check_statistics_nodal_consistency.py File Reference

Cross-check derived statistics VTK output against the convergence-history CSV. More...

Go to the source code of this file.

Namespaces
namespace	check_statistics_nodal_consistency

Functions
	check_statistics_nodal_consistency.require_numpy ()
	Import NumPy, which this checker needs to read the binary VTK payload.
	check_statistics_nodal_consistency.read_vts_point_arrays (path)
	Read appended raw Float64 point arrays from a PICurv .vts.
	check_statistics_nodal_consistency.read_checkpoint_periodicity (run_dir)
	Read the periodicity the solver recorded in any committed checkpoint.
	check_statistics_nodal_consistency.newest_window_vts (viz_dir, window)
	Locate the highest-step derived statistics file for one window.
	check_statistics_nodal_consistency.check_periodic_wrap (arrays, nodes, periodic)
	Verify every derived array wraps across each periodic layout boundary.
	check_statistics_nodal_consistency.main (argv=None)
	Run the consistency check.

Variables
str	check_statistics_nodal_consistency.USAGE = "usage: check_statistics_nodal_consistency.py RUN_DIR VIZ_SUBDIR WINDOW_NAME"

Detailed Description

Cross-check derived statistics VTK output against the convergence-history CSV.

The two are produced by paths that share no code: the CSV mean comes from PicurvWindowSpatialMean, which reduces the cell-centred accumulators over the resolved spatial target, while the VTK field comes from ComputeNodalAverage interpolating the derived staging buffer onto grid nodes. Agreement between them is therefore a real check rather than a restatement.

It catches the specific failure this check was written for: an interior-only producer leaving structural zeros on the layout boundary, which halves every boundary node and biases a whole-domain mean without any other symptom.

check_statistics_nodal_consistency.py RUN_DIR VIZ_SUBDIR WINDOW_NAME

Definition in file check_statistics_nodal_consistency.py.