setdiff vs. setequal inconsitency when columns have different types

[sigmaeon]

It seems like setequal compares column types while setdiff doesn't.

df1 <- data.frame(x=1)
df2 <- df1

# df1$x and df2$x are both of type `num`
setdiff(df1, df2) # --> will output 0-rowed data.frame
setequal(df1, df2) # --> will output TRUE

df2$x %<>% as.integer 

# df1$x is of type `num` while df2$x is of type `int`
setdiff(df1, df2) # --> will output 0-rowed data.frame
setequal(df1, df2) # --> will output FALSE

I can see how this could actually be the intended behavior.
However, in that case a hint in the docs would be great.

[hadley]

library(dplyr, warn.conflicts = FALSE)

df1 <- tibble(x = 1)
df2 <- tibble(x = 1L)

setdiff(df1, df2)
#> # A tibble: 0 × 1
#> # … with 1 variable: x <dbl>
#> # ℹ Use `colnames()` to see all variable names
setequal(df1, df2)
#> [1] FALSE

^{Created on 2022-07-22 by the reprex package (v2.0.1)}