infinity correct maxima



maxima-bounces at math.utexas.edu wrote on 09/02/2008 07:47:48 AM:

> I find simplus and friends pretty much undocumented and unreadable.  A 
> 100% rewrite doesn't sound like a bad idea to me.

(1) Simplus and friends (simplus, plusin, and pls) have about 20 "go
to" statements. Ouch.

(2) The "infinity correct" part of my simplus function can be easily
eliminated. Largely, I inserted this hunk of code as a demonstration
of the special case dispatch part of the altsimp/simplus code.

(3) A better name for "infinity correct" might be "infinity safe."

(4) Algorithmically, simplus is little more than 8th grade
mathematics. Maybe we shouldn't be so shy in replacing it. On my own,
I don't think I'll try to do that.

(5) Making multiplication or exponentiation infinity safe would almost
certainly cause lots of bugs in limit. The limit function might depend
too much on the particular way expressions Maxima simplifies
expressions.

(6) altsimp/simplus slows Maxima for sums with few terms (say 8 or
fewer). I don't know a good way to do that.


Barton