Re: large sparse matrix

[Jaroslav Hajek]

On Tue, Sep 1, 2009 at 1:12 AM, Carlo de Falco<address@hidden> wrote:
>
> On 31 Aug 2009, at 21:53, David Bateman wrote:
>>
>> Yes, but the configuration of SuiteSparse in terms of the pivot
>> threshold is the default that TIm Davis sets for Octave and whatever
>> mathworks chose to use for matlab. So don't expect exactly the
>> behavior.
> I see...
>
>> In Octave what does
>>
>> matrix_type(A)
>> give?
>
>
>  >> matrix_type (A)
> ans = Positive Definite
>  >>
>
>> If it doesn't flag your matrix as PD then there might be an issue
>> with the detection of the matrix and PD and the solver used in
>> Octave will be LU.
>
> on the other hand what surprises me is that:
>
>  >> issparse (B)
> ans =  1
>  >> matrix_type (B)
> ans = Rectangular
>  >> issparse (A)
> ans =  1
>  >> mat = [A B' spalloc(rows(A),1,0); B spalloc(rows(B),columns(A),1)
> E'; spalloc(columns(A),1,1) E 0];
>  >> matrix_type (mat)
> ans = Full
>  >> issparse (mat)
> ans =  1
>
> why is mat marked as full???
>

Full is what you get when nothing special about the matrix is
discovered. Did you expect something else? The matrix is obviously not
positive definite, as seen from its construction, and I can see no
apparent other structure (triangularity or bandedness), though it of
course depends on the values of A, B and E.

hth

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