Adjust default vs vp values #2189
Conversation
| a signaling Nan. This ensures every plugin that creates seismic outputs needs | ||
| to use a material model that actually fills those outputs. | ||
| <br> | ||
| (Rene Gassmoeller, 2018/04/27) |
There was a problem hiding this comment.
Would it be worth adding this documentation to the seismic_anomalies post-processor description? Alternatively, is there already an assert that produces a similar error message when the values are filled with Nan?
gassmoeller
force-pushed
the
adjust_default_vs_vp_values
branch
2 times, most recently
from
da3a0f0
to
3e35b91
Compare
| (out.template get_additional_output<MaterialModel::SeismicAdditionalOutputs<dim> >() != 0); | ||
|
|
||
| AssertThrow(material_model_provides_seismic_output, | ||
| ExcMessage("You requested the 'Seismic' postprocessor, " |
There was a problem hiding this comment.
please correct the name of the postprocessor
| (out.template get_additional_output<MaterialModel::SeismicAdditionalOutputs<dim> >() != 0); | ||
|
|
||
| AssertThrow(material_model_provides_seismic_output, | ||
| ExcMessage("You requested the 'Seismic' postprocessor, " |
source/postprocess/depth_average.cc
Outdated
| variables.push_back("Vp"); | ||
| if (std::find( output_variables.begin(), output_variables.end(), "Vs") != output_variables.end()) | ||
| { | ||
| AssertThrow(material_model_provides_seismic_output, |
There was a problem hiding this comment.
Can you move this assert above line 377 and combine it with the one below?
gassmoeller
force-pushed
the
adjust_default_vs_vp_values
branch
from
3e35b91
to
64fdba1
Compare
|
I addressed the comments and updated the test results. The one issue here is that the column numbers in the depth_average output files will change if no seismic velocities are supplied by the material model. This can break postprocessing scripts (only for columns viscosity and vertical heat flux). I do not see a way to prevent this though. |
| @@ -0,0 +1,5 @@ | |||
| Changed: The default value for AdditionalSeismicOutputs was changed from -1 to | |||
| a signaling Nan. This ensures every plugin that creates seismic outputs needs | |||
| to use a material model that actually fills those outputs. | |||
There was a problem hiding this comment.
If there's no way around changing the change in column number in the depth average file, that should be documented here too.
Also remove seismic velocities from depth average if not provided
gassmoeller
force-pushed
the
adjust_default_vs_vp_values
branch
from
a09fd96
to
6c291cd
Compare
|
I ran the check.sh script in benchmarks and cookbooks, and checked that the seismic velocity columns are not referenced in the manual or somewhere else. Good to go? |
Closes #1562. Lets see how many tests break.