How to set a double to NAN (not a number)?
I only found a IsNAN in the math unit.
Mattias
Mailing lists / fpc-pascal / 2007-01
NAN for double
10 messages · 4 people ·
- Mattias Gaertner 5
- Daniël Mantione 2
- Marco van de Voort 2
- Florian Klaempfl 1
Reply by email Subscribe to this list Your subscription
Read-only archive. Use j and k to move between messages.
Op Thu, 25 Jan 2007, schreef Mattias Gaertner:
How to set a double to NAN (not a number)? I only found a IsNAN in the math unit.
const nan=0/0;Daniël
Show quoted text 8 lines
Op Thu, 25 Jan 2007, schreef Mattias Gaertner:
> How to set a double to NAN (not a number)?
> I only found a IsNAN in the math unit.
const nan=0/0;
unit1.pas(28,10) Error: Division by zeroMattias
Show quoted text 6 lines
> > How to set a double to NAN (not a number)?
> > I only found a IsNAN in the math unit.
>
> const nan=0/0;
unit1.pas(28,10) Error: Division by zero
Afaik that is delphi mode only. Also check the "typ" unit of numlib (in packages)
Show quoted text 8 lines
> > > How to set a double to NAN (not a number)?
> > > I only found a IsNAN in the math unit.
> >
> > const nan=0/0;
>
> unit1.pas(28,10) Error: Division by zero
Afaik that is delphi mode only.
Me delphi mode? Are you kidding?
Also check the "typ" unit of numlib (in packages)
Thanks.
const
NANDoubleArray : ARRAY[0..7] OF BYTE = ($FF,$FF,$FF,$FF,$FF,$FF,$FF,$FF);
var
nan : double absolute NANDoubleArray;Maybe NAN constants for float types can be added to the Math unit?
Mattias
Show quoted text 11 lines
On Thu, 25 Jan 2007 12:51:06 +0100 (CET)
(Marco van de Voort) wrote:
> Also check the "typ" unit of numlib (in packages)
Thanks.
const
NANDoubleArray : ARRAY[0..7] OF BYTE = ($FF,$FF,$FF,$FF,$FF,$FF,$FF,$FF);
var
nan : double absolute NANDoubleArray;
(note this was never endian checked)
On Thu, 25 Jan 2007 12:10:44 +0100 (CET) unit1.pas(28,10) Error: Division by zero Mattias
Op Thu, 25 Jan 2007, schreef Mattias Gaertner:
Show quoted text 13 lines
On Thu, 25 Jan 2007 12:10:44 +0100 (CET)
Daniël Mantione wrote:
>
>
> Op Thu, 25 Jan 2007, schreef Mattias Gaertner:
>
> > How to set a double to NAN (not a number)?
> > I only found a IsNAN in the math unit.
>
> const nan=0/0;
unit1.pas(28,10) Error: Division by zero
Well, not here. See the program below. Can you check if this depends on the compiler version or mode? 0/0 should normally work.
Daniël
cat nan.pas
program nan;
const _nan=0/0;
begin
writeln(_nan);
end.ppc386 nan
Free Pascal Compiler version 2.0.4 [2006/08/20] for i386 Copyright (c) 1993-2006 by Florian Klaempfl
Target OS: Linux for i386
Compiling nan.pas
Linking nan
7 Lines compiled, 0.0 sec
./nan
Nan
How to set a double to NAN (not a number)? I only found a IsNAN in the math unit. Mattias
How to set a double to NAN (not a number)? I only found a IsNAN in the math unit.
Then you should search more ;) Acutally there is NaN
Show quoted text 5 lines
Mattias Gaertner schrieb:
> How to set a double to NAN (not a number)?
> I only found a IsNAN in the math unit.
Then you should search more ;) Acutally there is NaN
You are right. Sorry, probably a typo in the search. FPC is great.
Mattias
Op Thu, 25 Jan 2007, schreef Mattias Gaertner: Well, not here. See the program below. Can…
Show quoted text 20 lines
Op Thu, 25 Jan 2007, schreef Mattias Gaertner:
> On Thu, 25 Jan 2007 12:10:44 +0100 (CET)
> Daniël Mantione wrote:
>
> >
> >
> > Op Thu, 25 Jan 2007, schreef Mattias Gaertner:
> >
> > > How to set a double to NAN (not a number)?
> > > I only found a IsNAN in the math unit.
> >
> > const nan=0/0;
>
> unit1.pas(28,10) Error: Division by zero
Well, not here. See the program below. Can you check if this depends
on the compiler version or mode? 0/0 should normally work.
It depends on -Cr.
Mattias