tabensemb.trainer.Trainer.plot_partial_err#
method
- Trainer.plot_partial_err(program: str, model_name: str, feature, thres=0.8, ax=None, clr: Iterable | None = None, figure_kwargs: Dict | None = None, scatter_kwargs: Dict | None = None, hist_kwargs: Dict | None = None, savefig_kwargs: Dict | None = None, save_show_close: bool = True) Axes[source]#
Calculate prediction absolute errors on the testing dataset, and plot histograms of high-error samples and low-error samples respectively for a single feature.
- Parameters:
- program
The selected model base.
- model_name
The selected model in the model base.
- feature
The selected feature.
- thres
The absolute error threshold to identify high-error samples and low-error samples.
- ax
matplotlib.axes.Axes- clr
A seaborn color palette or an Iterable of colors. For example seaborn.color_palette(“deep”).
- figure_kwargs
Arguments for
plt.figure.- scatter_kwargs
Arguments for
ax.scatter()- hist_kwargs
Arguments for
ax.hist()- savefig_kwargs
Arguments for
plt.savefig- save_show_close
Whether to save, show (in the notebook), and close the figure if
axis not given.
- Returns:
- matplotlib.axes.Axes