Assign multiple columns when using mutate in dtplyr

Question

Is there a way of getting my data table to look like my target table when using dtplyr and mutate?`

A Dummy table

library(data.table)
library(dtplyr)
library(dplyr)

id <- rep(c("A","B"),each=3)
x1 <- rnorm(6)
x2 <- rnorm(6)

dat <- data.table(id,x1,x2)

A dummy function

my_fun <- function(x,y){
  cbind(a = x+10,b=y-10)  
}

And I would like to use this type of syntax

dat |> 
  group_by(id) |> 
  mutate(my_fun(x = x1,y = x2))

Where the end result will look like this

data.table(id, x1, x2, a=x1+10,b=x2-10)

I would like to have a generic solution that works for functions with variable number of columns returned but is that possible?

Answer 1

I think we would need more information about how this would work with a variable number of columns:

• Are the columns named in a specific way?
• Do the output columns need to be named in a specific way?
• Are there standard calculations being done to each column dependent on name? E.g., x1 = +10 and x2 = -10?

At any rate, here is a solution that works with your provided data to return the data.table you specified:

my_fun <- function(data, ...){
  dots <- list(...)
  cbind(data, 
        a = data[[dots$x]] + 10,
        b = data[[dots$y]] - 10
        )
}

dat |> 
  my_fun(x = "x1", y = "x2")

   id         x1         x2         a          b
1:  A  0.8485309 -0.3532837 10.848531 -10.353284
2:  A  0.7248478 -1.6561564 10.724848 -11.656156
3:  A -1.3629114  0.4210139  8.637089  -9.578986
4:  B -1.7934827  0.6717033  8.206517  -9.328297
5:  B -1.0971890 -0.3008422  8.902811 -10.300842
6:  B  0.4396630 -0.7447419 10.439663 -10.744742