maxima-bounces at math.utexas.edu wrote on 10/25/2006 07:54:33 AM:
> Hi,
>
> I work with Maxima 5.10.0/CLISP 2.38.
> Trying to use matrixexp package, I get:
You can avoid this bug by doing load(linearalgebra) instead of
load(matrixexp):
(%i1) load(linearalgebra)$
(%i2) spectral_rep(matrix([1,0],[0,1]));
(%o2) [[1],[matrix([1,0],[0,1])],matrix([0,0],[0,0])]
I'll think about the best way to fix this bug. (And it is
a bug; I think a user should never see "Maxima encountered
a Lisp error.")
Note: If you want the Jordan canonical form, use the diag
package. Both matrixexp and diag do functions of matrices.
For matrix exponentiation, I know of no advantages of one
package over the other (I haven't looked for any advantages
either). Not that it matters, but I don't recall which
package came first (but I think diag came first). I had some
particular use for the matrix spectral representation, so I
wrote the code for it; matrix exponentiation was a natural
application. Gosei Furuya wrote the code for the Jordan canonical
form. The overlap has caused some confusion. The two packages
aren't redundant.
We (the Maxima community) appreciate all worthy contributions.
If somebody writes code for matrix exponentiation that is better
than what we have, that's great. We'll take it. Another thing:
the location of a package (src, share, or share/contrib) doesn't
really reveal anything about its quality or utility. High-quality
useful code is well, useful and high-quality regardless of which
directory it is in.
Barton