unquote a column name given in another column

[qcta]

In dplyr verbs, I understand how to unquote a column name given as a character in the environment:

var <- "height"

starwars %>%
  transmute(test = !!sym(var))
# or
starwars %>%
  transmute(test = .data[[var]])

But what if the column name is given in another column in the same tibble? E.g.:

starwars %>%
  mutate(selected.var = "height") %>%
  transmute(test = !!sym(selected.var))
# Error in is_symbol(x) : object 'selected.var' not found

(Intended output would be the same as the two examples above.) I don't see this use-case covered in the programming vignette, and I'm seeing that the tidyeval book is no longer current. Apologies if I've missed this in the documentation.

[technocrat]

Unclear on the question

suppressPackageStartupMessages({
  library(dplyr)
})

var <- "height"

starwars %>%
  transmute(test = !!sym(var))
#> # A tibble: 87 x 1
#>     test
#>    <int>
#>  1   172
#>  2   167
#>  3    96
#>  4   202
#>  5   150
#>  6   178
#>  7   165
#>  8    97
#>  9   183
#> 10   182
#> # … with 77 more rows
# or
starwars %>%
  transmute(test = .data[[var]])
#> # A tibble: 87 x 1
#>     test
#>    <int>
#>  1   172
#>  2   167
#>  3    96
#>  4   202
#>  5   150
#>  6   178
#>  7   165
#>  8    97
#>  9   183
#> 10   182
#> # … with 77 more rows

[technocrat]

See the FAQ: How to do a minimal reproducible example reprex for beginners, which is what my post is. The reason for my question is that the reprex does not reproduce the error.

Are you using dplyr_1.0.5?

[qcta]

You're not reproducing the error for the final command? I omitted the output from the first block of code for brevity's sake since there is no error and it is working as intended.
Yes, I'm using 1.0.5.

[technocrat]

@qcta

starwars %>%
  mutate(selected.var = "height") %>%
  transmute(test = !!sym(selected.var))

Apologies. One of the advantages of a reprex is precisely that it will throw the error with one cut-and-paste pass.

This is how'd I'd do it

suppressPackageStartupMessages({
  library(dplyr)
})

var <- "height"

starwars %>%
  mutate(selected.var = !!sym(var)) %>%
  transmute(test = selected.var)
#> # A tibble: 87 x 1
#>     test
#>    <int>
#>  1   172
#>  2   167
#>  3    96
#>  4   202
#>  5   150
#>  6   178
#>  7   165
#>  8    97
#>  9   183
#> 10   182
#> # … with 77 more rows

or

suppressPackageStartupMessages({
  library(dplyr)
})

var <- "height"

starwars %>%
  transmute(test = !!sym(var))
#> # A tibble: 87 x 1
#>     test
#>    <int>
#>  1   172
#>  2   167
#>  3    96
#>  4   202
#>  5   150
#>  6   178
#>  7   165
#>  8    97
#>  9   183
#> 10   182
#> # … with 77 more rows

[qcta]

Thanks. I did use reprex - I just shortened the output afterward.

In my actual use-case, I really need to have selected.var be a column within the tibble, because it can vary. So it cannot be a static variable in the environment.

[technocrat]

@qcta

actual use-case, I really need to have selected.var be a column within the tibble

Glad to take a look at an illustration.