Relational operator in C while using real variables

Question

Why the output is not as expected?

#include <stdio.h>
void main(){
    float a,b,c;
    b=0.7;
    if( b<0.7 )
        printf(" It should NOT be here");
    else
        printf("It Should be here");
}

Please format your code properly and give the expected output, and what you get as output as well. — akshayk07, Sep 09 '17 at 08:29

score 0 · Answer 1 · answered Sep 09 '17 at 08:35

0

0.7 is double value. Try 0.7f in code. It should work.

answered Sep 09 '17 at 08:35

Ganeshdip

389
2
10

No, seems You don't understand floating point behaviours. Is 0.69999999999999996 or 0.699999988 depends on declaration – Jacek Cz Sep 09 '17 at 08:37
b is a float; 0.7 is a value of type double. The comparison between the two requires a conversion. The compiler will convert the float value to a double value ... and the value resulting from converting a float to a double is not the same as the value resulting from the compiler converting a string of text (the source code) to a double. But don't ever compare floating point values (float, double, or long double) with ==. – Ganeshdip Sep 09 '17 at 08:41
1

@JacekCz - I tried this code with 0.7f and it works. Please try by yourself. It should work. – Ganeshdip Sep 09 '17 at 10:03
@RudyVelthuis but type should be same. Float should be compared with float. – Ganeshdip Sep 09 '17 at 18:42
@Jaczek: If the conversion is correct, the value in `b` should be equal to `0.7f`: `b = 0.7f; if (b < 0.7f)`. – Rudy Velthuis Sep 09 '17 at 18:43
@Ganeshdip: I misunderstood your comment, sorry. You answer is a little short, but correct. – Rudy Velthuis Sep 09 '17 at 18:46
@RudyVelthuis That's ok :) – Ganeshdip Sep 09 '17 at 19:54

Jacek Cz · Answer 2 · 2017-09-09T08:44:22.827

-1

Floating point numbers

have limited accuracy
some value doesn't exist. 0.7 is such value. On my platform is 0.699999988. Reason: we write numbers as decimal, but internal are realised as binary. Many longer materials exist https://en.wikipedia.org/wiki/Floating-point_arithmetic#Representable_numbers.2C_conversion_and_rounding

So behaviour You are surprised

edited Sep 09 '17 at 08:44

answered Sep 09 '17 at 08:33

Jacek Cz

1,872
1
15
22

1

FWIW, `0.7` (double) is exactly `0.6999999999999999555910790149937383830547332763671875`, while `0.7f` (float) is exactly `0.699999988079071044921875`. So the float is slightly smaller than the *double* representation of `0.7`. – Rudy Velthuis Sep 09 '17 at 18:37

Nikolaos Polygenis · Answer 3 · 2017-09-20T11:15:00.027

-1

Please Try the below Code, it works!!!:

Code

 #include <stdio.h>

    int  main(void)
    {

    float a,b,c,temp;
    temp=0.7;
    b=0.7;

    if( b<temp )
        printf(" It should NOT be here");
    else
        printf("It Should be here");


    }

edited Sep 20 '17 at 11:15

answered Sep 09 '17 at 09:34

Nikolaos Polygenis

591
7
13

1

And did You try? Probably not – Jacek Cz Sep 09 '17 at 09:37
it works!! i have run it – Nikolaos Polygenis Sep 09 '17 at 09:39
Serious? On which platform? – Jacek Cz Sep 09 '17 at 09:40
Changing assignment for initialization doesn't change a thing. This is a platform dependent result that has to do with how floating point works. You just gave the OP their own code and said "try it". – StoryTeller - Unslander Monica Sep 09 '17 at 09:41
linux and windows. – Nikolaos Polygenis Sep 09 '17 at 09:41
in linux the code of OP, give the variables the below values:a=0.0,b=0.0,c=0.7.thats why i gave this declaration:double a=0.7, b=0.7, c = 0.7;.and with float declaration with one decimal point the code doesnt work properly. – Nikolaos Polygenis Sep 09 '17 at 09:46
1

@StoryTeller Not “their own code”. Unlike the code in the question, the code in this answer prints `It Should be here` for any compiler that rightfully defines `FLT_EVAL_METHOD` to zero, although interestingly it does not have to for a compiler that rightfully defines `FLT_EVAL_METHOD` to 2. – Pascal Cuoq Sep 10 '17 at 00:37
I have changed my code!! – Nikolaos Polygenis Sep 20 '17 at 11:38
It is not a good idea to compare with a value (like the value 0.7 in our case) because you never know the machine’s approximation for a variable assignment! (in that case the approximation was 0.699999988079071044921875) – Nikolaos Polygenis Sep 20 '17 at 11:38

Relational operator in C while using real variables

3 Answers3