Re: SVD, EIG and CHOL of a matrix

[Jaroslav Hajek]

On Mon, Jan 25, 2010 at 9:10 AM, Alberto Frigerio
<address@hidden> wrote:
>
> Do you mean that the previous versions of Octave, when runnig [R,p]=chol(A)
> with a not positive definite matrix, instead of giving error, returned the
> last acceptable tranformation of A before finding the error?

Yes.

> Because I tried
> to do the same thing, but I don't have the same results
> http://old.nabble.com/file/p27303247/cholmio.m cholmio.m
> http://old.nabble.com/file/p27303247/Results.txt Results.txt
>

The factorization is done by Octave on the upper triangle, whereas you
seem to work on the lower part, hence probably the difference.

-- 
RNDr. Jaroslav Hajek, PhD
computing expert & GNU Octave developer
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz