graph LR
dice_ml_counterfactual_explanations_CounterfactualExplanations["dice_ml.counterfactual_explanations.CounterfactualExplanations"]
dice_ml_diverse_counterfactuals_DiverseCounterfactuals["dice_ml.diverse_counterfactuals.DiverseCounterfactuals"]
dice_ml_diverse_counterfactuals_DiverseCounterfactuals -- "composes" --> dice_ml_counterfactual_explanations_CounterfactualExplanations
dice_ml_diverse_counterfactuals_DiverseCounterfactuals -- "delegates to" --> dice_ml_counterfactual_explanations_CounterfactualExplanations
The Explanation Output & Serialization subsystem in DiCE is responsible for the structured representation, persistence (serialization/deserialization), and presentation of generated counterfactual explanations. Its primary goal is to prepare the explanation data for various downstream uses, including display to users, storage, or integration with other analysis tools, while adhering to defined data schemas.
This class serves as the foundational data structure for a single set of counterfactual explanations. It encapsulates the explanation data (e.g., feature values, predictions), associated metadata, and provides essential functionalities for its lifecycle.
Related Classes/Methods:
dice_ml.counterfactual_explanations.CounterfactualExplanations:43-302dice_ml.counterfactual_explanations.CounterfactualExplanations:to_jsondice_ml.counterfactual_explanations.CounterfactualExplanations:from_jsondice_ml.counterfactual_explanations.CounterfactualExplanations:visualize_as_dataframedice_ml.counterfactual_explanations.CounterfactualExplanations:visualize_as_list
This class specializes in managing and presenting multiple, diverse counterfactual explanations. It acts as an aggregator and orchestrator for visualizing and handling collections of individual explanations.
Related Classes/Methods: