Valery Pipin wrote:
>On Thursday 11 March 2004 23:47, Richard Fateman wrote:
>| Rather than having 4 people write another graphics package for Maxima
>| to replace the packages that exist (I don't know if they work, but Xmaxima
>| plots for me..), why not write something that doesn't already exist.
>|
>| Having just looked at the list of product offerings from Mathworks
>| (vendors of Matlab), why not pick your favorite toolkit and see what
>| you can do symbolically. e.g. signal processing, control systems,
>| financial computations, pdes, ....
>These examples are not quite ingenious. Suppose we have all this happines. In
>processing all of these problems you need to visualize results. So truly
>good plotting package will be needed. I can not imagine the signal
>processing package without good plotting functions. Right now maxima's
>plotting package can serve as preview only. I
>
>Yet I think it is better to think on develop the interaction interface between
>maxima and, say, octave (or scilab) than transfer their routines to maxima.
>
Such routines will be strictly numerical, and cannot take advantage of
symbolic computation. The numerical routines can be useful, of course.
Access to matlab or octave (all of it) can probably be done today, about
the same way as I did it 20 years ago. You need 4 programs in maxima.
(a) matput... send a matrix to matlab
(b) matget .. get a matrix from matlab
(c) matrun ... send a string of commands to matlab.
(d) matinitialize .. start up a matlab process.
>
>
>Another problem is that right now maxima is more similar to imperative
>language like fortran or pascal. There is no good pattern matching package.
>So transfering program from mathematica will be not easy.
>
The lisp package mockmma has a mathematica-style syntax and a mathematica
pattern matcher (as of Mathematica 3.0), all written in Common Lisp.
Free/open source.
What is missing is the detail semantics of many of the routines, so it
is not
trivial to do the rest.
>
>
>rgds,
>Valery
>
>
>
>
>
>
>
>_______________________________________________
>Maxima mailing list
>Maxima@www.math.utexas.edu
>http://www.math.utexas.edu/mailman/listinfo/maxima
>