Objective: write a function that will take a data frame as its first
argument, and then two additional arguments, lists of argments, passed to
dplyr::select such that the function will return two data frames.
Here is a working example
my_select <- function(.data, df1, df2) {
DF1 <- dplyr::select(.data, rlang::UQS(df1))
DF2 <- dplyr::select(.data, rlang::UQS(df2))
list(DF1 = DF1, DF2 = DF2)
}
working_eg <-
my_select(mtcars,
df1 = alist(dplyr::contains("r"), dplyr::matches("^.p.*")),
df2 = alist(disp))
str(working_eg, max.length = 1L)
## List of 2
## $ DF1:'data.frame': 32 obs. of 5 variables:
## ..$ drat: num [1:32] 3.9 3.9 3.85 3.08 3.15 2.76 3.21 3.69 3.92 3.92 ...
## ..$ gear: num [1:32] 4 4 4 3 3 3 3 4 4 4 ...
## ..$ carb: num [1:32] 4 4 1 1 2 1 4 2 2 4 ...
## ..$ mpg : num [1:32] 21 21 22.8 21.4 18.7 18.1 14.3 24.4 22.8 19.2 ...
## ..$ hp : num [1:32] 110 110 93 110 175 105 245 62 95 123 ...
## $ DF2:'data.frame': 32 obs. of 1 variable:
## ..$ disp: num [1:32] 160 160 108 258 360 ...
I would prefer that the argments df1 and df2 take a list not a alist.
however, the function my_select will fail if the arguments are list
my_select(mtcars,
df1 = list(dplyr::contains("r"), dplyr::matches("^.p.*")),
df2 = list(disp))
## Error: Variable context not set
I don't want to ask end users to use alist, if possible, as I don't have a
good way to test that alist was used to pass arguments instead of list.
I've tried several combinations of rlang::UQ, rlang::UQE, rlang::UQS
with rlang::quo, rlang::enquos, and rlang::quos to fix this. What I
thought was the best approach was:
my_select2 <- function(.data, df1, df2) {
DF1 <- dplyr::select(.data, rlang::UQS(rlang::quos(df1)))
DF2 <- dplyr::select(.data, rlang::UQS(rlang::quos(df2)))
list(DF1 = DF1, DF2 = DF2)
}
my_select2(mtcars,
df1 = list(dplyr::contains("r"), dplyr::matches("^.p.*")),
df2 = list(disp))
## Error: `df1` must resolve to integer column positions, not a list
Is there a way to use the rlang
package with dplyr so that the
syntax of my_select2 will return the same object as my_select does when
arguments are passed via alist?
packageVersion("dplyr")
# [1] ‘0.7.4’
packageVersion("rlang")
# [1] ‘0.1.2’
@MrFlick's comment was the reminder I needed to solve the problem and improve the api.
my_select3 <- function(.data, df1, df2) {
DF1 <- dplyr::select(.data, rlang::UQS((df1)))
DF2 <- dplyr::select(.data, rlang::UQS((df2)))
list(DF1 = DF1, DF2 = DF2)
}
working_eg2 <-
my_select3(mtcars,
df1 = dplyr::vars(dplyr::contains("r"), dplyr::matches("^.p.*")),
df2 = dplyr::vars(disp))
all.equal(working_eg, working_eg2)
# [1] TRUE