>
> Hey,
>
> Here is one of these annoying things we sometimes run into:
>
> 1. From the menus select 'DATA:CONSTANT:COMPLEX' and put on your model,
> making it al little larger than default for displaying larger number
> formats.
> 2. Try entering e.g. (1E-20,1E-20) or even (1E-40,1E-40) into the constant.
> 3. What do you see? Does it give resp. (10E-21, 10) or (100E-42, 10) on your
> system? Why 10?????
>
I've seen similar problems with complexes in the FROM STRING and have
concluded that there's a bug in the 3.12 number builder.
Try using metric suffixes instead- this has been a good workaround
for me in some cases. 1e-20 would be .01a, etc. Or of course
you can use a "build complex".
No idea what the problem is. but what I have seen is generally restricted
to the second value in a coord or complex, so I don't believe it's
just the vagaraties of IEEE 64 bit floating point math.
Stan Bischof
stanb@sr.hp.com
> Hey,
>
> Here is one of these annoying things we sometimes run into:
>
> 1. From the menus select 'DATA:CONSTANT:COMPLEX' and put on your model,
> making it al little larger than default for displaying larger number
> formats.
> 2. Try entering e.g. (1E-20,1E-20) or even (1E-40,1E-40) into the constant.
> 3. What do you see? Does it give resp. (10E-21, 10) or (100E-42, 10) on your
> system? Why 10?????
>
I've seen similar problems with complexes in the FROM STRING and have
concluded that there's a bug in the 3.12 number builder.
Try using metric suffixes instead- this has been a good workaround
for me in some cases. 1e-20 would be .01a, etc. Or of course
you can use a "build complex".
No idea what the problem is. but what I have seen is generally restricted
to the second value in a coord or complex, so I don't believe it's
just the vagaraties of IEEE 64 bit floating point math.
Stan Bischof
stanb@sr.hp.com
Here is one of these annoying things we sometimes run into:
1. From the menus select 'DATA:CONSTANT:COMPLEX' and put on your model,
making it al little larger than default for displaying larger number
formats.
2. Try entering e.g. (1E-20,1E-20) or even (1E-40,1E-40) into the constant.
3. What do you see? Does it give resp. (10E-21, 10) or (100E-42, 10) on your
system? Why 10?????
We use vee 3.12 on HP-UX 9.05 on a 100MHz, 715 machine. However, vee 3.1 on some
PC with MS-windows 3.11 acts the same.
--
Remi Tuijtelaars
Philips Semiconductors, Dev. T&D, Building BY2.021,
Gerstweg 2, NL-6534 AE Nijmegen, Netherlands.
tel/fax: + 31 24 353 - 2812 / 3160.
email: Remi.Tuijtelaars@nym.sc.philips.com
(seri: nytuijte@nyhp02)