GSoC 2025 - Week 9

This week, I got started on the density component visualization. I initiated discussion #914 outlining the approaches and started implementing them.

I made the following progress:

• PR #916 (open)
  • Fixed the density entity to work with the style interface (for both API and GUI)
  • Added GUI support for density entities to show up in the 3D viewport panel
  • Added annotations support for density entities as well
  • Fixed a few other bugs encountered during this process
• PR #917 (open)
  • Refactored density entity to use MDAnalysis’s GridDataFormats library instead of mrcfile. GridDataFormats supports MRC files as well
  • Added a new Grids class that supports all the grid formats of the GridDataFormats library
    • .dx, .plt, .ccp4, .mrc, .map, .pickle, (and compressed variants like .dx.gz, .ccp4.gz, .ccp4.bz2, .mrc.gz, .mrc.bz2, .map.gz, .map.bz2)
  • Added support for different grid data types based on what openvdb supports
  • Support both older pyopenvdb (< blender 4.4.0) and newer openvdb library for generating and loading .vdb files in Blender
  • Min, Max support for thresholds in node socket inputs based on grid values

Here are some learning from this week:

• Blender’s openvdb support uses nanogrid to check and warn for leaks. Any vdb grids created must be ensured they are cleaned up to avoid memory errors, especially in pytests
• Blender seems to have some wierd caching of contents loaded from vdb files. Overwriting a vdb file that is already loaded and re-loading it as a new object will update the old object as well
• Though GridDataFormats supports a lot of commonly used grids in molecular simulations, most of them are limited to orthorhombic unitcells

Next week, I will fix any issues that come up during the review of the open PRs and wrap up work on densities by adding a new style node for ISO surfaces that allows different colors for positive and negative ISO values and also slicing along all the major axes.