[racket] matrix-solve and approximation errors

Interesting; even more when using rounded numbers returns a correct
solution!

To protect against such errors, maybe `matrix-solve` could run a post-check
to verify that M×X=B up to some epsilon, depending on an optional argument?
Or maybe just mention to do this check systematically in the docs?


On Wed, Apr 16, 2014 at 1:31 PM, Jens Axel Søgaard <jensaxel at soegaard.net>wrote:

> 2014-04-16 12:18 GMT+02:00 Jens Axel Søgaard <jensaxel at soegaard.net>:
> > Hi Laurent,
> >
> > I think the underlying problem is that the matrix is *very* close to
> > an invertible one:
>
> non-invertible
>
>
> >> (matrix-determinant
> >      (matrix [[ 1     0     9/10  1]
> >                  [ 0     1     1/10  1]
> >                  [ 9/10  1/10 82/100 1]
> >                  [ 1     1      1   0]]))
> > 0
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140416/38a028cc/attachment.html>