[Scilab-users] sparse() to build a block-diagonal matrix?

Chin Luh Tan chinluh.tan at bytecode-asia.com
Tue Nov 26 03:43:45 CET 2019


Hi, 



from the first glance on the documentation you show: 



"Beside this function, you can also use sparse() primitive to build a block diagonal sparse matrix."



it sound to me that this statement said the sparse could be used to create block diagonal matrix with the similar inputs to sysdiag.





But after a trying to stare at it again... I think it might mean:


" to convert the block diagonal matrix to block diagonal sparse matrix "

which likely is the one shown in the example: 

S=sysdiag([1 2; 3 4], [5 6; 7 8], [9 10; 11 12], [13 14; 15 16])
S=sparse.html(S)

which simply to say that we could use sparse to convert the full matrix created by sysdiag to sparse?



Thanks.

CL













---- On Mon, 25 Nov 2019 23:02:42 +0800 Samuel Gougeon <mailto:sgougeon at free.fr> wrote ----


Hello,

In thehttps://help.scilab.org/docs/6.0.2/en_US/sysdiag.html, it is stated that sparse() can be used to build a
    block-diagonal matrix:
 "Beside this function,
      you can also use https://help.scilab.org/docs/6.0.2/fr_FR/sparse.html() primitive to build a block
      diagonal sparse matrix."
 
 But i did not find any description nor example of such a feature
      in thehttps://help.scilab.org/docs/6.0.2/en_US/sparse.html. 
 I have quickly tried some "exotic" syntaxes, without result.
  Does anyone know if this feature
          actually exists, and if so what's its syntax?

I know that sysdiag() works also
          with sparse objects.
 There are several issues in the sysdiag() page, and i would
          like to have clear ideas about the reference to sparse.
 
 Thanks
 Samuel
 




_______________________________________________
users mailing list 
mailto:users at lists.scilab.org 
http://lists.scilab.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/users/attachments/20191126/28f8b79c/attachment.htm>


More information about the users mailing list