Scatter of each method's relative squared error against how much of the variable is missing,
to see whether imputation gets worse as missingness grows.
Usage
plot_rse_vs_missingness(data, rse_df)
Arguments
- data
The data frame with the missing values (used for the percent missing of each variable).
- rse_df
A [compare_imputation_rse()] result or anything with `variable`, `method`, `rse`.
If it has a `pct_miss` column that is used instead of `data`.
Value
A ggplot2 object. Variables of `rse_df` that are not columns of `data` are dropped with a
warning.
