August 2008 - petsc-dev - mailman3.cels.anl.gov

Re: how do you access the nonzero elements in SeqAIJ matrices
by Zi-Hao Wei 04 Aug '08

04 Aug '08

On Mon, Aug 4, 2008 at 10:53 PM, Ahmed El Zein <ahmed(a)azein.com> wrote: > my code looks like: > ierr = MatCreate(PETSC_COMM_WORLD,&A); CHKERRQ(ierr); > ierr = MatSetSizes(A,M,N,M,N); CHKERRQ(ierr); > ierr = MatSetType(A, MATSEQAIJ); CHKERRQ(ierr); > ierr = MatAssemblyBegin(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); > [...lots of MatSetValue() calls...] > ierr = MatAssemblyEnd(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); I think that the code should be ierr = MatCreate(PETSC_COMM_WORLD,&A); CHKERRQ(ierr); ierr = MatSetSizes(A,M,N,M,N); CHKERRQ(ierr); ierr = MatSetType(A, MATSEQAIJ); CHKERRQ(ierr); [...lots of MatSetValue() calls...] ierr = MatAssemblyBegin(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); ierr = MatAssemblyEnd(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); The routines MatAssemblyBegin and MatAssemblyEnd should be called after completing all calls to MatSetValues().

1 0

Re: how do you access the nonzero elements in SeqAIJ matrices
by Barry Smith 04 Aug '08

04 Aug '08

>>> >>> >>> >> Make sure this is a MPI_SeqAIJ matrix. > I thought that this would work with any seqaij matrix. does it have to > be MPISEQAIJ or can it just be MATSEQAIJ? >> Sorry, my mistake. Yes a MATSEQAIJ matrix. Barry >> It is only valid to access this data after you have put the values >> into >> the matrix and called MatAssemblyBegin() and MatAssemblyEnd(). > my code looks like: > ierr = MatCreate(PETSC_COMM_WORLD,&A); CHKERRQ(ierr); > ierr = MatSetSizes(A,M,N,M,N); CHKERRQ(ierr); > ierr = MatSetType(A, MATSEQAIJ); CHKERRQ(ierr); > ierr = MatAssemblyBegin(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); > [...lots of MatSetValue() calls...] > ierr = MatAssemblyEnd(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); > >> If you are creating a matrix from i, j, and a look at >> MatCreateSeqAIJWithArrays() > I would like to extract the data from already assembled matrix (to > copy > to a GPU actually). Ok, what is the code where you access the pointers? Barry > > > Ahmed >> >> Barry >> >> >>> However when accessing directly I get different values for ptr and >>> SIGSEGV when accessing val or ind. >>> >>> also I get a bogus number for a->nz (134630032 instead of 21) >>> >>> Can someone please explain when I am doing wrong? >>> >>> Ahmed >>> >

1 0

Re: how do you access the nonzero elements in SeqAIJ matrices
by Ahmed El Zein 04 Aug '08

04 Aug '08

On Mon, 2008-08-04 at 04:17 -0500, Matthew Knepley wrote: > On Mon, Aug 4, 2008 at 3:14 AM, Ahmed El Zein <ahmed(a)azein.com> wrote: > > I am working on a project where I would like to copy a sparse matrix in > > CSR format. > > > > I have tried MatGetRow() which works OK but I would really like to get > > pointers to the 3 arrays directly. > > > > I also tried MatGetRowIJ() which allows me to get the i and j arrays but > > I can't see how to access the nonzero elements. > > You can use MatGetArray(). Thanks. The man page states that: The result of this routine is dependent on the underlying matrix data structure, and may not even work for certain matrix types. How do I find out which matrix types support it? and is there a method that works across all matrix types? Thanks, Ahmed > > Matt > > > and finally I attempted to access the arrays directly like this: > > Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; > > MatScalar *val = a->a; > > PetscInt *ptr = a->i; > > PetscInt *ind = a->j; > > > > However when accessing directly I get different values for ptr and > > SIGSEGV when accessing val or ind. > > > > also I get a bogus number for a->nz (134630032 instead of 21) What is the correct way to get the number of nonzeros? > > > > Can someone please explain when I am doing wrong? > > > > Ahmed > > > > > > >

1 0

Re: how do you access the nonzero elements in SeqAIJ matrices
by Ahmed El Zein 04 Aug '08

04 Aug '08

On Mon, 2008-08-04 at 07:57 -0500, Barry Smith wrote: > On Aug 4, 2008, at 3:14 AM, Ahmed El Zein wrote: > > > I am working on a project where I would like to copy a sparse matrix > > in > > CSR format. > > > > I have tried MatGetRow() which works OK but I would really like to get > > pointers to the 3 arrays directly. > > > > I also tried MatGetRowIJ() which allows me to get the i and j arrays > > but > > I can't see how to access the nonzero elements. > > > > and finally I attempted to access the arrays directly like this: > > Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; > > MatScalar *val = a->a; > > PetscInt *ptr = a->i; > > PetscInt *ind = a->j; > > > Make sure this is a MPI_SeqAIJ matrix. I thought that this would work with any seqaij matrix. does it have to be MPISEQAIJ or can it just be MATSEQAIJ? > > It is only valid to access this data after you have put the values > into > the matrix and called MatAssemblyBegin() and MatAssemblyEnd(). my code looks like: ierr = MatCreate(PETSC_COMM_WORLD,&A); CHKERRQ(ierr); ierr = MatSetSizes(A,M,N,M,N); CHKERRQ(ierr); ierr = MatSetType(A, MATSEQAIJ); CHKERRQ(ierr); ierr = MatAssemblyBegin(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); [...lots of MatSetValue() calls...] ierr = MatAssemblyEnd(A, MAT_FINAL_ASSEMBLY); CHKERRQ(ierr); > If you are creating a matrix from i, j, and a look at > MatCreateSeqAIJWithArrays() I would like to extract the data from already assembled matrix (to copy to a GPU actually). Ahmed > > Barry > > > > However when accessing directly I get different values for ptr and > > SIGSEGV when accessing val or ind. > > > > also I get a bogus number for a->nz (134630032 instead of 21) > > > > Can someone please explain when I am doing wrong? > > > > Ahmed > >

1 0

Re: how do you access the nonzero elements in SeqAIJ matrices
by Barry Smith 04 Aug '08

04 Aug '08

On Aug 4, 2008, at 3:14 AM, Ahmed El Zein wrote: > I am working on a project where I would like to copy a sparse matrix > in > CSR format. > > I have tried MatGetRow() which works OK but I would really like to get > pointers to the 3 arrays directly. > > I also tried MatGetRowIJ() which allows me to get the i and j arrays > but > I can't see how to access the nonzero elements. > > and finally I attempted to access the arrays directly like this: > Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; > MatScalar *val = a->a; > PetscInt *ptr = a->i; > PetscInt *ind = a->j; > Make sure this is a MPI_SeqAIJ matrix. It is only valid to access this data after you have put the values into the matrix and called MatAssemblyBegin() and MatAssemblyEnd(). If you are creating a matrix from i, j, and a look at MatCreateSeqAIJWithArrays() Barry > However when accessing directly I get different values for ptr and > SIGSEGV when accessing val or ind. > > also I get a bogus number for a->nz (134630032 instead of 21) > > Can someone please explain when I am doing wrong? > > Ahmed >

1 0

Re: how do you access the nonzero elements in SeqAIJ matrices
by Matthew Knepley 04 Aug '08

04 Aug '08

On Mon, Aug 4, 2008 at 3:14 AM, Ahmed El Zein <ahmed(a)azein.com> wrote: > I am working on a project where I would like to copy a sparse matrix in > CSR format. > > I have tried MatGetRow() which works OK but I would really like to get > pointers to the 3 arrays directly. > > I also tried MatGetRowIJ() which allows me to get the i and j arrays but > I can't see how to access the nonzero elements. You can use MatGetArray(). Matt > and finally I attempted to access the arrays directly like this: > Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; > MatScalar *val = a->a; > PetscInt *ptr = a->i; > PetscInt *ind = a->j; > > However when accessing directly I get different values for ptr and > SIGSEGV when accessing val or ind. > > also I get a bogus number for a->nz (134630032 instead of 21) > > Can someone please explain when I am doing wrong? > > Ahmed > > -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener

1 0

how do you access the nonzero elements in SeqAIJ matrices
by Ahmed El Zein 04 Aug '08

04 Aug '08

I am working on a project where I would like to copy a sparse matrix in CSR format. I have tried MatGetRow() which works OK but I would really like to get pointers to the 3 arrays directly. I also tried MatGetRowIJ() which allows me to get the i and j arrays but I can't see how to access the nonzero elements. and finally I attempted to access the arrays directly like this: Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; MatScalar *val = a->a; PetscInt *ptr = a->i; PetscInt *ind = a->j; However when accessing directly I get different values for ptr and SIGSEGV when accessing val or ind. also I get a bogus number for a->nz (134630032 instead of 21) Can someone please explain when I am doing wrong? Ahmed

1 0

Fwd: New approach for external direct solvers in PETSc
by Barry Smith 01 Aug '08

01 Aug '08

Ok, the deed is done. I have pushed the changes to petsc-dev Please report and problems to petsc-dev or petsc-maint as soon as you see them. Barry Begin forwarded message: > From: Barry Smith <bsmith(a)mcs.anl.gov> > Date: July 25, 2008 5:02:31 PM CDT > To: petsc-dev(a)mcs.anl.gov > Subject: New approach for external direct solvers in PETSc > Reply-To: petsc-dev(a)mcs.anl.gov > > > petsc-dev users, > > I have finally completed my changes to PETSc for a new approach > to accessing external direct solvers > in PETSc like Spooles, MUMPS etc. I will be pushing it to petsc-dev > sometime the middle of next week. > If you are using the direct solvers you might consider cloning from > > http://petsc.cs.iit.edu/hg/petsc/petsc-dev-new-solvers or > > ssh://[email protected]//hg/petsc/petsc-dev-new-solvers > > and trying it out before then. Please report problem to petsc-maint(a)mcs.anl.gov > as you hit them. > > > Barry > > From the changes file > > The "parallel direct solver" matrix types like > MATAIJSPOOLES are ALL gone. Now you use -pc_factor_mat_solver_package > spooles etc or PCFactorSetMatSolverPackage() or if working directly > with > matrices, MatGetFactor(A,MATSPOOLES,...) >

1 0