float numbers

[Cagliostro]

But the way, what is the binary format of float/bouble numbers? For example run this code you to understand me:
#include<iostream>
using namespace std;
struct x
{
union
{
int i;
float f;
};
};
int main(int _args,char** _cmdline)
{
x a;
a.f=1.;
cout<<a.i<<endl;
a.i=1;
cout<<a.f<<endl;
return 0;
} John Fill

ivfmd@mail.md

 

[Zyrenthian]

I was trying to figure this out a while ago as well... if you should figure it out please post.

Matt

 

[Cagliostro]

ok, I think is a question for asm programmers. John Fill

ivfmd@mail.md

 

[StukA]

Unfortunately, I don't have the book with me here, but the general binary format for floating point numbers is defined by the IEEE. It involves a sign bit, several bits for the exponent, and the rest for the mantissa. You can probably find it online, as I believe that VC++ uses the same format that the Pentium processor does.

 

[Zyrenthian]

Most likely... and I am not very good at it. I did copy my idea and merged it with your post... here is what I was thinking along the lines of

#include <iostream.h>

struct x
{
union
{
int i;
float f;
};
};
int main(int _args,char** _cmdline)
{
x a;
a.f=1.;
cout<<a.i<<endl;
cout<<&quot;BINARY: &quot;;
for(int i = 31;i+1;i--)
{
cout<< ((a.i & 1<<i) ? 1:0);
}
cout<<endl<<endl;
a.i=1;
cout<<a.f<<endl;
cout<<&quot;BINARY: &quot;;
for(int j = 31;j+1;j--)
{
cout<< ((a.i & 1<<j) ? 1:0);
}
cout<<endl;
return 0;
}

 

[KeithBrautigam]

A good place to start reading about floating point formats is

http://developer.intel.com/design/pentium/manuals/

From this page, the link you want is to the Intel Architecture Software Developer's Manual, Volume 1: Basic Architecture.

Then go to chapter 7 of this manual.

 

[copton]

hi there
as far as i know float values are defined like:
byte 0: exponent (e)
byte 1-5: mantissa (m)
as the msb is used as sign bit, it is always regarded as set for the mantisse. therefore 1/2 <= m < 1 is always valid.
signing for the exponent is realised by offsetting. that means -128 is 0x00.
the float value is m * 2[sup]e[/sup]

hope it helps
yours
copton

 

[Cagliostro]

4 bites float
Bit Description
31 - Sign
30-23 - Power
22-0 - Mantissa (real bites of the number)

8 bites float (double)
63 - Sign
62-52 - Power
51-0 - Mantissa

10 bites float (long double)
79 - Sign
78-64 - Power
63-0 - Mantissa John Fill

ivfmd@mail.md