Generate Observational Data

Generates synthetic observational data (Y, Z, X) with known treatment effect tau. The propensity score is a sigmoid function of the covariates.

Usage

generate_data(N = 10000, k = 2, tau = 10)

Arguments

N

Integer. Number of observations. Default 10000.

k

Integer. Number of covariates. Default 2.

tau

Numeric. True average treatment effect. Default 10.

Value

A list with components Y, Z, and X.

Examples

data <- generate_data(N = 1000, k = 2, tau = 5)
obs <- observational(data$Y, data$Z, data$X)
est_via_aipw(obs)
#> Causal Model Estimation Result
#> ------------------------------
#>   ATE:      4.929815
#>   SE:       0.073875
#>   z:        66.7319
#>   p-value:  0.0000
#>   95% CI:   [4.785023, 5.074608]