xlnstorch.viz.plot_precision_heatmap_grid

xlnstorch.viz.plot_precision_heatmap_grid(results, *, figsize=None, cmap='viridis')

Create a grid of heatmaps showing error patterns at different precisions.

Parameters:

• results (Dict[int, Dict]) – Results from precision_sweep_analysis.

• figsize (Tuple[int, int], optional) – Size of the figure to create.

• cmap (str, optional) – Colormap to use for heatmaps. Default is ‘viridis’.

Returns:

The figure and list of axes containing the heatmaps.

Return type:

Tuple[matplotlib.figure.Figure, List[matplotlib.axes.Axes]]

Raises:

ImportError – If matplotlib is not installed, an ImportError is raised.