0.1 + 0.7 != 0.8

Question

Why in some programming languages the expression in title evaluates to true? I've tried it in php, ruby and python.

http://en.wikipedia.org/wiki/Floating_point#Accuracy_problems — Ignacio Vazquez-Abrams, Dec 31 '10 at 16:21
possible duplicate of [Is JavaScript's Math broken?](http://stackoverflow.com/questions/588004/is-javascripts-math-broken) And about a hundred other questions. — dan04, Dec 31 '10 at 16:23
The accepted answer to that one is marvelous. http://stackoverflow.com/questions/588004/is-javascripts-math-broken/588014#588014 — Paul Tomblin, Dec 31 '10 at 16:25
It evaluates to true because it is true (when 0.1, etc are floating-point literals). — Stephen Canon, Dec 31 '10 at 16:36

score 18 · Accepted Answer · answered Dec 31 '10 at 16:22

18

Please read What Every Programmer Should Know About Floating-Point Arithmetic .

answered Dec 31 '10 at 16:22

ismail

46,010
9
86
95

score 2 · Answer 2 · answered Dec 31 '10 at 16:27

2

double TOLERANCE < 1.0E-10;
if(fabs(0.1+0.7-0.8)< TOLERANCE)
{
    std::cout << "0.1 + 0.7 == 0.8" << std::endl;
}
else
{
    std::cout << "0.1 + 0.7 != 0.8" << std::endl;
}

answered Dec 31 '10 at 16:27

Cesar A. Rivas

1,355
1
10
13

score 0 · Answer 3 · answered Dec 31 '10 at 16:25

0

Because of internal interpretation of floating-point numbers, they do not exactly equal to what you have wrote.

answered Dec 31 '10 at 16:25

Nickolay Olshevsky

13,706
1
34
48

0.1 + 0.7 != 0.8

3 Answers3