Subject: Re: CCM3.6 on SUN
From: Erik Kluzek (erik@ncar.ucar.edu)
Date: Wed May 03 2000 - 15:25:28 MDT
Anita
Regarding your message as below...
On Wed, 3 May 2000, Anita Rodrigues de Moraes Drumond wrote:
> Although, when I try to run the model I get the following messages at the
> end of the model run.
>
> Note: IEEE floating-point exception flags raised:
> Inexact; Underflow;
> IEEE floating-point exception traps enabled:
> overflow; division by zero; invalid operation;
> Nonstandard floating-point mode enabled
> See the Numerical Computation Guide, ieee_flags(3M), ieee_handler(3M),
> ieee_sun(3M)
>
The above is actually good news! What it's telling you is that the math
completed successfully (ie. there wasn't any overflow, division by zero,
or invalid operation).
It's also telling you that it checked for those
things and nothing bad happen (doesn't happen by default with IEEE so this
is also good news). If you've ever had a program run successfully and just
create "NaN" (not a number) as output or "Infinity" you know how frustrating
that can be. Now, the only flags that were raised were "Inexact" and "Underflow".
Inexact is what happens with floating point arithmetic. So the result of
pi*2.58 get's rounded off rather than the exact number it is, since it's only
represented by a finite decimal point number. Underflow just means that results
that were very small got rounded off to zero. For the CCM this is quite normal.
Erik Kluzek, (CGD at NCAR)
National Center for Atmospheric Research
Boulder CO, (off) (303)497-1326 (fax) (303)497-1324
--------- Home page and public PGP key---------------
http://www.cgd.ucar.edu/~erik
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
This archive was generated by hypermail 2b27 : Thu Jan 04 2001 - 10:01:48 MST