I am trying to mutate a new column in an R tibble which is a logical equation of existing factor columns. I keep encountering the error the condition has length > 1 despite the simple control flow of my problem. My actual code has a more complex logical flow and works fine on some factor columns but not others.

Here is an MRE:

my_tibble <- tibble::tibble(a = as.factor(c("dog","cat","dog")), b = as.factor(c(1,5,6)))
my_fn <- function(a, b){
    if (a == "dog"){
        if (b == 1){
            return(12)
        }
        else{
            return(3)
        }
    } else{
        return(5)
    }
}
dplyr::mutate(my_tibble, c = my_fn(a,b)) # Error here!

However, my_tibble[1,1] == "dog" evaluates to TRUE.

The error itself is:

Error in `mutate()`:
ℹ In argument: `c = my_fn(a, b)`.
Caused by error in `if (a == "dog") ...`:
! the condition has length > 1

I have to imagine there is something about logical expressions involving factors I am missing here. Any guidance is much appreciated.

type here