Interface: IdentifyEffectResponse
@kortexya/reasoninglayer / IdentifyEffectResponse
Interface: IdentifyEffectResponse
Section titled “Interface: IdentifyEffectResponse”Defined in: src/types/causal.ts:432
Response for effect identification.
Remarks
Section titled “Remarks”Either a proof-carrying estimand (assumption ledger + licensed derivation + persisted certificate id) or an explicit refusal carrying the hedge witness when the effect is not identifiable.
Properties
Section titled “Properties”adjustmentSet?
Section titled “adjustmentSet?”
optionaladjustmentSet?:string[] |null
Defined in: src/types/causal.ts:434
The adjustment set, when the estimand is an adjustment form.
assumptions
Section titled “assumptions”assumptions:
CausalAssumptionDto[]
Defined in: src/types/causal.ts:436
Every assumption the result consumes.
certificateId?
Section titled “certificateId?”
optionalcertificateId?:string|null
Defined in: src/types/causal.ts:438
Psi-term id of the persisted identification certificate.
derivation
Section titled “derivation”derivation:
CausalDerivationStepDto[]
Defined in: src/types/causal.ts:440
The licensed derivation.
estimand
Section titled “estimand”estimand:
object
Defined in: src/types/causal.ts:442
The identified estimand (serialized expression tree), when identifiable.
hedge?
Section titled “hedge?”
optionalhedge?:CausalHedgeDto|null
Defined in: src/types/causal.ts:444
Hedge witness, when the ID algorithm proved non-identifiability.
hierarchyLevel
Section titled “hierarchyLevel”hierarchyLevel:
number
Defined in: src/types/causal.ts:446
Causal-hierarchy level of the answered query (2 = interventional).
identifiable
Section titled “identifiable”identifiable:
boolean
Defined in: src/types/causal.ts:448
Whether the effect is identifiable from the declared model.
optimalAdjustmentSet?
Section titled “optimalAdjustmentSet?”
optionaloptimalAdjustmentSet?:string[] |null
Defined in: src/types/causal.ts:455
The Henckel variance-optimal adjustment set (O-set), when it is fully observed and re-verified against the backdoor criterion. Prefer these covariates when your data has them: the estimation gate accepts either certified set.
outcome
Section titled “outcome”outcome:
string
Defined in: src/types/causal.ts:457
Outcome variable.
reason?
Section titled “reason?”
optionalreason?:string|null
Defined in: src/types/causal.ts:459
Refusal reason, when not identifiable.
treatment
Section titled “treatment”treatment:
string
Defined in: src/types/causal.ts:461
Treatment variable.