causalEffect: causalEffect class

In ericdunipace/causalOT: Optimal Transport Weights for Causal Inference

causalEffect class

Description

causalEffect class

causalEffect constructor function

Usage

causalEffect(data, causalWeights, model.outputs, augment.estimate, call)

Arguments

data	an object of class dataHolder
causalWeights	an object of class causalWeights
model.outputs	Outputs of the estimate_model() function
augment.estimate	Is the estimate to be the augmented (doubly robust) estimator? TRUE/FALSE
call	the call used to calculate the treatment effects

Details

The variables in slot augmentedData are

• weights: The causalWeights targeting the causal estimand.

• y_obs: The vector of the observed outcomes for each observation

• y_0: The outcome under the control condition. Missingness respects the design of the experiment. i.e., Y(0) | Z = 1 = NA.

• y_hat_0: The conditional mean outcome under the control condition. Estimated from a model.

• y_hat_1: The conditional mean outcome under the treatment condition. Estimated from a model.

• x: The columns denoting the covariates.

• z: The treatment indicator.

The slot fit is a list with slots control, treated, and overall_sample. Control and treated will be filled if estimate.separately is TRUE in estimate_effect. overall_sample will be filled if estimate.separately is FALSE.

Value

an object of class causalEffect

Slots

estimate

The estimated treatment effect.

estimand

The estimand of interest

weights

The weights as an object of class causalWeights

augmentedData

The data as a data.frame with variables weights, y_obs, y_0, y_1, y_hat_0, y_hat_1, x, and z. See details for more info.

fit

The fitted model if present. See details.

call

The call from the estimate_effect() function.

ericdunipace/causalOT documentation built on Aug. 8, 2024, 6:14 p.m.