[petsc-users] DMGetCoordinatesLocal and DMPlexGetCellCoordinates in PETSc > 3.18

Berend van Wachem berend.vanwachem at ovgu.de
Mon May 22 03:39:09 CDT 2023


Dear Matt,

I'm really sorry for this stupid bug!

I can confirm that setting the coordinates with both 
CellCoordinatesLocal and CoordinatesLocal works.

Best regards, Berend.

Many thanks and best regards, Berend.

On 5/17/23 23:04, Matthew Knepley wrote:
> On Wed, May 17, 2023 at 2:01 PM Berend van Wachem 
> <berend.vanwachem at ovgu.de <mailto:berend.vanwachem at ovgu.de>> wrote:
> 
>     Dear Matt,
> 
>     I tried it, but it doesn't seem to work.
>     Attached is a very small working example illustrating the problem.
>     I create a DMPlexBox Mesh, periodic in the Y direction. I then scale
>     the Y coordinates with a factor 10, and add 1.0 to it. Both
>     DMGetCoordinatesLocal and DMGetCellCoordinatesLocal.
>     Then I evaluate the coordinates with DMPlexGetCellCoordinates. Most
>     of the Y coordinates are correct, but not all of them - for
>     instance, the minimum Y coordinate is 0.0, and this should be 1.0.
> 
>     Am I doing something wrong?
> 
> 
> Quickly, I see that
> 
>    a *= 10.0 + 1.0;
> 
> is the same as
> 
>    a *= 11.0;
> 
> not multiply by 10 and add 1. I will send it back when I get everything 
> the way I want.
> 
>    Thanks,
> 
>      Matt
> 
>     Thanks and best regards,
> 
>     Berend.
> 
>     On 5/17/23 17:58, Matthew Knepley wrote:
>      > On Wed, May 17, 2023 at 11:20 AM Berend van Wachem
>     <berend.vanwachem at ovgu.de <mailto:berend.vanwachem at ovgu.de>
>     <mailto:berend.vanwachem at ovgu.de <mailto:berend.vanwachem at ovgu.de>>>
>     wrote:
>      >
>      >     Dear Matt,
>      >
>      >     Is there a way to 'redo' the DMLocalizeCoordinates() ? Or to
>     undo it?
>      >     Alternatively, can we make the calling of
>     DMLocalizeCoordinates() in the  DMPlexCreate...() routines optional?
>      >
>      >     Otherwise, we would have to copy all arrays of coordinates
>     from DMGetCoordinatesLocal() and DMGetCellCoordinatesLocal() before
>      >     scaling them.
>      >
>      >
>      > I am likely not being clear. I think all you have to do is the
>     following:
>      >
>      >    DMGetCoordinatesLocal(dm, &xl);
>      >    VecScale(xl, scale);
>      >    DMSetCoordinatesLocal(dm, xl);
>      >    DMGetCellCoordinatesLocal(dm, &xl);
>      >    VecScale(xl, scale);
>      >    DMSetCellCoordinatesLocal(dm, xl);
>      >
>      > Does this not work?
>      >
>      >    Thanks,
>      >
>      >       Matt
>      >
>      >     Best regards, Berend.
>      >
>      >     On 5/17/23 16:35, Matthew Knepley wrote:
>      >      > On Wed, May 17, 2023 at 10:21 AM Berend van Wachem
>     <berend.vanwachem at ovgu.de <mailto:berend.vanwachem at ovgu.de>
>     <mailto:berend.vanwachem at ovgu.de <mailto:berend.vanwachem at ovgu.de>>
>      >     <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de> <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de>>>> wrote:
>      >      >
>      >      >     Dear Matt,
>      >      >
>      >      >     Thanks for getting back to me so quickly.
>      >      >
>      >      >     If I scale each of the coordinates of the mesh (say, I
>     want to cube each
>      >      >     co-ordinate), and I do this for both:
>      >      >
>      >      >     DMGetCoordinatesLocal();
>      >      >     DMGetCellCoordinatesLocal();
>      >      >
>      >      >     How do I know I am not cubing one coordinate multiple
>     times?
>      >      >
>      >      >
>      >      > Good question. Right now, the only connection between the
>     two sets of coordinates is DMLocalizeCoordinates(). Since
>      >     sometimes
>      >      > people want to do non-trivial things to
>      >      > coordinates, I prefer not to push in an API for "just"
>     scaling, but I could be convinced
>      >      > the other way.
>      >      >
>      >      >    Thanks,
>      >      >
>      >      >       Matt
>      >      >
>      >      >     Thanks, Berend.
>      >      >
>      >      >     On 5/17/23 16:10, Matthew Knepley wrote:
>      >      >      > On Wed, May 17, 2023 at 10:02 AM Berend van Wachem
>      >      >      > <berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de> <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de>> <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de>
>      >     <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de>>> <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de> <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de>>
>      >      >     <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de> <mailto:berend.vanwachem at ovgu.de
>     <mailto:berend.vanwachem at ovgu.de>>>>> wrote:
>      >      >      >
>      >      >      >     Dear PETSc Team,
>      >      >      >
>      >      >      >     We are using DMPlex, and we create a mesh using
>      >      >      >
>      >      >      >     DMPlexCreateBoxMesh (.... );
>      >      >      >
>      >      >      >     and get a uniform mesh. The mesh is periodic.
>      >      >      >
>      >      >      >     We typically want to "scale" the coordinates
>     (vertices) of the mesh,
>      >      >      >     and
>      >      >      >     to achieve this, we call
>      >      >      >
>      >      >      >     DMGetCoordinatesLocal(dm, &coordinates);
>      >      >      >
>      >      >      >     and scale the entries in the Vector coordinates
>     appropriately.
>      >      >      >
>      >      >      >     and then
>      >      >      >
>      >      >      >     DMSetCoordinatesLocal(dm, coordinates);
>      >      >      >
>      >      >      >
>      >      >      >     After this, we localise the coordinates by calling
>      >      >      >
>      >      >      >     DMLocalizeCoordinates(dm);
>      >      >      >
>      >      >      >     This worked fine up to PETSc 3.18, but with
>     versions after this, the
>      >      >      >     coordinates we get from the call
>      >      >      >
>      >      >      >     DMPlexGetCellCoordinates(dm, CellID, &isDG,
>     &CoordSize,
>      >      >      >     &ArrayCoordinates, &Coordinates);
>      >      >      >
>      >      >      >     are no longer correct if the mesh is periodic.
>     A number of the
>      >      >      >     coordinates returned from calling
>     DMPlexGetCellCoordinates are wrong.
>      >      >      >
>      >      >      >     I think, this is because DMLocalizeCoordinates
>     is now automatically
>      >      >      >     called within the routine DMPlexCreateBoxMesh.
>      >      >      >
>      >      >      >     So, my question is: How should we scale the
>     coordinates from a periodic
>      >      >      >     DMPlex mesh so that they are reflected
>     correctly when calling both
>      >      >      >     DMGetCoordinatesLocal and
>     DMPlexGetCellCoordinates, with PETSc versions
>      >      >      >       >= 3.18?
>      >      >      >
>      >      >      >
>      >      >      > I think we might have to add an API function. For
>     now, when you scale
>      >      >      > the coordinates,
>      >      >      > can you scale both copies?
>      >      >      >
>      >      >      >    DMGetCoordinatesLocal()
>      >      >      >    DMGetCellCoordinatesLocal();
>      >      >      >
>      >      >      > and then set them back.
>      >      >      >
>      >      >      >    Thanks,
>      >      >      >
>      >      >      >       Matt
>      >      >      >
>      >      >      >     Many thanks, Berend.
>      >      >      >
>      >      >      > --
>      >      >      > 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
>      >      >      >
>      >      >      > https://www.cse.buffalo.edu/~knepley/
>     <https://www.cse.buffalo.edu/~knepley/>
>     <https://www.cse.buffalo.edu/~knepley/
>     <https://www.cse.buffalo.edu/~knepley/>>
>      >     <https://www.cse.buffalo.edu/~knepley/
>     <https://www.cse.buffalo.edu/~knepley/>
>     <https://www.cse.buffalo.edu/~knepley/
>     <https://www.cse.buffalo.edu/~knepley/>>>
>     <http://www.cse.buffalo.edu/~knepley/
>     <http://www.cse.buffalo.edu/~knepley/>
>      >     <http://www.cse.buffalo.edu/~knepley/
>     <http://www.cse.buffalo.edu/~knepley/>>
>      >      >     <http://www.cse.buffalo.edu/~knepley/
>     <http://www.cse.buffalo.edu/~knepley/>
>     <http://www.cse.buffalo.edu/~knepley/
>     <http://www.cse.buffalo.edu/~knepley/>>>>
>      >      >
>      >      >
>      >      >
>      >      > --
>      >      > 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
>      >      >
>      >      > https://www.cse.buffalo.edu/~knepley/
>     <https://www.cse.buffalo.edu/~knepley/>
>     <https://www.cse.buffalo.edu/~knepley/
>     <https://www.cse.buffalo.edu/~knepley/>>
>     <http://www.cse.buffalo.edu/~knepley/
>     <http://www.cse.buffalo.edu/~knepley/>
>      >     <http://www.cse.buffalo.edu/~knepley/
>     <http://www.cse.buffalo.edu/~knepley/>>>
>      >
>      >
>      >
>      > --
>      > 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
>      >
>      > https://www.cse.buffalo.edu/~knepley/
>     <https://www.cse.buffalo.edu/~knepley/>
>     <http://www.cse.buffalo.edu/~knepley/
>     <http://www.cse.buffalo.edu/~knepley/>>
> 
> 
> 
> -- 
> 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
> 
> https://www.cse.buffalo.edu/~knepley/ <http://www.cse.buffalo.edu/~knepley/>


More information about the petsc-users mailing list