From bug-octave-request at che dot utexas dot edu  Wed Jun 29 07:35:37 1994
Subject: octave for linux
From: janne at avocado dot pc dot helsinki dot fi (Janne Sinkkonen)
To: matthew at redbird dot umsl dot edu (Matthew Feldt)
Cc: bug-octave at che dot utexas dot edu
Date: Wed, 29 Jun 94 15:23 EET DST


 > am confused about some results I have seen.  If you could offer insight
 > to the following results or need more information please contact me at
 > the e-mail address below.

 > octave:1> a = [ 1 2 3; 4 5 6; 7 8 9]

What's the problem? Your matrix is singular:

octave:262> cond(a)
ans =  6.0262e+16

octave:263> svd(a)
ans =

   1.6848e+01
   1.0684e+00
   2.7958e-16

So, don't expect very accurate results from inverse().

I guess determinant is most probably NOT calculated by the familiar
sum of products of permutations rule, but instead by doing something
like prod(eig(a)) but more clever. In fact the answer you show is
quite near the actual result which is 0.

The smaller a is not singular.

--
Janne