[Trilinos-Users] question on extracting submatrices
Williams, Alan B
william at sandia.gov
Mon Dec 22 13:04:53 MST 2008
Did anyone else reply to you about this already?
Perhaps you could give a little more detail about your scenario.
Do you want B to be a 'view' such that any changes made to A automatically appear in B?
Do you want the underlying numbering to remain compatible such that mathematical operations can be performed mixing A and B (adding and/or multiplying)?
> -----Original Message-----
> From: trilinos-users-bounces at software.sandia.gov
> [mailto:trilinos-users-bounces at software.sandia.gov] On Behalf
> Of rrossi at cimne.upc.edu
> Sent: Thursday, December 11, 2008 1:29 PM
> To: trilinos-users at software.sandia.gov
> Subject: [Trilinos-Users] question on extracting submatrices
> Dear all,
> i am a new user of the Epetra, Amesos, Aztect and IFpack which we are
> currently using as an option for our nonlinear FE analysis
> in parallel.
> First of all my felicitations to the developers. We found the
> very good and much simpler than alternatives we looked at (PETSC)
> I'd like however to ask a practical question to see if i can make
> effective use also of the meros package ;-)
> suppose i have a matrix of type Epetra_FECrsMatrix
> distributed between two
> say that such matrix is
> A =
> 1 2 3 4
> 5 6 7 8
> 9 10 11 12
> 13 14 15 16 17
> distributed so that the first 2 lines are on proc0 and the
> others on proc1
> i would like to create a new matrix containing (say) the
> first and third
> rows and columns of A
> defined for exampel as
> B =
> 1 3
> 9 11
> or another containing (say the elements)
> G =
> where all the elements of B can stay on the same processors
> they were in A
> does it exist any "easy" way of doing this?
> Trilinos-Users mailing list
> Trilinos-Users at software.sandia.gov
More information about the Trilinos-Users