Some general comments from a casual user of matlab, octave, etc.
First, I don't see much advantage to being able to call some CAS from within octave?
I can understand why it is useful to have some Maple routines
available in matlab.
1) Matlab has a whole graphic interface. It is an entire environment
from which to work.
2) Maple and Matlab, are, separately, expensive packages. So, an
add-on from one to the other with a small extra expsnse is
sensible. Actually, I tried both together (the symbolic toolkit in
matlab) and each separately, and I find I prefer the latter. For
me, the symbolic implementation in Matlab is cumbersome. The main
reason for it is to save money, not to increase functionality. If
price were not a serious consideration, I would simply have both.
In this connection, I briefly tried SAGE as well. It does many
good things, but I did not see much advantage to calling maxima up
from within it. I preferred to translate data and use maxima
separately. (Admittedly, I did not use it much, so I can imagine
the utility of some big number-theoretic or group-theoretic operation carried out in
SAGE which might make use of a maxima routine). That is for the
significant users to discuss.
Also Octave (as far as I have seen, and I am not a big user) is a free command
line interface to many routines available in matlab. It does not
(again as far as I know), have a Win-XP version. So, if one wants to
do CAS at the command line in unix, why not just fire up maxima
itself. Of course there is some data exchange necessary, read-write
of matrices, etc.
There would be some advantage to calling up graphics routines
(e.g. matplotlib) from within maxima. But that looks like having
octave available to do routines from within maxima (or just the
matplotlib routines themselves), not the other way
around. This, again AFAIK, would be problematic in Win-XP.
So, before a real commitment is made to get octave and maxima to work,
one from within the other, it seems to me some indication of the
potential benefits should be given.
-sen
---------------------------------------------------------------------------
| Sheldon E. Newhouse | e-mail: sen1 at math.msu.edu |
| Mathematics Department | |
| Michigan State University | telephone: 517-355-9684 |
| E. Lansing, MI 48824-1027 USA | FAX: 517-432-1562 |
---------------------------------------------------------------------------
On Thu, 1 Feb 2007, Doug Stewart wrote:
> Nicolas Pettiaux wrote:
>> Maple is the CAS (computer algebra software) of Matlab.
>>
>> How difficult would it be to have maxima more closely linked with
>> octave in a similar way ?
>>
>> I know that there is the SAGE project that is very interesting, but
>> this is yet another language of its own, aka python, on top of the
>> octave and maxima ones.
>>
>> Thanks,
>>
>> Nicolas
>>
>>
>
> I did some work on this and it is not hard to run Maxima from Octave.
> The problem is that you now have 2 different languages to learn. Octave
> and Maxima.
>
> Or we would have to write a converter between the two systems. ie. Have
> a complete symbolic syntax addition to octave, and then a converter
> that runs behind the scene that would change to Maxima syntax. This
> wouldn't be too hard for integration and derivatives and some series.
> But Maxima is huge and we would have to decide how much to implement.
>
> If we don't go with the converter system and we make everyone learn
> Maxima then I do not see much advantage to join them together. I now use
> mainly Octave, and when I have to I fire up Maxima or Axium and then
> bring the results back to Octave.
>
> I found that there is big learning curve for Maxima
>
> Just my 2 C worth.
>
> Doug
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima
>