voiage.methods.perspective_diagnostics.evop_convergence_profile

evop_convergence_profile

evop_convergence_profile([positional or keyword] tensor: NetBenefitTensor = None, [keyword-only] choose_under: str = None, [keyword-only] evaluate_under: str = None, [keyword-only] draw_counts: Iterable[int] = None, [keyword-only] repeats: int = 20, [keyword-only] decision_rule: DecisionRule | str = DecisionRule.EXPECTED_VALUE, [keyword-only] selection_tie_policy: TiePolicy | str = TiePolicy.SPLIT, [keyword-only] seed: int | None = 0, [keyword-only] atol: float = 1e-12) -> list[EVoPConvergenceRow]

Summarise EVoP stability across random draw subsets of increasing size.

Parameters:

• tensor NetBenefitTensor
• choose_under str
• evaluate_under str
• draw_counts Iterable[int]
• repeats int (default: 20)
• decision_rule DecisionRule | str (default: DecisionRule.EXPECTED_VALUE)
• selection_tie_policy TiePolicy | str (default: TiePolicy.SPLIT)
• seed int | None (default: 0)
• atol float (default: 1e-12)

Returns: list[EVoPConvergenceRow]