value-diffs.md
In versus: Compare Data Frames

value_diffs with a single column works

Code
  value_diffs(comp, mpg)
Output
  # A tibble: 2 x 3
    mpg_a mpg_b car       
    <dbl> <dbl> <chr>     
  1  14.3  16.3 Duster 360
  2  24.4  26.4 Merc 240D

value_diffs works when the supplied columns have no diffs

Code
  value_diffs(comp, hp)
Output
  # A tibble: 0 x 3
  # i 3 variables: hp_a <dbl>, hp_b <dbl>, car <chr>

Code
  value_diffs_stacked(comp, c(hp, drat))
Output
  # A tibble: 0 x 4
  # i 4 variables: column <chr>, val_a <dbl>, val_b <dbl>, car <chr>

Error when `comparison` isn't a comparison

Code
  value_diffs(mtcars, mpg)
Condition
  Error in `value_diffs()`:
  ! Problem with argument `comparison = mtcars`
  i `comparison` must be the output of `versus::compare()`

Code
  value_diffs_stacked(mtcars, c(mpg, disp))
Condition
  Error in `value_diffs_stacked()`:
  ! Problem with argument `comparison = mtcars`
  i `comparison` must be the output of `versus::compare()`

Error on value_diffs with empty selection

Code
  value_diffs(comp, where(is.factor))
Condition
  Error in `value_diffs()`:
  ! Problem with argument `column = where(is.factor)`:
  * Must select at least one item.

Code
  value_diffs_stacked(comp, where(is.factor))
Condition
  Error in `value_diffs_stacked()`:
  ! Problem with argument `column = where(is.factor)`:
  * Must select at least one item.

Error on value_diffs when column doesn't exist

Code
  value_diffs(comp, bear)
Condition
  Error in `value_diffs()`:
  ! Problem with argument `column = bear`:
  * Must select columns from `comparison$intersection`
  i column `bear` is not part of the supplied comparison

Code
  value_diffs_stacked(comp, c(bear, mpg))
Condition
  Error in `value_diffs_stacked()`:
  ! Problem with argument `column = c(bear, mpg)`:
  * Must select columns from `comparison$intersection`
  i column `bear` is not part of the supplied comparison

Error on value_diffs() with multiple columns

Code
  value_diffs(comp, c(mpg, disp))
Condition
  Error in `value_diffs()`:
  ! Must select only one column.
  i Columns selected: mpg, disp
  i For multiple columns, use `value_diffs_stacked()`

value_diffs_stacked() works

Code
  value_diffs_stacked(comp, c(mpg, disp))
Output
  # A tibble: 4 x 4
    column val_a val_b car           
    <chr>  <dbl> <dbl> <chr>         
  1 mpg     14.3  16.3 Duster 360    
  2 mpg     24.4  26.4 Merc 240D     
  3 disp   109   108   Datsun 710    
  4 disp   259   258   Hornet 4 Drive

Code
  value_diffs_stacked(comp, where(is.numeric))
Output
  # A tibble: 5 x 4
    column val_a val_b car           
    <chr>  <dbl> <dbl> <chr>         
  1 mpg     14.3  16.3 Duster 360    
  2 mpg     24.4  26.4 Merc 240D     
  3 cyl      6    NA   Hornet 4 Drive
  4 disp   109   108   Datsun 710    
  5 disp   259   258   Hornet 4 Drive

value_diffs_stacked() works without column

Code
  value_diffs_stacked(comp)
Output
  # A tibble: 5 x 4
    column val_a val_b car           
    <chr>  <dbl> <dbl> <chr>         
  1 mpg     14.3  16.3 Duster 360    
  2 mpg     24.4  26.4 Merc 240D     
  3 cyl      6    NA   Hornet 4 Drive
  4 disp   109   108   Datsun 710    
  5 disp   259   258   Hornet 4 Drive

value_diffs_stacked() coerces to char on incompatible ptypes

Code
  value_diffs_stacked(comp)
Message
  i Columns converted to character: val_a
Output
  # A tibble: 5 x 4
    column val_a val_b car           
    <chr>  <chr> <dbl> <chr>         
  1 mpg    14.3   16.3 Duster 360    
  2 mpg    24.4   26.4 Merc 240D     
  3 cyl    6      NA   Hornet 4 Drive
  4 disp   109   108   Datsun 710    
  5 disp   259   258   Hornet 4 Drive