Article View: comp.sys.next.programmer
Article #1894Re: trapping NaN
From: Frank Mitchell
Date: Mon, 23 Nov 1992 16:12
Date: Mon, 23 Nov 1992 16:12
25 lines
1146 bytes
1146 bytes
In article <1992Nov21.111335.4014@arizona.edu> writes:
> In article <1992Nov20.221038.1170@proforma.com> writes:
> > I would like to trap NaN -- the "not a number" condition that floating
point
> > values can take.
> It turns out that (NaN != NaN) is true. So, in some code I did before, I
used:
>
> if(x == x) { /* x is NOT NaN */
> < use x>
> }
> else {
> <whatever for the NaN case>
> }
>
I tend to use the finite() call, which also traps infinities (which may
be good or bad). Check out its man page for details.
--
Frank Mitchell, Business Systems Analyst, First National Bank of Chicago
email:frank@fnbc.com (NeXTmail)
A happy child is merely an unhappy child who is having fun.
-- Morticia Addams
Technical Details
Message-ID:
<1992Nov23.161217.23247@fnbc.com>Path:
arc1-ca.pugleaf.net!sparky!uunet!psinntp!juliet!newsReferences:
<1992Nov21.111335.4014@arizona.edu>Show Full Headers
Complete Article Headers
Newsgroups: comp.sys.next.programmer Path: arc1-ca.pugleaf.net!sparky!uunet!psinntp!juliet!news From: frank@fnbc.com (Frank Mitchell) Subject: Re: trapping NaN Message-ID: <1992Nov23.161217.23247@fnbc.com> Sender: news@fnbc.com Organization: First National Bank Of Chicago, Chicago IL, USA References: <1992Nov21.111335.4014@arizona.edu> Distribution: world,local Date: Mon, 23 Nov 92 16:12:17 GMT Lines: 25