| Re: Puzzle with floating point underflow [message #26452 is a reply to message #26334] |
Fri, 24 August 2001 07:45   |
Paul van Delst
Messages: 364
Registered: March 1997
|
Senior Member |
|
|
Martin Schultz wrote:
>
> Paul van Delst <paul.vandelst@noaa.gov> writes:
>
>> Martin Schultz wrote:
>>>
>>> Not a real problem because underflows are considered harmless
>>
>> Aargh! wash your mouth out with soap! :o)
>>
>> Sigh.
>>
>> Paul "!EXCEPT=2" van Delst
>>
>
> Can you recommend a soap brand that doesn't taste too bad?
Well, there's this brewery in Pennsylvannia called Yuengling that makes a pretty good
lager.... oh. _Soap_. Sorry. :o)
> In fact, I only quoted from the IDL online help:
>
> "In the vast majority of cases, floating-point underflow errors
> are harmless and can be ignored."
Oh man, what a ridiculous statement to make in a manual. Makes me think the writer(s)
never actually thought IDL would be used for anything _really_ important. I hope anyone
using IDL to, say, process MRI images to look for anomalies never read that part of the
manual. Given tortuous enough code, the same can be said for fp overflow errors. Or fp
inexact errors. Divide by zero will probably break things everytime depending on how the
result is handled. IDL has a great way of handling these sorts of errors (typically with
minimum user intervention) that doesn't translate into other programming languages, and
bad habits can be hard to break.
>
> I don't like statements like this either, but !EXCEPT=2 really
> slows down my code too much ;-)
Well, I don't recommend it (!EXCEPT=2) as something one has turned on all the time. But
for code testing it's a must.
paulv
--
Paul van Delst A little learning is a dangerous thing;
CIMSS @ NOAA/NCEP Drink deep, or taste not the Pierian spring;
Ph: (301)763-8000 x7274 There shallow draughts intoxicate the brain,
Fax:(301)763-8545 And drinking largely sobers us again.
Alexander Pope.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
