On 11/17/05, <b class="gmail_sendername">Robert Ramey</b> &lt;<a href="mailto:ramey@rrsd.com">ramey@rrsd.com</a>&gt; wrote:<div><span class="gmail_quote"></span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">







<div><font face="Arial" size="2">The problem is given something like +/inf how do I 
create a float with that value?&nbsp; </font></div>
<div><font face="Arial" size="2"></font>&nbsp;</div>
<div><font face="Arial" size="2">Can I use something like this?</font></div>
<div><font face="Arial" size="2"></font>&nbsp;</div>
<div><font face="Arial" size="2">double x;</font></div>
<div><font face="Arial" size="2">x = 
std::numeric_limits&lt;T&gt;::quiet_NaN();</font></div>
<div><font face="Arial" size="2"></font>&nbsp;</div>
<div><font face="Arial" size="2">assert(x == 
std::numeric_limits&lt;T&gt;::quiet_NaN());</font></div></blockquote><div><br>Not sure why you'd need the check if you've just done the assignment, but I don't believe that NaN-ness and Inf-ness can be tested like that.&nbsp; You'd can use  isnan() (which is spelled _isnan on Windows) to test for NaN.&nbsp; Not sure if there is any way to distinguish between signaling and non-signaling NaNs however.&nbsp; BSD-derived systems provide an isinf() as well, but Windows seems to omit this one.&nbsp; There is also the C99 fpclassify stuff others have alluded to.&nbsp; Win32 has _fpclass it would seem.
<br></div><br></div><br>-- <br>Caleb Epstein<br>caleb dot epstein at gmail dot com