[petsc-users] parallel HDF5 output of DMDA data with dof>1

[Barry Smith]

   I have run your code and looked at the output in Matlab and then looked at your code. 

   It seems you expect the output to be independent of the number of MPI ranks? It will not be because you have written

>  ierr = VecGetSubVector(U0,is[0],&uField); CHKERRQ(ierr);
>   PetscObjectSetName((PetscObject) uField, "S");
>   ierr = VecView(uField,viewer); CHKERRQ(ierr);
>   ierr = VecRestoreSubVector(U0,is[0],&uField); CHKERRQ(ierr);

The subvector you create loses the DMDA context (information that the vector is associated with a 2d array sliced up among the MPI ranks) since you just taking a part of the vector out via VecGetSubVector() which only sees an IS so has no way to connect the new subvector to a DMDA of dimension 2 (with a single field) and automatically do the reordering to natural ordering when the vector is directly associated with a DMDA and then viewed to a file.

In order to get the behavior you hope for, you need to have your subvector be associated with the DMDA that comes out as the final argument to DMCreateFieldDecomposition(). There are multiple ways you can do this, none of them are particularly "clear", because it appears no one has bothered to include in the DM API a clear way to transfer vectors and parts of vectors between DMs and sub-DMs (that is a DM with the same topology as the original DM but fewer or more "fields"). 

  I would suggest using 

   DMCreateGlobalVector(daField[0], &uField);
   VecStrideGather(U0,0,uField,INSERT_VALUES);
   PetscObjectSetName((PetscObject) uField, "S");
   ierr = VecView(uField,viewer); CHKERRQ(ierr);
   DMRestoreGlobalVector(daField[0], &uField);

  For the second field you would use U0,1 instead of U0,0. 

   This will be fine for DMDA but I cannot say if it is appropriate for all types of DMs in all circumstances. 

Barry

  
  



> On Jul 15, 2021, at 10:44 AM, Matteo Semplice <matteo.semplice at uninsubria.it> wrote:
> 
> Hi.
> 
> When I write (HDF5 viewer) a vector associated to a DMDA with 1 dof, the output is independent of the number of cpus used.
> 
> However, for a DMDA with dof=2, the output seems to be correct when I run on 1 or 2 cpus, but is scrambled when I run with 4 cpus. Judging from the ranges of the data, each field gets written to the correct part, and its the data witin the field that is scrambled. Here's my MWE:
> 
> #include <petscversion.h>
> #include <petscdmda.h>
> #include <petscviewer.h>
> #include <petscsys.h>
> #include <petscviewerhdf5.h>
> 
> int main(int argc, char **argv) {
> 
>   PetscErrorCode ierr;
>   ierr = PetscInitialize(&argc,&argv,(char*)0,help); CHKERRQ(ierr);
>   PetscInt Nx=11;
>   PetscInt Ny=11;
>   PetscScalar dx = 1.0 / (Nx-1);
>   PetscScalar dy = 1.0 / (Ny-1);
>   DM dmda;
>   ierr = DMDACreate2d(PETSC_COMM_WORLD,
>                       DM_BOUNDARY_NONE,DM_BOUNDARY_NONE,
>                       DMDA_STENCIL_STAR,
>                       Nx,Ny, //global dim
>                       PETSC_DECIDE,PETSC_DECIDE, //n proc on each dim
>                       2,1, //dof, stencil width
>                       NULL, NULL, //n nodes per direction on each cpu
>                       &dmda);      CHKERRQ(ierr);
>   ierr = DMSetFromOptions(dmda); CHKERRQ(ierr);
>   ierr = DMSetUp(dmda); CHKERRQ(ierr); CHKERRQ(ierr);
>   ierr = DMDASetUniformCoordinates(dmda, 0.0, 1.0, 0.0, 1.0, 0.0, 1.0); CHKERRQ(ierr);
>   ierr = DMDASetFieldName(dmda,0,"s"); CHKERRQ(ierr);
>   ierr = DMDASetFieldName(dmda,1,"c"); CHKERRQ(ierr);
>   DMDALocalInfo daInfo;
>   ierr = DMDAGetLocalInfo(dmda,&daInfo); CHKERRQ(ierr);
>   IS *is;
>   DM *daField;
>   ierr = DMCreateFieldDecomposition(dmda,NULL, NULL, &is, &daField); CHKERRQ(ierr);
>   Vec U0;
>   ierr = DMCreateGlobalVector(dmda,&U0); CHKERRQ(ierr);
> 
>   //Initial data
>   typedef struct{ PetscScalar s,c;} data_type;
>   data_type **u;
>   ierr = DMDAVecGetArray(dmda,U0,&u); CHKERRQ(ierr);
>   for (PetscInt j=daInfo.ys; j<daInfo.ys+daInfo.ym; j++){
>     PetscScalar y = j*dy;
>     for (PetscInt i=daInfo.xs; i<daInfo.xs+daInfo.xm; i++){
>       PetscScalar x = i*dx;
>       u[j][i].s = x+2.*y;
>       u[j][i].c = 10. + 2.*x*x+y*y;
>     }
>   }
>   ierr = DMDAVecRestoreArray(dmda,U0,&u); CHKERRQ(ierr);
> 
>   PetscViewer viewer;
>   ierr = PetscViewerHDF5Open(PETSC_COMM_WORLD,"solutionSC.hdf5",FILE_MODE_WRITE,&viewer); CHKERRQ(ierr);
>   Vec uField;
>   ierr = VecGetSubVector(U0,is[0],&uField); CHKERRQ(ierr);
>   PetscObjectSetName((PetscObject) uField, "S");
>   ierr = VecView(uField,viewer); CHKERRQ(ierr);
>   ierr = VecRestoreSubVector(U0,is[0],&uField); CHKERRQ(ierr);
>   ierr = VecGetSubVector(U0,is[1],&uField); CHKERRQ(ierr);
>   PetscObjectSetName((PetscObject) uField, "C");
>   ierr = VecView(uField,viewer); CHKERRQ(ierr);
>   ierr = VecRestoreSubVector(U0,is[1],&uField); CHKERRQ(ierr);
>   ierr = PetscViewerDestroy(&viewer); CHKERRQ(ierr);
> 
>   ierr = PetscFinalize();
>   return ierr;
> }
> 
> and my xdmf file
> 
> <?xml version="1.0" ?>
> <Xdmf xmlns:xi="http://www.w3.org/2001/XInclude" Version="2.0">
>   <Domain>
>     <Grid GridType="Collection" CollectionType="Temporal">
>       <Time TimeType="List">
>         <DataItem Dimensions="1">1.0</DataItem>
>       </Time>
>       <Grid GridType="Uniform" Name="domain">
>         <Topology TopologyType="2DCoRectMesh" Dimensions="11 11"/>
>         <Geometry GeometryType="ORIGIN_DXDY">
>           <DataItem Format="XML" NumberType="Float" Dimensions="2">0.0 0.0</DataItem>
>           <DataItem Format="XML" NumberType="Float" Dimensions="2">0.1 0.1</DataItem>
>         </Geometry>
>         <Attribute Name="S" Center="Node" AttributeType="Scalar">
>           <DataItem Format="HDF" Precision="8" Dimensions="11 11">solutionSC.hdf5:/S</DataItem>
>         </Attribute>
>         <Attribute Name="C" Center="Node" AttributeType="Scalar">
>           <DataItem Format="HDF" Precision="8" Dimensions="11 11">solutionSC.hdf5:/C</DataItem>
>         </Attribute>
>       </Grid>
>     </Grid>
>   </Domain>
> </Xdmf>
> 
> Steps to reprduce: run code and open the xdmf with paraview. If the code was run with 1,2 or 3 cpus, the data are correct (except the plane xy has become the plane yz), but with 4 cpus the data are scrambled.
> 
> Does anyone have any insight?
> 
> (I am using Petsc Release Version 3.14.2, but I can compile a newer one if you think it's important.)
> 
> Best
> 
>     Matteo
>