why is 1.2 * 30 = 35?

问题

Why does this:

int main(void)
{
    short w = 30;
    return  1.2 * w;
}

return 35?

回答1:

If you want to get more suitable result, try the following:

return 12*w/10

回答2:

1.2 * w is 36.0. It has the double type meaning it is not represented exactly.

Likely it turns out to be slightly less than 36, maybe 35.99999 so when you return it the fractional part is discarded and the integer part only is returned. That's how you get 35.

---

P.S. All operations with floating point are not precise. You should expect little discrepancies. Also when you compare a floating point value against a fixed value, you mustn't do a direct comparison but do a range comparison.

Wrong: if (value == 36.0) { /* ... */ }

Correct: if (abs (value - 36.0) < 0.0001) { /* ... */ }

回答3:

It's an issue with binary floating point precision. 1.2 is very slightly less than 1.2, so the result of the multiplication is slightly less than 36.

回答4:

Because of representation: 1.2 is really something like 1.1999999

回答5:

Since floating point math could be inexact, use round() before casting to integer, to get a better result.

#include <math.h>
...
return (int)round(some_real_value);

回答6:

Don't expect absolutely exact result in floating point operations. Multiplication result may be, for example, 35.9999, which is rounded to 35.

回答7:

short w = 30; return 1.2 * w;

In return statement first sort value is type cast in double because 1.2 is double type so it will multiply 1.2 * 30.000000 and result become around 35.999999 and function return type is int so decimal part is truncate and it return 35 only.

来源：https://stackoverflow.com/questions/3811355/why-is-1-2-30-35