facet: bug where character & NAs

[davidhodge931]

facet: bug where character & NAs

[davidhodge931]

Solution:

• if character or logical: drop = TRUE
• if factor: drop = FALSE

[davidhodge931]

Kind of a pain to do, as have to get the factor/character/logical value's of factor and factor2 first

[davidhodge931]

Relates to #5401 ggplot2 bug

[davidhodge931]

Either wait for ggplot2 to fix this or add code like this:

if (is.character(rlang::eval_tidy(facet, data)) | 
    is.logical(rlang::eval_tidy(facet, data))) {
  data <- data |> 
    mutate(across(!!facet, as.factor))
}
if (is.character(rlang::eval_tidy(facet2, data)) | 
    is.logical(rlang::eval_tidy(facet2, data))) {
  data <- data |> 
    mutate(across(!!facet2, as.factor))
}

[davidhodge931]

This is fine - must have been something in my environment