multiplying matrices into tensors?

[James Cloos]

I have a pair of matrices, such as:

a = [  0  20  10 -10 -20   0 ;
     -25  -5  15  15  -5 -25 ;
       0   0   0   0   0   0 ;
       5   1   1   1   1   5 ]

b = [ 0  100  100    0   0  100  100    0   0  100 ;
      0    0  100  100   0    0  100  100   0    0 ;
      0    0    0    0   0    0    0    0   0    0 ;
      1    1    1    1   1    1    1    1   1    1 ]

representing lists of points where each column is an xyzw point in R³P.

(In this case they are actually lists of points in R²P, but the
underlying storage always stores in R³P and wants them ordered
as above.)

I need to mupliply the points together piecewise to generate a 3-d tensor
representing a 2-d array of xywz points, using the typical:

             $[ c_{i,j} = a_i × b_j $]

formulation.

Does octave have an easy way of accomplishing that?  Especially given
the homogeneous coordinates?

Or will I need to write a funtion to loop through the two and multiply
as needed?

I'm more comfortable doing this stuff in C or Perl, but I wanted to see
whether octave would make prototyping quicker.  I've looked through the
docs and some of the code (including octave-forge code), but haven't yet
found the right clue....

Thanks,

-JimC
-- 
James Cloos <address@hidden>         OpenPGP: 1024D/ED7DAEA6