zoomy_amrex.run_case

zoomy_amrex.run_case#

zoomy_amrex.run_case — importable structured-grid entry point + VTK output.

Runs a model on a structured AMReX grid and writes a ParaView-ready VTK series (.pvd + per-snapshot/-level .vtu). One call handles both solver families:

  • standard hyperbolic (SWE / SME) via the AmrCore driver (main.cpp);

  • non-hydrostatic VAM / ML-VAM via the Chorin pressure-projection driver (chorin_main.cpp / chorin_amr.cpp) — auto-selected when the model exposes chorin_split.

The case scripts under thesis/cases/amrex are thin: build a model (+IC/BC), give a settings dict (domain / n_cells / dimension / time_end / amr / …), call run_case → a .pvd for ParaView. This is the same entry point the GUI/CLI server adapter uses. AMReX emits VTK only; a store-HDF5 (for the GUI viz) is obtained from the shared zoomy_prepost.vtk_to_hdf5 on this VTK if ever needed.

Module Contents#

Functions#

gmsh_to_rasters

Project a gmsh .msh + its $NodeData (bed/depth) onto the structured grid — generalises malpasset.prepare_rasters (REQ-89). Returns dem/release.

run_case

Run model on a structured AMReX grid built from settings; write a ParaView .pvd (+ .vtu series) and return its path.

API#

zoomy_amrex.run_case.gmsh_to_rasters(msh_path, out_dir, geom)#

Project a gmsh .msh + its $NodeData (bed/depth) onto the structured grid — generalises malpasset.prepare_rasters (REQ-89). Returns dem/release.

zoomy_amrex.run_case.run_case(model, settings, output_dir, on_progress=None)#

Run model on a structured AMReX grid built from settings; write a ParaView .pvd (+ .vtu series) and return its path.

settings: dimension (1|2), domain ([x0,x1] or [x0,x1,y0,y1]), n_cells (nx or [nx,ny]), time_end, cfl, spatial_order, output_snapshots, max_level (AMR; 0 = single level), optional mesh_msh (gmsh + $NodeData IC), and for Chorin models optional inflow / pin_all / params. The solver family (standard vs Chorin) is auto-selected from the model.

Contents