20

% is an integer operator - use fmod or fmodf for doubles or floats.

Alternatively if you expect your float to represent integer values then convert it to an int first, e.g.:

if ((int)someFloat % 2 == 1) // if f is an odd integer value
{
    ...
}

answered Oct 11 '13 at 11:42

Paul R

208,748
37
389
560

1

Huh, I never knew that. Only been doing this 15 years.... Just had a look in my trusty old K&R, and that is so. – mjs Oct 11 '13 at 11:46
1

30 years for me, and still learning... ;-) – Paul R Oct 11 '13 at 11:47
1

@PaulR as with practically any other science, we can never completely learn a language. :) – Oct 11 '13 at 11:48
2

True - of course after 30 years you begin to suspect that there are some things that you once knew, then forgot, then had to learn again... ;-) – Paul R Oct 11 '13 at 11:49
1

@Paul R thanks for fmod hint – Chris P Jan 23 '19 at 18:18

score 3 · Answer 2 · answered Oct 11 '13 at 11:42

3

The modulus operator only works on integers. For floating point values use fmod or fmodf.

answered Oct 11 '13 at 11:42

Jon

428,835
81
738
806

1

@downvoter (we both know who you are): Hey there! That was really mature. – Jon Oct 11 '13 at 11:54

score 2 · Answer 3 · answered Oct 11 '13 at 11:46

% this only works with integers use fmod for floating point or double values**

double fmod(double x, double y)


x -- This is the floating point value with the division numerator i.e. x.

y -- This is the floating point value with the division denominator i.e. y.

error: invalid operands to binary % when taking modulus of float

3 Answers3

Linked

Related