[CIG-SHORT] Pylith turn off vtk writers

[Francisco Delgado]

H Brad, thanks for the info but I figured it out how to do it, you have to
add the following for every material in the model

[pylithapp.problem.materials.material_name.output]
cell_info_fields = []
cell_data_fields = []

If you another two materials material called "mantle" and "crust", then add

*[pylithapp.problem.materials.crust.output]*
*cell_info_fields = []*
*cell_data_fields = []*
*[pylithapp.problem.materials.mantle.output]*
*cell_info_fields = []*
*cell_data_fields = []*

Now the simulations run much much faster

Another question, why should I use* ksp_rtol=ksp_atol=1.0e-12*  instead of
1e-11 or 1e-10? Again, this can save a lot of time

Francisco

On Thu, Aug 23, 2018 at 10:27 AM, Surendra Nadh Somala <surendra at iith.ac.in>
wrote:

> One other reason why vtk files might still be there would be if output
> settings for one of the materials is either missing (in which case pylith
> defaults that to VTK) or misspelt.
>
> -Surendra
>
> On Thu, 23 Aug 2018 at 7:50 PM, Francisco Delgado <fjd49 at cornell.edu>
> wrote:
>
>> I'm using the H5 output, it's much better than the vtk. I added the
>> following to my file
>>
>> [pylithapp.problem.materials.crust.output]
>> cell_info_fields = []
>> cell_data_fields = []
>>
>> but it still writes the vtk output. Is there a solution to this?
>>
>> On Thu, Aug 23, 2018 at 10:13 AM, Surendra Nadh Somala <
>> surendra at iith.ac.in> wrote:
>>
>>> Use
>>>
>>> vertex_data_fields=[]
>>> vertex_info_fields=[]
>>>
>>> Displacement is a vertex field whereas stress and strain are cell fields.
>>>
>>> But then you won’t have any output if you want neither displacements nor
>>> stress/strain outputs.
>>>
>>> Charles suggestion of using HDF5writer is what you might be looking for.
>>>
>>> -Surendra
>>>
>>> On Thu, 23 Aug 2018 at 7:10 PM, Francisco Delgado <fjd49 at cornell.edu>
>>> wrote:
>>>
>>>> Hi Brad and Charles, yes it's the VTK writer for the predicted ground
>>>> displacement of a pressurized magma chamber.
>>>>
>>>> My file looks like the following
>>>>
>>>> [pylithapp.timedependent.implicit]
>>>> output = [subdomain]
>>>> output.subdomain = pylith.meshio.OutputSolnSubset
>>>> #output.points = pylith.meshio.OutputSolnPoints
>>>>
>>>> # Surface
>>>> [pylithapp.problem.formulation.output.subdomain]
>>>> writer = pylith.meshio.DataWriterHDF5
>>>> label = face_zpos
>>>> vertex_data_fields = [displacement]
>>>> writer.filename = surface.h5
>>>>
>>>> but when I add the following lines
>>>>
>>>> cell_info_fields = []
>>>> cell_data_fields = []
>>>>
>>>> I get the following error
>>>>
>>>> * >> visco_shell.cfg:133:*
>>>>
>>>> * -- pyre.inventory(error)*
>>>>
>>>> * -- timedependent.implicit.output.outputsolnsubset.cell_data_fields <-
>>>> '[]'*
>>>>
>>>> * -- unrecognized property
>>>> 'timedependent.implicit.output.outputsolnsubset.cell_data_fields'*
>>>>
>>>> * >> visco_shell.cfg:132:*
>>>>
>>>> * -- pyre.inventory(error)*
>>>>
>>>> * -- timedependent.implicit.output.outputsolnsubset.cell_info_fields <-
>>>> '[]'*
>>>>
>>>> * -- unrecognized property
>>>> 'timedependent.implicit.output.outputsolnsubset.cell_info_fields'*
>>>>
>>>>
>>>>
>>>> *How can I fix it?*
>>>>
>>>> *Thanks*
>>>>
>>>> On Wed, Aug 22, 2018 at 10:47 PM, Brad Aagaard <baagaard at usgs.gov>
>>>> wrote:
>>>>
>>>>> For each writer set the vertex/cell info and data fields to emtpy
>>>>> arrays, like
>>>>>
>>>>> cell_info_fields = []
>>>>> cell_data_fields = []
>>>>>
>>>>> On Aug 22, 2018 5:22 PM, Francisco Delgado <fjd49 at cornell.edu> wrote:
>>>>>
>>>>> Dear Pylith developers, how can I turn off the vtk writer? They take a
>>>>> lot of time and waste a lot of hard drive space.
>>>>>
>>>>> Thanks
>>>>>
>>>>> --
>>>>> Francisco Delgado
>>>>> PhD, Geological Sciences
>>>>> Cornell University
>>>>> https://fdelgadodelapuente.wixsite.com/home
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> CIG-SHORT mailing list
>>>>> CIG-SHORT at geodynamics.org
>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Francisco Delgado
>>>> PhD, Geological Sciences
>>>> Cornell University
>>>> https://fdelgadodelapuente.wixsite.com/home
>>>> _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org
>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>> --
>>> ___________
>>> *Surendra Nadh Somala*
>>> Assistant Professor
>>> Department of Civil Engineering (Block-C, 112/D)
>>> Indian Institute of Technology (IIT) Hyderabad
>>> Kandi, Sangareddy
>>> Telangana, India - 502285
>>> http://civil.iith.ac.in/surendra/
>>>
>>
>>
>>
>> --
>> Francisco Delgado
>> PhD, Geological Sciences
>> Cornell University
>> https://fdelgadodelapuente.wixsite.com/home
>>
> --
> ___________
> *Surendra Nadh Somala*
> Assistant Professor
> Department of Civil Engineering (Block-C, 112/D)
> Indian Institute of Technology (IIT) Hyderabad
> Kandi, Sangareddy
> Telangana, India - 502285
> http://civil.iith.ac.in/surendra/
>



-- 
Francisco Delgado
PhD, Geological Sciences
Cornell University
https://fdelgadodelapuente.wixsite.com/home
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/cig-short/attachments/20180823/b7d33e08/attachment.html>