Dimitri Pourbaix a écrit :
> Luc,
>
>> Concerning MATH321, the partial fix I proposed has been withdrawn by
>> the complete rewrite of SVD.
>> This partial fix also was not considered mathematically sound by Dimitri
>> since the dimensions of the matrices did not match complete SVD. Well,
>> this was in fact desired as MATH321 explicitly asks for nonfull SVD.
>>
>> So I'm not sure anymore what we need to do. If we follow Dimitri's
>> rationale, we should close this as WON'T FIX. If we decide to fix it, I
>> think we should postpone it to after 2.1 to have time to decide how we
>> solve it. SVD has already been vastly improved by Dimitri's work, so I
>> would really much like to see it published now.
>>
>> Dim, what do you think about this ?
>
> The version I uploaded does not benefit from any sparsity of the matrix
> to be decomposed. I have not looked at any algorithm for this kind of
> matrix as it would likely involve a different way of storing the matrix
> itself. I agree to put "WON'T FIX" for 321 unless someone can come up
> with an example showing that that, for a sparse matrix, the present
> version does not return the right answer.
I think there are two parts in the isse.
The first one is related to sparse matrix and we don't have an answer
yet. The second part is related to compute a partial set of singular
values. This is used for example in image compression or to find a
matrix with reduce rank that is the closest possible to an input matrix.
for this part, we may have an answer.
So for the first part, I think WON'T FIX may be an answer (or
INCOMPLETE, asking for a patch ...). For the second part, I would
suggest to simply postpone to 3.0.
Luc
>
> Dim.
> 
>
> Dimitri Pourbaix *
> Institut d'Astronomie et d'Astrophysique * Don't worry, be happy
> CP 226, office 2.N4.211, building NO * and CARPE DIEM.
> Universite Libre de Bruxelles *
> Boulevard du Triomphe * Tel : +322650.35.71
> B1050 Bruxelles * Fax : +322650.42.26
> http://sb9.astro.ulb.ac.be/~pourbaix * mailto:pourbaix@astro.ulb.ac.be
>
> 
> To unsubscribe, email: devunsubscribe@commons.apache.org
> For additional commands, email: devhelp@commons.apache.org
>

To unsubscribe, email: devunsubscribe@commons.apache.org
For additional commands, email: devhelp@commons.apache.org
