Assumptions

Assumption objects used to gate estimators and document identification.

Assumption dataclass

Represents a causal identification assumption.

Estimand

Not applicable.

Assumptions: None. Inputs: name: Stable identifier for the assumption. description: Short human-readable description. Failure modes: Mislabeling assumptions can lead to applying an estimator outside its identification scope.

to_dict()

Return a dictionary representation.

AssumptionSet

Collection of assumptions used to validate estimator applicability.

Estimand

Not applicable.

Assumptions: None. Inputs: assumptions: Iterable of Assumption objects. Outputs: None. Use has/require to check membership. Failure modes: If assumptions are missing, estimators should refuse to run.

has(name)

Return True if the assumption name is present.

names()

Return assumption names in sorted order.

require(names)

Raise ValueError if any required assumption is missing.

to_dict()

Return a dictionary keyed by assumption name.

Default assumption objects for CRL estimands.