I have a with users, the groups they belong to, and two columns I want to expand (below). I want to expand it so that each user keeps the groups that they're in but for each group they're in gets all possible permutations of country and season.

Currently I'm expanding all of the columns then doing a semi-join to the table to drop all of the user and group combinations which don't exist.

df <- data.frame(User_ID = c(1, 2, 1, 3), 
    Group = c(1,2,2,1), 
    Country = c("France", "Spain", "Italy","Italy"), 
    Season = c("Summer", "Autumn", "Winter", "Spring")) 

df %>% 
    expand(User_ID, Group, Country, Season) %>%
    semi_join(df, by = c("User_ID", "Group")

Is this the most efficient way to do this, or am I missing something with the expand function?

CodePudding user response：

Is this what you want (4 user/groups x 3 countries x 4 seasons, i.e. 48 rows)?

library(tidyverse)

tribble(
  ~User_ID, ~Group, ~Country, ~Season,
  1, 1, "France", "Autumn",
  2, 2, "Spain", "Summer",
  1, 2, "Italy", "Winter",
  3, 2, "Italy", "Spring"
) |> 
  complete(nesting(User_ID, Group), Country, Season)

#> # A tibble: 48 × 4
#>    User_ID Group Country Season
#>      <dbl> <dbl> <chr>   <chr> 
#>  1       1     1 France  Autumn
#>  2       1     1 France  Spring
#>  3       1     1 France  Summer
#>  4       1     1 France  Winter
#>  5       1     1 Italy   Autumn
#>  6       1     1 Italy   Spring
#>  7       1     1 Italy   Summer
#>  8       1     1 Italy   Winter
#>  9       1     1 Spain   Autumn
#> 10       1     1 Spain   Spring
#> # … with 38 more rows

^{Created on 2022-06-14 by the reprex package (v2.0.1)}