CODEWITHSUNDEEP
rdplyrcase-when
tic-toc-choc
tic-toc-choc

Reputation: 961

Evaluation in dplyr::case_when()

Following an example given in the dplyr::case_when() documentation:

x <- 1:50
case_when(x %% 35 == 0 ~ "fizz buzz",
          x %% 5 == 0 ~ "fizz",
          x %% 7 == 0 ~ "buzz",
          TRUE ~ as.character(x))

I expect that the number 35 will produce "buzz" but it produces "fizz buzz"

My reasoning is that case_when() evaluates all statements one by one regardless if a previous one is true or not (since it does evaluate TRUE ~ as.character(x) which is the last one) and that 35 %% 7 is obviously 0.

What am I missing?

Upvotes: 2

Views: 3645

Answers (1)

bmrn
bmrn

Reputation: 500

case_when() evaluates all statements one by one regardless if a previous one is true or not (since it does evaluate TRUE ~ as.character(x) which is the last one)

This is misleading, the output of case_when() is based on the first statement that is true.

TRUE ~ as.character(x) means that if x is not divisible by 5 or 7 then then x will be returned as a string ie for x = 5, "5" will be returned.

If x is divisible by 5 or 7, casewhen() does not evaluate subsequent cases. "fizz" and "buzz" are not passed to as.character(x) and they do not have to be because they are already character strings.

Upvotes: 3

Related Questions