Fractional part from a float value in C++ without any inbuilt libraries

Question

float f = 735.64;    
int i = (int)f;    
float f2 = f - i;    
cout<<f2<<endl;

The value outputted by f2 is simply not 0.64 but something like 0.6400015 I just want to have the value in f2 as either 0.64 or 0.640000.. so could I do it without using any inbuilt functions as such?

Thanks.

[What Every Computer Scientist Should Know About Floating-Point Arithmetic](http://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html) — Cory Kramer, Oct 02 '14 at 16:35
@MarkRansom: Indeed; but you'll never get all the way there with a binary representation. — Mike Seymour, Oct 02 '14 at 17:01
Don't use floating point but Fixed Point. Search the web for "Fixed point notation". — Thomas Matthews, Oct 02 '14 at 18:02

score 1 · Answer 1 · answered Oct 02 '14 at 16:59

you have to understand that there is an error in each number used in a computer, if you find the value of that error and subtracted from the solution, then you will get the actual values. I advise you to study all about "relative error" and "absolute error" these concepts are related to numerical mathematics

Fractional part from a float value in C++ without any inbuilt libraries

1 Answers1