[petsc-users] Writing VTK output for vectors derived from DM Global vector

[Garnet Vaz]

Hi Matt,

I can just as easily write the real and imaginary parts into two components.
This would be more flexible as well. I do not know yet if I can select only
the real/imaginary parts of the solution in VisIT.

Thanks.

-
Garnet


On Fri, Sep 27, 2013 at 10:20 AM, Garnet Vaz <garnet.vaz at gmail.com> wrote:

> Hi Matt,
>
> I did run into some documentation that said that complex output
> should use complex instead of scalars and have
> Real[0]
> Imag[0]
> ...
> Real[N]
> Imag[N].
>
>
> I will try to find the documentation.
>
> -
> Garnet
>
>
> On Fri, Sep 27, 2013 at 10:14 AM, Matthew Knepley <knepley at gmail.com>wrote:
>
>> On Fri, Sep 27, 2013 at 10:12 AM, Garnet Vaz <garnet.vaz at gmail.com>wrote:
>>
>>> Hi,
>>>
>>> Does the vtk output write complex vectors? I wrote an example
>>> to test the usage but only the real part of the vector is written
>>> to the file. Is complex output supported?
>>>
>>
>> I had never considered that since I did not know how VTK would plot it.
>> Does VTK support it?
>> If so, I can change the code to output what you want.
>>
>>   Thanks,
>>
>>      Matt
>>
>>
>>> Thanks.
>>> -
>>> Garnet
>>>
>>>
>>> On Tue, Sep 24, 2013 at 6:44 AM, Garnet Vaz <garnet.vaz at gmail.com>wrote:
>>>
>>>> Hi Matt,
>>>>
>>>> Yes. Thanks. I got it to work. The shell on emacs had not
>>>> updated $PETSC_DIR and was using 3.4.2.
>>>>
>>>> -
>>>> Garnet
>>>>
>>>>
>>>> On Tue, Sep 24, 2013 at 6:26 AM, Matthew Knepley <knepley at gmail.com>wrote:
>>>>
>>>>> On Tue, Sep 24, 2013 at 6:16 AM, Garnet Vaz <garnet.vaz at gmail.com>wrote:
>>>>>
>>>>>> Hi Matt,
>>>>>>
>>>>>> I did pull yesterday. My configure.log file has
>>>>>>           Defined "VERSION_GIT" to
>>>>>> ""b4e904d1358c92c43842a940440f995d2117e40b""
>>>>>>           Defined "VERSION_DATE_GIT" to ""2013-09-23 12:36:08 +0100""
>>>>>>
>>>>>> Is the change added after this?
>>>>>>
>>>>>
>>>>> I see it here:
>>>>>
>>>>>
>>>>> https://bitbucket.org/petsc/petsc/src/33dd83b134909f5b4acb842818b1e0de2db2f4eb/include/petscdm.h?at=master#cl-35
>>>>>
>>>>> and it has been there since this commit in July:
>>>>>
>>>>>
>>>>> https://bitbucket.org/petsc/petsc/commits/38221697d93bc7c9a6149445938a5de22e85bc83
>>>>>
>>>>>       Matt
>>>>>
>>>>>
>>>>>>  -
>>>>>> Garnet
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Tue, Sep 24, 2013 at 6:08 AM, Matthew Knepley <knepley at gmail.com>wrote:
>>>>>>
>>>>>>> On Tue, Sep 24, 2013 at 5:54 AM, Garnet Vaz <garnet.vaz at gmail.com>wrote:
>>>>>>>
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> Thanks Jed. I will set the field names before I write them.
>>>>>>>> I need to create a clone and then change the default section
>>>>>>>> to have dof to 1 before I write them. I am not able to use clone
>>>>>>>> with
>>>>>>>>
>>>>>>>> DMClone(originaldm,&newdm);
>>>>>>>>
>>>>>>>> The error I get when compiling is
>>>>>>>> "'DMClone' was not declared in this scope".
>>>>>>>>
>>>>>>>> I do have #include<petscdmplex.h> which includes the
>>>>>>>> petscdm.h header. Other than clone I am able to use the
>>>>>>>> other functions.
>>>>>>>>
>>>>>>>
>>>>>>> Have you updated? I think I forgot the declaration initially.
>>>>>>>
>>>>>>>   Matt
>>>>>>>
>>>>>>>
>>>>>>>> Thanks.
>>>>>>>> -
>>>>>>>> Garnet
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, Sep 23, 2013 at 3:30 PM, Jed Brown <jedbrown at mcs.anl.gov>wrote:
>>>>>>>>
>>>>>>>>> Garnet Vaz <garnet.vaz at gmail.com> writes:
>>>>>>>>>
>>>>>>>>> > Hi,
>>>>>>>>> >
>>>>>>>>> > I would like to write out my solution to a VTK file where I use
>>>>>>>>> unstructured
>>>>>>>>> > grids in DMPlex.
>>>>>>>>> >
>>>>>>>>> > The solution corresponds to  harmonic frequencies (~ 5-10
>>>>>>>>> distinct) which
>>>>>>>>> > are
>>>>>>>>> > currently saved into a Vec* object. These vectors were created by
>>>>>>>>> > duplicating the
>>>>>>>>> > global vector for the DM object. (I know that they correspond to
>>>>>>>>> the right
>>>>>>>>> > cell.)
>>>>>>>>> >
>>>>>>>>> > Is it possible to have one file which would contain the mesh
>>>>>>>>> from the DM
>>>>>>>>> > object and all the Vec* objects?
>>>>>>>>> > Would I have to clone the original DM and use
>>>>>>>>> >
>>>>>>>>> http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Viewer/PetscViewerVTKAddField.html
>>>>>>>>> > to include all the vectors?
>>>>>>>>>
>>>>>>>>> This is a private function.
>>>>>>>>>
>>>>>>>>> > In the above function does
>>>>>>>>> > PetscViewerVTKWriteFunction
>>>>>>>>> > just need to be
>>>>>>>>> >
>>>>>>>>> > myWriteFunc(vec,viewer)
>>>>>>>>> > {
>>>>>>>>> > VecView(vec,viewer)
>>>>>>>>> > }
>>>>>>>>> > if I do not need to make any changes to the vector?
>>>>>>>>>
>>>>>>>>> This should work, but you may want to use PetscObjectSetName() so
>>>>>>>>> that
>>>>>>>>> the names of the different fields are sensible.
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>> Regards,
>>>>>>>> Garnet
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> 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
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Regards,
>>>>>> Garnet
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> 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
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Regards,
>>>> Garnet
>>>>
>>>
>>>
>>>
>>> --
>>> Regards,
>>> Garnet
>>>
>>
>>
>>
>> --
>> 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
>>
>
>
>
> --
> Regards,
> Garnet
>



-- 
Regards,
Garnet
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20130927/acf93ff9/attachment.html>