floating point precision giving wrong output. please help in bypassing such errors

Question

So i have got a problem with floating point variables. can someone explain? the code is given below...

#include<stdio.h>

void main() {
    float b=3.6;
    if(b==3.6)
    {
        printf("one");
    }
    if(b<=3.6)
    {
        printf("two");
    }
    if(b>=3.6)
    {
        printf("three");
    }
    else printf("four");
}

Now why is the Output "two" and "four"? and how do i avoid such things in pprograms?

See http://blog.frama-c.com/index.php?post/2011/11/08/Floating-point-quiz — Pascal Cuoq, May 17 '14 at 06:16
possible duplicate of [Floating point comparison](http://stackoverflow.com/questions/7011184/floating-point-comparison) — Pascal Cuoq, May 17 '14 at 06:18
there are already lots of duplicates about "floating point precision" here — phuclv, May 17 '14 at 07:27
also, avoid float unless needed, most probably in storage, normal temporary calculations should use double instead — phuclv, May 17 '14 at 07:28

score 0 · Accepted Answer · answered May 17 '14 at 06:23

0

You are comparing floats to doubles. Try comparing them to 3.6f instead of 3.6

answered May 17 '14 at 06:23

Jcl

27,696
5
61
92

by default does it take 3.6 as double when comparing? – Sam May 17 '14 at 06:26
You might want to check this: [What's the point of the L, U and f suffixes on numeric literals?](http://www.parashift.com/c++-faq-lite/numeric-literal-suffixes.html) – Jcl May 17 '14 at 06:29
And answering to your comment... yes, every unsuffixed literal with decimals is taken as a `double`, so it's actually promoting `b` to `double`, and making the comparison afterwards – Jcl May 17 '14 at 06:31

floating point precision giving wrong output. please help in bypassing such errors

1 Answers1