I think your approach is correct.  Do you have an example code?

--Junchao Zhang

On Tue, Dec 5, 2023 at 5:15 PM Sreeram R Venkat <srvenkat@utexas.edu> wrote:

Hi, I have a follow up question on this. 

Now, I'm trying to do a scatter and permutation of the vector. Under the same setup as the original example, here are the new Start and Finish states I want to achieve:

 Start                                Finish

Proc | Entries                 Proc | Entries

    0   |  0,...,8                     0   | 0, 12, 24

    1   |  9,...,17                   1   | 1, 13, 25

    2   |  18,...,26                 2   | 2, 14, 26

    3   |  27,...,35                 3   | 3, 15, 27

    4   |  None                     4   | 4, 16, 28

    5   |  None                     5   | 5, 17, 29

    6   |  None                     6   | 6, 18, 30

    7   |  None                     7   | 7, 19, 31

    8   |  None                     8   | 8, 20, 32

    9   |  None                     9   | 9, 21, 33

    10   |  None                   10 | 10, 22, 34

    11   |  None                   11  | 11, 23, 35

So far, I've tried to use ISCreateGeneral(), with each process giving an idx array corresponding to the indices it wants (i.e. idx on P0 looks like [0,12,24] P1 -> [1,13, 25], and so on). 

Then I use this to create the VecScatter with VecScatterCreate(x, is, y, NULL, &scatter).

However, when I try to do the scatter, I get some illegal memory access errors. 

Is there something wrong with how I define the index sets?

Thanks,

Sreeram

On Thu, Oct 5, 2023 at 12:57 PM Sreeram R Venkat <srvenkat@utexas.edu> wrote:

Thank you. This works for me.

Sreeram

On Wed, Oct 4, 2023 at 6:41 PM Junchao Zhang <junchao.zhang@gmail.com> wrote:

Hi, Sreeram,
You can try this code. Since x, y are both MPI vectors, we just need to say we want to scatter x[0:N] to y[0:N]. The 12 index sets with your example on the 12 processes would be [0..8], [9..17], [18..26], [27..35], [], ..., [].  Actually, you can do it arbitrarily, say, with 12 index sets [0..17], [18..35], .., [].  PETSc will figure out how to do the communication. 

PetscInt rstart, rend, N;

IS ix;

VecScatter vscat;

Vec y;

MPI_Comm comm;

VecType type;

PetscObjectGetComm((PetscObject)x, &comm);

VecGetType(x, &type);

VecGetSize(x, &N);

VecGetOwnershipRange(x, &rstart, &rend);

VecCreate(comm, &y);

VecSetSizes(y, PETSC_DECIDE, N);

VecSetType(y, type);

ISCreateStride(PetscObjectComm((PetscObject)x), rend - rstart, rstart, 1, &ix);

VecScatterCreate(x, ix, y, ix, &vscat);

--Junchao Zhang

On Wed, Oct 4, 2023 at 6:03 PM Sreeram R Venkat <srvenkat@utexas.edu> wrote:

Suppose I am running on 12 processors, and I have a vector "v" of size 36 partitioned over the first 4. v still uses the PETSC_COMM_WORLD, so it has a layout of (9, 9, 9, 9, 0, 0, ..., 0). Now, I would like to repartition it over all 12 processors, so that the layout becomes (3, 3, 3, ..., 3). I've been trying to use VecScatter to do this, but I'm not sure what IndexSets to use for the sender and receiver.

The result I am trying to achieve is this:

Assume the vector is v = <0, 1, 2, ..., 35> 

     Start                                Finish

Proc | Entries                 Proc | Entries

    0   |  0,...,8                     0   | 0, 1, 2

    1   |  9,...,17                   1   | 3, 4, 5

    2   |  18,...,26                 2   | 6, 7, 8

    3   |  27,...,35                 3   | 9, 10, 11

    4   |  None                     4   | 12, 13, 14

    5   |  None                     5   | 15, 16, 17

    6   |  None                     6   | 18, 19, 20

    7   |  None                     7   | 21, 22, 23

    8   |  None                     8   | 24, 25, 26

    9   |  None                     9   | 27, 28, 29

    10   |  None                   10 | 30, 31, 32

    11   |  None                   11  | 33, 34, 35

Appreciate any help you can provide on this.

Thanks,

Sreeram