ncem.train.GridSearchContainer.plot_best_model_by_hyperparam

GridSearchContainer.plot_best_model_by_hyperparam(graph_model_class, baseline_model_class, partition_show='test', metric_show='r_squared_linreg', partition_select='val', metric_select='r_squared_linreg', param_x='um_radius', param_hue='model', rename_levels=None, subset_hyperparameters=None, cv_mode='mean', yaxis_limit=None, xticks=None, rotate_xticks=True, figsize=(3.5, 4.0), fontsize=None, example_cellradius=10, plot_mode='lineplot', palette={'NCEM': 'C1', 'baseline': 'C0'}, color=None, save=None, suffix='best_by_hyperparam.pdf', show=True, return_axs=False)[source]

Plot best model by hyperparameter.

Parameters

  • graph_model_class (str) – Graph model class.

  • baseline_model_class (str) – Baseline model class.

  • partition_show (str) – Showing partition.

  • metric_show (str) – Showing metric.

  • partition_select (str) – Selected partition.

  • metric_select (str) – Selected metric.

  • param_x (str) – Parameter on x axis.

  • param_hue (str) – Parameter for hue.

  • rename_levels (list, optional) – Rename levels with stated logic.

  • subset_hyperparameters (list, optional) – Subset hyperparameters.

  • cv_mode (str) – Cross validation mode.

  • yaxis_limit (tuple, optional) – y axis limits.

  • xticks (list, optional) – List of x ticks.

  • rotate_xticks (bool) – Whether to rotate x ticks.

  • figsize (tuple) – Figure size.

  • fontsize (int, optional) – Font size.

  • plot_mode (str) – Plotting mode, can be boxplot, lineplot or mean_lineplot.

  • palette (dict, optional) – Palette.

  • color (str, optional) – Color.

  • save (str, optional) – Whether (if not None) and where (path as string given as save) to save plot.

  • suffix (str) – Suffix of file name to save to.

  • show (bool) – Whether to display plot.

  • return_axs (bool) – Whether to return axis objects.

  • example_cellradius (Optional[int]) –

Return type

axis if return_axs is True.