Re: new version of gsvd

[David Bateman]

Pascal Dupuis wrote:
> Hello,
>
> I have the pleasure to send a revision of the gsvd routine, which
> David included in octave-forge/main/linear-algebra.
>
> Main changes:
> - only returns the non-trivial extended singular values, no padding
> with ones and 0
> - If called with a 6th output arg, returns R such that
>   U'*A*X = C*R; V'*B*X = S*R
> - works correctly if the number of rows of A is smaller than the
> number of columns of B
> - a case test, test_gsvd.m, has been included. It verifies the basic
> properties on all cases of possible rank-deficiencies: rank(A) <
> rank(B), rank(B) < rank(A), rank([A' B']) < min(rows([A' B']),
> columns([A' B'])). Real and Complex computations are tested.
> - updated the doc
>
> Todo: I don't know that much about texinfo, the 'help gsvd' complains
> and display the raw code. There is some issue with tags opening and
> closing, I can't figure it.
>
> Best regards
>
> Pascal Dupuis 
>
>   
Pascal,

I have very poor connection to sourceforge at the moment as I can't
commit from work and don't have a working modem at home. Therefore, I
propose to add your sourceforge id to octave-forge and can you commit
these changes yourself.. What is your sourceforge ID?

Regards
David


-- 
David Bateman                                address@hidden
Motorola Labs - Paris                        +33 1 69 35 48 04 (Ph) 
Parc Les Algorithmes, Commune de St Aubin    +33 6 72 01 06 33 (Mob) 
91193 Gif-Sur-Yvette FRANCE                  +33 1 69 35 77 01 (Fax) 

The information contained in this communication has been classified as: 

[x] General Business Information 
[ ] Motorola Internal Use Only 
[ ] Motorola Confidential Proprietary