Skip to contents

Fill in missing values and convert data to canonical form

Source: R/patch_data_set.R
patch_data_set.Rd

This method will fill in means for all missing covariate values and return a dataset where rows with missing outcome, cluster or block membership, and treatment assignment are all dropped and everything else is filled in. To do this, it will convert categorical covariates to a set of dummy variables (dropping the reference group).

Usage

patch_data_set(
  formula = NULL,
  data = NULL,
  control_formula = NULL,
  warn_missing = FALSE
)

Arguments

formula

Notation for Y ~ Z | clusterID | blockID (| blockID is optional). If NULL, `data` is assumed to already be in canonical form.

data

Dataframe to patch.

control_formula

What variables to control for, in the form of "~ X1 + X2".

warn_missing

If TRUE, warn when rows are dropped or covariates are imputed.

Details

It will drop missing data indicators that are co-linear with prior missing data indicators.

If there are categorical covariates, then it will convert those to dummy variables regardless of missingness. The `control_formula` attribute handed back will give the updated control_formula in this case.

This attribute will also have all missing value indicators added in as covariates.