redundant_ifelse

What it does

This checks for cases of ifelse(), dplyr::if_else(), and data.table::fifelse() where the output is always a boolean. In those cases, using the condition directly is enough, the function call is redundant.

Why is this bad?

This rule looks for 4 cases:

• ifelse(condition, TRUE, FALSE)
• ifelse(condition, FALSE, TRUE)
• ifelse(condition, TRUE, TRUE)
• ifelse(condition, FALSE, FALSE)

The first two cases can be simplified to condition and !condition respectively. The last two cases are very likely to be mistakes since the output is always the same.

This rule has a safe fix and doesn’t handle calls to dplyr::if_else() and data.table::fifelse() when they have additional arguments.

Example

ifelse(x %in% letters, TRUE, FALSE)
dplyr::if_else(x > 1, FALSE, TRUE)

Use instead:

x %in% letters
!(x > 1) # (or `x <= 1`)