utils.load_run_data#

utils.load_run_data(path: str | Path) → DataFrame[source]#

Load HDF5 run data as DataFrame for plotting.

Loads time-series data (residuals) along with metadata as columns. This format is optimized for seaborn plotting with hue parameter.

Parameters:

pathstr or Path: Path to HDF5 file.

Returns:

pd.DataFrame: DataFrame with columns: - iteration: Iteration number (0, 1, 2, …) - residual: Residual value at each iteration - u_residual, v_residual, continuity_residual: Component residuals (if available) - Re: Reynolds number (from metadata) - converged: Whether solver converged (from metadata) - All other metadata fields as additional columns

Examples

>>> df = load_run_data('run.h5')
>>> df.head()
   iteration  residual  Re  converged  mesh_path  ...
0          0  1.000000  100       True  fine.msh  ...
1          1  0.500000  100       True  fine.msh  ...

>>> # Plot multiple runs
>>> import seaborn as sns
>>> df1 = load_run_data('run1.h5').assign(run='Run 1')
>>> df2 = load_run_data('run2.h5').assign(run='Run 2')
>>> df = pd.concat([df1, df2])
>>> sns.lineplot(data=df, x='iteration', y='residual', hue='run')