Handling Exported Color Names

Currently, colors are defined as variables in the environment. This works well for sourcing a script. This does not work so well for a package namespace. As you can imagine, if each theme has say 8 colors then at 17 existing themes, we have 136 colors that would need to be documented. This type (a) pollutes the search path, making it harder to use autocomplete and (b) gets out of control fast.

Solution A: Export 2 lists, one with theme colors and one with discrete color palettes

This would look like defining a common list with common names (a la #11)

theme_park <- list(
    barbie = c(text_color= '#a62675ff', panel_color = '#fdf6faff', border_color = '#d74ea2ff', 
                      lighter_color = '#f5d1e6ff', light_color = '#eeb4d7ff', medium_color = '#d74ea2ff', 
                      dark_color= '#bf2986ff'),
  ...
)

and a second common list for discrete color palettes:

theme_palettes <- list(
    nemo = c(
      '#E9F4FB', '#FE691D', '#015DC2', '#FCDD2E', '#7867A0',
      '#BE1D57', '#798A5A', '#005478','#1B1A3D'
    ), 
...
)

Solution B: Have two color vectors for each

Theme colors would be added to the theme Rd and might look like:

theme_barbie_colors <- c(text_color= '#a62675ff', panel_color = '#fdf6faff', border_color = '#d74ea2ff', 
                      lighter_color = '#f5d1e6ff', light_color = '#eeb4d7ff', medium_color = '#d74ea2ff', 
                      dark_color= '#bf2986ff')

Colors that you might want to use within plots would be added to the scale Rd and might look like:

nemo_colors <- c(
  '#E9F4FB', '#FE691D', '#015DC2', '#FCDD2E', '#7867A0',
  '#BE1D57', '#798A5A', '#005478','#1B1A3D'
)

Solution C:

Don't export colors. This is the easiest, but probably the least fun.

Of course, we can also mix and match, but probably want to pick something consistent that is maintainable.

MatthewBJane / ThemePark