Float print gives extra numbers

Question

i am having a hard time printing float variable. It gives me extra numbers, that shouldn´t be there. Here is example:

float number;
char temp[50];

fgets ( temp, sizeof temp, fr );  //reading string from file, example: 99.10   
number=atof(temp);

printf("%lf",number);             //console output: 99.101563

This is obviously wrong output. Any suggestions?

http://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html — Carl Norum, Apr 12 '15 at 22:19
It's obviously correct, according to the spec. Floating point is not an exact representation. — Matthew Read, Apr 12 '15 at 22:19
"This is obviously wrong output" -- nope, this is actually correct output. — Paul R, Apr 12 '15 at 22:22
@MatthewRead This is as useful a characterization as “Integer is not an exact representation”. Binary floating-point is exact for values that lend themselves to that representation in the same way that integer types are. — Pascal Cuoq, Apr 12 '15 at 22:36
@PascalCuoq Integers represent their *range* exactly; floats do not. It's not useful to pretend that is not the relevant point of confusion. — Matthew Read, Apr 12 '15 at 22:38
@MatthewRead Integers do not represent 1.10 any better than floats do although it is in range for both, and this is completely relevant to the question. — Pascal Cuoq, Apr 12 '15 at 22:44
@PascalCuoq No, 1.10 is not in the range of integers. It doesn't exist. — Matthew Read, Apr 12 '15 at 22:46

score 1 · Answer 1 · answered Apr 12 '15 at 22:20

1

There is nothing unexpected about this output. 99.10 can't be represented exactly in IEEE754. Change your format to restrict output to two decimal places if that's what you want:

printf("%.2f", number);

answered Apr 12 '15 at 22:20

Carl Norum

219,201
40
422
469

Float print gives extra numbers

1 Answers1