An NA means the covariate is unknown for that row, so it resolves to the
average of the fitted covariate: evenly over the levels of a factor, and to
the mean of a continuous one.
Usage
handle_missing_values(data)
Arguments
- data
A data frame holding the formula covariates, with categorical
columns already declared as factors over the fitted levels and continuous
columns already z-scored against the fitted rows
Value
A data frame with missing values handled: