Model data

Create data containing the treatment variable and all other covariates. This data is required to build the design (i.e., x) matrix in either a propensity score model or a covariate adjusted outcomes model. Only variables required by the model are kept.

make_xdata(
  formula,
  data,
  imputation = c("multi", "none"),
  missing_dummy = FALSE,
  id = "patient_id",
  n_impute = 5
)

Arguments

formula	A formula for the propensity score with the treatment variable to the left of the ~ operator and the model terms on the right.
data	The dataset containing the variables specified in formula.
imputation	Imputation strategy to use. Use "multi" to multiply impute data with multi_impute(). No imputation is performed when the "none" is chosen.
missing_dummy	Whether to create dummy variables for variables when observations are missing.
id	The variable uniquely identifying each patient.
n_impute	Number of imnputations to perform. Only used when imputation = "multi".

Value

An object of class model_data inheriting from dplyr::tibble. It contains all variables used in formula and, if multiple imputation is used, an additional column named imp indicating the imputation number.