On Dec 12, 2007 5:30 AM, Paul Zimmermann <[EMAIL PROTECTED]> wrote:
>
> > > > >sage: pari('2').besselk(3) # random
> > > "Random" here doesn't mean what you think. [...]
> > We really need to kill all of those and add "..." to account for the
> > imprecision caused by different CPUs/operating systems/compilers.
>
> even better would be to adopt a computational model such that all
> numerical computations can give only *one* correct result. Then you
> could simply compare to the expected result with utilities like "diff".
That would be nice but isn't realistic, since Sage includes systems like
Numpy / Scipy / PARI, etc, and none of these systems adopt such a model,
for whatever reason.
The realistic thing is what Mabshoff suggests, which is changing the
doctests to look like this:
sage: pari('2').besselk(3)
0.061510458471742...
where the ...'s the bits that are random.
David -- maybe changing some of these would be something
you might want to do?
William
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-support@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/sage-support
URLs: http://sage.math.washington.edu/sage/ and http://sage.scipy.org/sage/
-~----------~----~----~----~------~----~------~--~---
