gcubed.projections._ev_calculation

Pure regional lifetime equivalent variation with the example's terminal tail.

@dataclass(frozen=True)

class EVResult:

Percentage EV, utility diagnostics and optional anchored monetary amounts.

EVResult( ev_percent: float, phi: float, reference_utility: float, scenario_utility: float, terminal_tail_weight_share: float, ev_level: float | None = None, reference_wealth: float | None = None, counterfactual_wealth: float | None = None)

ev_percent: float

phi: float

reference_utility: float

scenario_utility: float

terminal_tail_weight_share: float

ev_level: float | None = None

reference_wealth: float | None = None

counterfactual_wealth: float | None = None

class EVInputError(builtins.ValueError):

An input does not satisfy the regional EV calculation contract.

Inherited Members

builtins.BaseException: BaseException; with_traceback; add_note; args

def calculate_ev( reference_path: pandas.DataFrame, scenario_path: pandas.DataFrame, *, weights: pandas.Series, sigma: float, time_preference: float, growth: float, include_monetary: bool = False) -> EVResult:

Compare declared household-factor paths under common preferences.

Weights must sum to one within absolute tolerance 1e-12. The initial observation is undiscounted; a frozen continuous tail follows the final annual observation. Inputs are read without mutation. Monetary results use reference prct/prid index levels and decimal intr; they remove the common quantity trend from the valuation year in log space.