Re: platform dependency?
Since the Jama code is the same byte-code on all platforms, the question
"Is Jama code platform dependent?" is not the right question. Jama (and any
other pure Java program) is identical on all platforms. Given identical Java
Virtual Machines, it will always (by definition) produce the same result.
And there's the rub...You are running two different JVMs.
Floating-point results *are* JVM-dependent (even though they shouldn't be).
This shows up more in numerical methods because intense numerical work
highlights the differences in floating point hardware (such as
different-length internal accumulation registers, truncate vs. round
decisions, and so on.)
So, while Jama is not platform dependent, the *results* you get from Jama
(or any other Java program using floating point) *are* dependent on the JVM.
One could say that "Jama behaves in a platform-dependent manner", but that
really would miss the point that the difference is not in Jama.
----- Original Message -----
From: "michael ege" <firstname.lastname@example.org>
To: "Multiple recipients of list" <email@example.com>
Sent: Sunday, September 16, 2001 2:31 PM
Subject: platform dependency?
> I'am using Jama on a NT System and on an Unix Server (Sun Solaris).
> The problem is:
> We do for example a rank calculation of a 250 dimensional quadratic
> Thereby we get different results, the calculation on the NT yields full
> and the one on the Unix machine rank 1. The bigger eigenvalues of this
> matrix nearly the same. But there are negativ eigenvalues on the Unix
> calculation against no negativ eigenvalues on the NT OS.
> Could the Jama code be platform depend?
> Michael Ege
Date Index |
Thread Index |
Problems or questions? Contact firstname.lastname@example.org