2 replies ›

1. I’m officially freaking out – thanks for publishing

   Loading...

2. As far as I can tell the “starwars” example in the quiz has been wrong since it was introduced in the dplyr 0.7.0 announcement of 2017-06-13. The work around I think is to add extra “!!“s to the code (meaning perhaps dplyr is fine, and the example was what is wrong). But frankly that is just me trying things until something works like the documentation claims it should.

   Below is the issue: dplyr “pronoun mode” renames columns- which means if you try to use those columns you run into trouble (unless you add the extra “!!“s, which don’t seem to be part of the example or pronoun documentation).

   suppressPackageStartupMessages(library("dplyr"))
   
   # explicit code, works
   starwars %>%
     select(homeworld) %>%
     group_by(homeworld) %>%
     select(homeworld) %>%
     summarize(count = n()) %>%
     head()
   #> # A tibble: 6 x 2
   #>        homeworld count
   #>            <chr> <int>
   #> 1       Alderaan     3
   #> 2    Aleen Minor     1
   #> 3         Bespin     1
   #> 4     Bestine IV     1
   #> 5 Cato Neimoidia     1
   #> 6          Cerea     1
   
   grouping_column <- "homeworld"
   
   # augmented (extra "!!"s) pronound code
   # seems to work.
   starwars %>%
     select(.data[[!!grouping_column]]) %>%
     group_by(.data[[!!grouping_column]]) %>%
     select(.data[[!!grouping_column]]) %>%
     summarize(count = n()) %>%
     head()
   #> # A tibble: 6 x 2
   #>        homeworld count
   #>            <chr> <int>
   #> 1       Alderaan     3
   #> 2    Aleen Minor     1
   #> 3         Bespin     1
   #> 4     Bestine IV     1
   #> 5 Cato Neimoidia     1
   #> 6          Cerea     1
   
   # "pronoun code" as described in announcement,
   # does not work.
   starwars %>%
     select(.data[[grouping_column]]) %>%
     group_by(.data[[grouping_column]]) %>%
     select(.data[[grouping_column]]) %>%
     summarize(count = n()) %>%
     head()
   #> Error: Must subset with a string
   

   Loading...