problem about converting "float" to "int"

[pango]

I write below code in my program:
float f=0.371f;
int i=(int)(f*1000.0f);

I think the result of "i" should be "371",but in fact it is "370",why?How to solve it?

[doug mansell]

pango wrote:

I write below code in my program:
float f=0.371f;
int i=(int)(f*1000.0f);

I think the result of "i" should be "371",but in fact it is "370",why?How to solve it?

If you look at f in your debugger you will see it is actually
0.37099999. This is the nature of floating point representation. The
other key thing to note is that float to integer conversions don't
round. If you want rounding, you write your own.
And, as an exercise in confusion, inspect this with your debugger:

int i2 = (0.371 * 1000.0);
doug.

[Ronald Laeremans [MSFT]]

Search for "what every programmer should know about floating point" using
your favorite search engine for the classic (and well written) article
describing this and other relevant issues.

Ronald Laeremans
Visual C++ team

"doug mansell" <do**********@hotmail.com> wrote in message
news:uS**************@TK2MSFTNGP12.phx.gbl...

pango wrote:

I write below code in my program:
float f=0.371f;
int i=(int)(f*1000.0f);

I think the result of "i" should be "371",but in fact it is "370",why?How
to solve it?

If you look at f in your debugger you will see it is actually 0.37099999.
This is the nature of floating point representation. The other key thing
to note is that float to integer conversions don't round. If you want
rounding, you write your own.
And, as an exercise in confusion, inspect this with your debugger:

int i2 = (0.371 * 1000.0);
doug.