Generates data from a two part distribution with a point mass at zero and heterogeneous treatment effects

Generates semicontinuous data with heterogeneity of treatment effect

sim_semicontinuous_data(n.obs = 1000, n.vars = 25)

Arguments

n.obs	number of observations
n.vars	number of variables. Must be at least 10

Value

returns list with values y for outcome, x for design matrix, trt for treatment assignments, betanonzero for true coefficients for treatment-covariate interactions for model for whether or not a response is nonzero, betapos for true coefficients for treatment-covariate interactions for positive model, treatment_risk_ratio for the true covariate-conditional treatment effect risk ratio for each observation, pi.x for the true underlying propensity score