ParaView is an open-source, multi-platform data analysis and visualization application.
Caution: Although ParaView is available accross all nodes, the application works only on the login nodes.
ParaView version 5.9.0 is installed on the OMNI cluster. To use it, you need to load the module paraview
:
module load paraview
The ParaView documentation can be found here.
ParaView Desktop
ParaView provides a powerful graphical user interface (GUI) to explore and filter simulation data. It is capable of rendering images and videos with user defined color and configuration settings.
In order to use ParaView desktop on the OMNI cluster, connect to the cluster login nodes via SSH using X support, i.e. with the -X
option. Then the ParaView GUI can be launched with:
module load paraview
paraview --mesa &
The option --mesa
is required because Mesa support is necessary.
Note also that in this example, ParaView is launched in the background by appending &
. This is not required, but it allows working with the same console while the ParaView window is open. More information about foreground and background processes can be found in our Linux tutorial.
Python interface
ParaView offers a Python application programming interface (API) to automate more extensive data processing or reoccuring tasks. You can either write such a python script yourself (see ParaView’s Python API documentation) or record the sequence from a ParaView session.
Caution: When running a script, keep in mind that you share the login nodes with everyone and do not run compute-intensive tasks for longer periods of time. We reserve the right to kill long-running processes without prior warning if we find that they slow down the login nodes.
Example script
The following example script uses the ParaView Python API to generate a sphere and save an image:
from paraview.simple import *
sphere = Sphere()
rep = Show()
ColorBy(rep, ("POINTS", "vtkProcessId"))
Render()
rep.RescaleTransferFunctionToDataRange(True)
Render()
WriteImage("parasphere.png")
To run this example, save the code on the cluster as DistributedSphere.py
and call the script on one of the cluster login nodes with the following commands:
module load paraview
pvbatch --mesa DistributedSphere.py
The ParaView Python API also supports parallel execution using MPI, in the following example 4 parallel processes are used:
module load paraview
mpiexec -np 4 pvbatch --mesa DistributedSphere.py