How to put mapdl plots into the pyvista.plotter subplots?

[MrMechanics]

Hi
I have a question about the pyvista.plotter. I was wondering if I could take mapdl plots (mapdl.eplot, mapdl.post_processing.nodal_displacement('NORM'), mapdl.post_processing.plot_nodal_eqv_stress( ), etc...) and put them into pyvista.plotter subplots. This would make the jupyter lab notebook easier to scroll through and read when there are a lot of results I think. I assume the code will look something like this:

import pyvista as pv
from ansys.mapdl.core import launch_mapdl
mapdl = launch_mapdl()

...

meshElements = mapdl.eplot()
displacement = mapdl.post_processing.plot_nodal_displacement(component='NORM')
vonMisesStress = mapdl.post_processing.plot_nodal_eqv_stress()
plotter = pv.Plotter(shape=(1, 3))
plotter.subplot(0, 0)
plotter.add_mesh(meshElements, cpos='xy')
plotter.subplot(0, 1)
sbar_kwargs = {'color': 'black', 'title': 'Displacement (mm)', 'vertical': True, 'n_labels': 9}
plotter.add_mesh(displacement, scalar_bar_args=sbar_kwargs, cpos='xy')
plotter.subplot(0, 2)
sbar_kwargs = {'color': 'black', 'title': 'VonMises Stress [MPa]', 'vertical': True, 'n_labels': 9}
plotter.add_mesh(vonMisesStress, scalar_bar_args=sbar_kwargs, cpos='xy')
# Render all plots
plotter.show()    

Am I right in assuming this is possible, or will the mesh added to the subplot with the plotter.add_mesh command need to be of some other format?

Best regards
Mats

[germa89]

Hi @MrMechanics

Remember that all the plotting functions (which uses vtk=True) pass by the function general_plotter hence they accepts any arguments that this function accept.
So you could use the argument plotter to tell PyMAPDL to reuse an specific plotter.

Having said that, you can do something like:

from ansys.mapdl.core import launch_mapdl
import pyvista as pv
# import code...

# Run model

# Plotting 
plotter = pv.Plotter(shape=(1, 3))

plotter.subplot(0, 0)
mapdl.eplot(plotter=plotter, return_plotter=True) # 'return_plotter' is to avoid plotting yet

plotter.subplot(0, 1)
mapdl.nplot(plotter=plotter, return_plotter=True)

plotter.subplot(0, 2)
mapdl.post_processing.plot_element_stress("X", plotter=plotter, return_plotter=True)

plotter.show()

and it should work :)

[RobPasMue]

I'd recommend creating a new virtual environment and running

pip install pyansys[mapdl-all]

[MrMechanics]

Hi @RobPasMue
For some reason, right now I was not allowed to install using

python -m pip install pyansys[mapdl-all] 

But I installed using

python -m pip install ansys-mapdl-core
python -m pip install ansys-mapdl-reader
python -m pip install ansys-dpf-core
python -m pip install ansys-dpf-reader

And now the multiplot functionality works if I just run mapdl in a python script.
If I try to run it in jupyter lab however, I get this error message:

C:\Users\MATMATH\Anaconda3\envs\testenv\Lib\site-packages\ansys\mapdl\core\plotting.py:353: UserWarning: The kwargs 'off_screen', 'notebook' and 'theme' are ignored when using 'plotter' kwarg.
  warn(
C:\Users\MATMATH\Anaconda3\envs\testenv\Lib\site-packages\pyvista\jupyter\notebook.py:58: UserWarning: Failed to use notebook backend: 

No module named 'trame'

Falling back to a static output.
  warnings.warn(

Now I installed pyansys first, and then installed the other packages I need (matplotlib, jupyter, jupyterlab, ipyvtklink, pyvista)? Earlier I just assumed one should always install the conda packages first, and then the pip install packages last. When I did that earlier I get interactive 3D plots in jupyter lab, but multiplot functionality doesn't work at all then.

Best regards,
Mats

[RobPasMue]

@germa89 is multiplot available in the latest PyAnsys metapackage release?

[germa89]

@RobPasMue I think it is not in the latest release, that is why I asked @MrMechanics to update PyMAPDL.

[RobPasMue]

In that case, I'd recommend doing the following:

python -m pip install pyansys[mapdl-all] --force-reinstall
python -m pip install ansys-mapdl-core==X.Y.Z

Where X.Y.Z stands for the desired version of PyMAPDL that you are interested on.