summarize_plot_visualize_exp

hybrid_learning.experimentation.fuzzy_exp.fuzzy_exp_vis.summarize_plot_visualize_exp(experiment_root, model_key, split, logic_type, formula_dirs=None, per_image_stats_csv='per_image_stats.csv', force_reevaluate=True, visualize=None, num_interesting_samples=15, verbose=False, only_filenames=None)

Concatenation of per image stats gathering, stats summary, and optional visualization of formula body masks. Summary of hybrid_learning.experimentation.fuzzy_exp.fuzzy_exp_eval.gather_exp_stats(), hybrid_learning.experimentation.fuzzy_exp.fuzzy_exp_eval.summarize_exp_stats(), plot_summaries(), and visualize_most_interesting_samples() with option to loop over several formulas. Formula body values are reevaluated.

Parameters

• per_image_stats_csv (str) – file path to a CSV file for caching the per image statistics (will contain information for one image per row); file path relative to the respective logic_type/metrics results directory

• visualize (Optional[Sequence[Literal['best', 'worst']]]) – list of different arguments for key best in visualize_most_interesting_samples(); set to None or an empty sequence to disable visualization of example images

• num_interesting_samples (int) – see visualize_most_interesting_samples()

• force_reevaluate (bool) – see hybrid_learning.experimentation.fuzzy_exp.fuzzy_exp_eval.gather_exp_stats()

• only_filenames (Optional[List[str]]) – only take into account the files with given names in analysis and visualization

• experiment_root (str) –

• model_key (str) –

• split (str) –

• logic_type (str) –

• formula_dirs (Optional[list]) –

• verbose (bool) –