I/O and Visualization

cartan-io writes cartan meshes and DEC fields to VTK XML and Blender vertex caches. It is the bridge from a running solver to ParaView, PyVista, and Blender.

Writers

Module	Output	Use
`vtu`	`.vtu` UnstructuredGrid	dimension-generic mesh with point and cell data
`vtp`	`.vtp` PolyData	surface fields and director glyphs
`pvd`	`.pvd` collection	time series of frames indexed by simulation time
`diagnostics`	`.csv`	scalar diagnostics per frame
`blender`	`.obj` and `.mdd`	base mesh and vertex-cache animation
`feec_fields`	per-cell arrays	cochain reconstruction at cell barycenters
`run`	run directory	`RunWriter` orchestrates the writers above

Run Directory

RunWriter records one simulation frame at a time. Each push_frame writes a .vtu, appends a diagnostics row, records an MDD frame, and on the first call writes the Blender base mesh.

use cartan_io::run::RunWriter;
use std::path::Path;
 
let mut writer = RunWriter::new(Path::new("out/run"))?;
writer.push_frame(time, &complex, &coords, &b, &e, energy, residual)?;
writer.finish()?;

The frame layout is:

out/run/
  frames.pvd
  frame_0000.vtu ...
  diagnostics.csv
  blender/base.obj
  blender/motion.mdd

Dashboard

cartan-viz, a Python package built on PyVista and matplotlib, reads a run directory and renders a dashboard video: a 3D field view on top and rolling diagnostic plots on the bottom. Frames are synchronized by simulation time.

cargo run -p cartan-maxwell --example maxwell_record -- out/run
cd cartan-viz && uv run cartan-viz render out/run --quantity energy

cartan-viz is a uv package and is not published to crates.io.