pystra.decision.ddo.DDO

Use it: User guide · Worked example · Theory

class DDO(*, study, criterion, objective=None)

Bases: object

Evaluate a decision context with an objective and acceptability criterion.

Construct directly from the three pieces:

ddo = DDO(
    study=study,
    objective=CostBenefitModel(...),
    criterion=LQI.from_country(...),
)

Construction is keyword-only by design: study, objective and criterion are easy to transpose positionally, which would silently misassign them. run() evaluates every alternative and caches the table; optimize() returns the best feasible alternative and economic_optimum() the unconstrained economic best.

Methods

economic_optimum	Return the alternative with the largest objective, ignoring feasibility.
feasible_results	Return evaluated alternatives satisfying the criterion.
optimize	Return the feasible alternative with the largest objective value.
plot	Plot results from the most recent run.
run	Evaluate every alternative; a failed rerun clears the previous table.
summary	Return the key decision points as a small labeled table.

property results: DataFrame | None

Independent table from the last completed run, or None.

run()

Evaluate every alternative; a failed rerun clears the previous table.

Return type:

DataFrame

economic_optimum()

Return the alternative with the largest objective, ignoring feasibility.

Return type:

Series

feasible_results()

Return evaluated alternatives satisfying the criterion.

Return type:

DataFrame

optimize()

Return the feasible alternative with the largest objective value.

Return type:

Series

summary()

Return the key decision points as a small labeled table.

One row for the unconstrained economic optimum and, when the criterion admits one, a second for the best feasible alternative. Columns are the design variable, pf, beta (when present), the objective, and the criterion’s feasibility flag. This is the table a designer reads off a study, without rebuilding it row by row.

Return type:

DataFrame

plot(design=None, quantities=None, **kwargs)

Plot results from the most recent run.

Return type:

tuple[Figure, ndarray]