[vtk-developers] Adding IsNan() to vtkMath

[Brad King]

Moreland, Kenneth wrote:
> int vtkMath::IsNan(double x)
> {
>   return isnan(x);
> }
> 
> Does anyone know any portability issues with this?

Getting this to work everywhere (not just compile) is somewhat
involved.  The vxl folks have code here:

http://vxl.svn.sourceforge.net/viewvc/vxl/trunk/core/vnl/vnl_math.cxx?revision=24028&view=markup

Note that there should be overloads for float, double, and
perhaps long double.

-Brad