Diagnostic plots for logical values

[lillemets]

I'm unable to create plots for class mids objects using densityplot, stripplot or bwplot because the mids object contains imputed logical values.

Error in bind_rows_(x, .id) : 
  Column `variablename` can't be converted from logical to numeric

Is this expected behavior? Is there a solution?

[stefvanbuuren]

These lattice functions are for numeric outcomes. One way to circumvent the error is to manually convert logical by as.numeric before imputation, and use the default imputation method (pmm) for these variables. However, the resulting plots might not make much sense since there are only two different values.