Skip to content

voiage.methods.preference.PreferenceHeterogeneityResult

Structured preference heterogeneity result.

analysis_id : str Stable analysis identifier for the preference analysis. decision_problem_id : str Stable decision-problem identifier for the preference analysis. value : float Weighted switching value relative to the reference preference profile. individualized_care_value : float Weighted value of tailoring strategy choice to each preference profile. preference_profile_ids : list[str] Preference-profile identifiers in analysis order. preference_profile_labels : list[str] Human-readable preference-profile labels. strategy_names : list[str] Strategy labels in analysis order. expected_net_benefits : numpy.ndarray Expected net benefit with shape (n_profiles, n_strategies). optimal_strategy_indices : numpy.ndarray Preference-profile-specific optimal strategy index. optimal_strategy_names : list[str] Preference-profile-specific optimal strategy name. optimal_expected_net_benefits : numpy.ndarray Expected net benefit of each preference-profile-specific optimum. regret_matrix : numpy.ndarray Matrix where row i and column j is regret in profile i when using the strategy optimal under profile j. switching_values : numpy.ndarray Regret avoided by switching away from the reference profile’s strategy toward each profile’s own optimal strategy. consensus_strategy_index : int Strategy maximizing weighted expected net benefit across profiles. consensus_strategy_name : str Name of the consensus strategy. consensus_weighted_expected_net_benefit : float Weighted expected net benefit of the consensus strategy. individualized_care_strategy_index : int Strategy maximizing weighted expected net benefit when tailoring by preference profile. individualized_care_strategy_name : str Name of the individualized-care strategy summary. individualized_care_weighted_expected_net_benefit : float Weighted expected net benefit of the individualized-care strategy. robust_strategy_index : int Strategy maximizing the minimum expected net benefit across profiles. robust_strategy_name : str Name of the robust maximin strategy. pareto_strategy_indices : list[int] Strategy indices not dominated across profiles. pareto_strategy_names : list[str] Strategy names not dominated across profiles. preference_profile_weights : numpy.ndarray Normalized preference-profile weights used for consensus and value summaries. reference_preference_profile : str Preference profile used as the reference decision rule. method_maturity : str Maturity label for the fixture-backed preference surface. diagnostics : dict[str, object] Deterministic diagnostics for downstream reporting.

to_dict([positional or keyword] self: None = None) -> dict[str, object]

Serialize to a JSON-compatible payload.

Parameters:

  • self

Returns: dict[str, object]