Should tidy_select_variables() rearrange the order of the variables according to include?

[larmarange]

Currently, the include parameter preserves the original order of the variables.

library(broom.helpers)

mod <- lm(Sepal.Length ~ ., data = iris)

mod |> 
  tidy_plus_plus() |> 
  dplyr::select(term, variable, label, estimate)
#> # A tibble: 6 × 4
#>   term              variable     label        estimate
#>   <chr>             <chr>        <chr>           <dbl>
#> 1 Sepal.Width       Sepal.Width  Sepal.Width     0.496
#> 2 Petal.Length      Petal.Length Petal.Length    0.829
#> 3 Petal.Width       Petal.Width  Petal.Width    -0.315
#> 4 Speciessetosa     Species      setosa          0    
#> 5 Speciesversicolor Species      versicolor     -0.724
#> 6 Speciesvirginica  Species      virginica      -1.02
  

mod |> 
  tidy_plus_plus(include = c(Species, Petal.Length)) |> 
  dplyr::select(term, variable, label, estimate)
#> # A tibble: 4 × 4
#>   term              variable     label        estimate
#>   <chr>             <chr>        <chr>           <dbl>
#> 1 Petal.Length      Petal.Length Petal.Length    0.829
#> 2 Speciessetosa     Species      setosa          0    
#> 3 Speciesversicolor Species      versicolor     -0.724
#> 4 Speciesvirginica  Species      virginica      -1.02

mod |> 
  tidy_plus_plus(include = c(Species, Petal.Length, everything())) |> 
  dplyr::select(term, variable, label, estimate)
#> # A tibble: 6 × 4
#>   term              variable     label        estimate
#>   <chr>             <chr>        <chr>           <dbl>
#> 1 Sepal.Width       Sepal.Width  Sepal.Width     0.496
#> 2 Petal.Length      Petal.Length Petal.Length    0.829
#> 3 Petal.Width       Petal.Width  Petal.Width    -0.315
#> 4 Speciessetosa     Species      setosa          0    
#> 5 Speciesversicolor Species      versicolor     -0.724
#> 6 Speciesvirginica  Species      virginica      -1.02

^{Created on 2022-10-07 with reprex v2.0.2}

To be consistent with dplyr::select(), should the table be reordered according to the include parameters?

@ddsjoberg do you have an advice about it?

[ddsjoberg]

Oh good point. I think it would be preferable to have the order adjusted with include= if it's not too hairy to make happen.

[larmarange]

I will explore it

[larmarange]

It is finally not too complicated to do it. See #184