James Amundson [amundson@users.sourceforge.net] wrote:
>=20
> I use Octave every day. I definitely think that some sort Maxima-Octave
> link would be useful. Even though I have thought about linking the two
> quite a bit, I still have conceptual questions about how the link should
> work. Unfortunately, I have not seen the Matlab symbolic toolkit.
[...]
> How would you imagine the interface working?
It's always seemed to me that the best way to do that is to have
special-case algorithms and data types. i.e. have the Matrix type able
to store itself as a numeric (fortran-style) matrix when it's known not
to contain symbolic entries and the precision requested is less than or
equal to the machine precision. Algorithms like multiply and
determinant should be able to operate on them using octave/lapack
algorithms.
This is the way Maple does it, and in addition has other numeric
algorithms like integrate, etc.
Of course this requiires lisp <-> fortran or lisp <-> C callability
which I don't know how to do.
Cheers,
Bob McElrath [Univ. of California at Davis, Department of Physics]