class econml.inference.GenericSingleTreatmentModelFinalInference[source]

Bases: econml.inference._inference.GenericModelFinalInference

Inference based on predict_interval of the model_final model. Assumes that treatment is single dimensional. Thus, the predict(X) of model_final gives the const_marginal_effect(X). The single dimensionality allows us to implement effect_interval(X, T0, T1) based on the const_marginal_effect_interval.




ate_inference([X, T0, T1])

ate_interval([X, T0, T1, alpha])


const_marginal_ate_interval([X, alpha])


const_marginal_effect_interval(X, *[, alpha])

effect_inference(X, *, T0, T1)

effect_interval(X, *, T0, T1[, alpha])

fit(estimator, *args, **kwargs)

Fits the inference model.

marginal_ate_inference(T[, X])

marginal_ate_interval(T[, X, alpha])

marginal_effect_inference(T, X)

marginal_effect_interval(T, X, *[, alpha])

prefit(estimator, *args, **kwargs)

Performs any necessary logic before the estimator's fit has been called.

fit(estimator, *args, **kwargs)[source]

Fits the inference model.

This is called after the estimator’s fit.

prefit(estimator, *args, **kwargs)

Performs any necessary logic before the estimator’s fit has been called.