Number precision
-
arne-helge
- Newbie
- Posts: 1
- Joined: Tue Feb 07, 2006 2:47 pm
Number precision
Is there a way I can represent the numbers 1.000005 and 1000005 in a Vensim model? In my Vensim double-precision version 5.3 these numbers are rounded off to 1.00001 and 1000010 causing wrong results.
Number precision
Hi
Normally SD is used to understand apparently complex systems. Although Vensim can deliver precise results, going to this level of precision in a context of structural complexity that generates imprecisions much more important than this scale of measure may be highly unuseful.
What justifies going to this level of precision?
Regards.
J.J. Laublé
Normally SD is used to understand apparently complex systems. Although Vensim can deliver precise results, going to this level of precision in a context of structural complexity that generates imprecisions much more important than this scale of measure may be highly unuseful.
What justifies going to this level of precision?
Regards.
J.J. Laublé
Number precision
A possible solution would be to normalyze the values.
You could define a normal constant = 1000 and the normalysed value = 1000 * Normal constant and the difference value = Normal constant * 0.005 and the exact value = normalyzed value + difference value.
What is the sensibilty of the results to the value?
Regards.
J.J. Laublé
You could define a normal constant = 1000 and the normalysed value = 1000 * Normal constant and the difference value = Normal constant * 0.005 and the exact value = normalyzed value + difference value.
What is the sensibilty of the results to the value?
Regards.
J.J. Laublé