Float/Double Substraction Bug ?

[cybertof]

Hello,

I have the strange beeing :

float f1 = 108.88F;
float f2 = 111.11F;
float f3 = f1 - f2;
double d1 = 108.88;
double d2 = 111.11;
double d3 = d1 - d2;

f3 contains -2.23000336
d3 contains -2.230000000000004

instead of -2.23.
Am i missing something ?
Regards,
Cybertof.

[Miha Markic]

Hi cybertor,

This is normal behavious of binary based floating point variables.
They are only rounded to the value you want.
You might try using decimal variables instead.

--
Miha Markic - RightHand .NET consulting & development
miha at rthand com

"cybertof" <cy****************@gmx.net> wrote in message
news:MP***********************@msnews.microsoft.co m...

Hello,

I have the strange beeing :

float f1 = 108.88F;
float f2 = 111.11F;
float f3 = f1 - f2;
double d1 = 108.88;
double d2 = 111.11;
double d3 = d1 - d2;

f3 contains -2.23000336
d3 contains -2.230000000000004

instead of -2.23.
Am i missing something ?
Regards,
Cybertof.

[cybertof]

Hi Miha !

How do you explain ?
Does it mean that with single & double, you always get calculation
errors ?

When can you use binary based floating point variables if you cannot
make correct calculations with them ?
Regards,
Cybertof.

In article <O1**************@TK2MSFTNGP11.phx.gbl>, "Miha Markic" <miha
at rthand com> says...

Hi cybertor,

This is normal behavious of binary based floating point variables.
They are only rounded to the value you want.
You might try using decimal variables instead.

[Eric Gunnerson [MS]]

This document may help explain it:

http://portal.acm.org/citation.cfm?i...TOKEN=87893990

--
Eric Gunnerson

Visit the C# product team at http://www.csharp.net
Eric's blog is at http://blogs.gotdotnet.com/ericgu/

This posting is provided "AS IS" with no warranties, and confers no rights.
"cybertof" <cy****************@gmx.net> wrote in message
news:MP************************@msnews.microsoft.c om...

Hi Miha !

How do you explain ?
Does it mean that with single & double, you always get calculation
errors ?

When can you use binary based floating point variables if you cannot
make correct calculations with them ?
Regards,
Cybertof.

In article <O1**************@TK2MSFTNGP11.phx.gbl>, "Miha Markic" <miha
at rthand com> says...

Hi cybertor,

This is normal behavious of binary based floating point variables.
They are only rounded to the value you want.
You might try using decimal variables instead.

[Jon Skeet [C# MVP]]

cybertof <cy****************@gmx.net> wrote:

How do you explain ?
Does it mean that with single & double, you always get calculation
errors ?
Often, yes - but not always.
When can you use binary based floating point variables if you cannot
make correct calculations with them ?

In fact, the calculations may well be performing without any error, but
with the initial values not being what you think they were.

See http://www.pobox.com/~skeet/csharp/floatingpoint.html for more
information.

--
Jon Skeet - <sk***@pobox.com>
http://www.pobox.com/~skeet
If replying to the group, please do not mail me too