Re: Problem With Solving Matrix Equations

[Quentin Spencer]

MoneyHaus wrote:
> The problem I am having with Octave seems rather elementary, but I can't seem
> to get my head around it. Solving linear equations using the "\" operator,
> such as x=A\B, just won't ever work for me.
>
> For example, the following should solve easily, yet my Octave installation
> (latest) won't:
>
> octave:17> A=[0,2,0,1;2,2,3,2;4,-3,0,1.;6,1,-6,-5];
> octave:18> B=[0;-2;-7;6];
> octave:19> inv(A)*B
> ans =
>
>   -0.50000
>    1.00000
>    0.33333
>   -2.00000
>
> octave:20> A\B
> warning: matrix singular to machine precision, rcond = nan
> warning: attempting to find minimum norm solution
>
> And there it stops... it'll just sit there and never solve before. I'm
> pretty sure I shouldn't be getting those warning messages to being with.
>
> Any help would be greatly appreciated!
>   

Like the others have said, it would help if you provided more 
information. Having said that, this looks like a problem with your 
LAPACK library rather than octave. Errors like this have come up before, 
particularly with some Linux distributions when they switched to gcc 
4.0. I don't know what you're running, but I would suggest checking for 
a lapack update.

-Quentin