Downstream Visualization: ParaView Streamline Tutorial & State File

[jameslehoux]

Labels: documentation, visualization, phase:0-polish
Priority: Medium / High (Crucial for User Publications)

Description

Visualising multi-billion-voxel HPC datasets presents a "Data Gravity" problem. We cannot expect users to download 50GB output files to their local laptops to view them.

Because OpenImpala is built on AMReX, it natively outputs standard AMReX plotfiles. ParaView (and VisIt) already support these plotfiles out-of-the-box and feature a robust client-server architecture (where the heavy rendering happens on the HPC node and streams to the user's local GUI).

We do not need to write any new visualization code for this tier. Instead, we must provide pristine documentation showing users exactly how to load OpenImpala outputs and generate publication-quality 3D renders of their flux fields.

The Standard Pipeline to Document

The documentation must walk users through this exact filter pipeline:

1. Open AMReX plotfile in ParaView.
2. Apply Cell Data to Point Data filter (to smooth the finite difference voxels).
3. Apply Threshold filter (to hide the solid phase and isolate the pore network).
4. Apply StreamTracer filter using the computed flux_x/y/z vector field.
5. Apply Tube filter (to give the streamlines 3D volume for rendering).

Acceptance Criteria

• Add a dedicated user-guide/visualization-paraview.md page to the Sphinx documentation.
• Provide step-by-step screenshots of the pipeline described above.
• State File Provision: Create and commit an openimpala_standard_render.pvsm (ParaView State File) to a new assets/ or examples/ directory. This will allow users to load their plotfile, load the state file, and instantly get a beautiful, pre-configured render pipeline without learning ParaView's complex UI.
• Document the steps required to connect a local ParaView GUI client to an pvserver instance running on an HPC Slurm node.