Bug [1954846] bessel_i(1/2,0) -> divide by zero error
Subject: Bug [1954846] bessel_i(1/2,0) -> divide by zero error
From: Robert Dodier
Date: Sat, 5 Jul 2008 09:57:23 -0600
On 7/5/08, Stavros Macrakis <macrakis at alum.mit.edu> wrote:
> On Sat, Jul 5, 2008 at 4:07 AM, Dieter Kaiser <drdieterkaiser at web.de> wrote:
> > Might it be possible to introduce to Maxima the correct handling of
> > infinities, so we can define the singularities of functions more completly?
>
> That would be nice, but several attempts over the years have not
> resulted in usable code.... Also, it is not clear what the correct
> handling is.
For the moment, I think it's OK to trigger an error.
When/if there is some code to handle infinity in arithmetic,
we can revise the Bessel code (and other stuff) to return infinity.
Another approach is to throw something instead of triggering an error.
Callers can catch specific conditions and take appropriate action;
"there is an error" is too blunt. Actually we could put code in place
to throw conditions (in the Bessel code, I mean) even without
having handlers to catch them; without a handler, the user sees
the condition as a Lisp error, I'm pretty sure.
best
Robert Dodier