How to make your plot match with the global app theme?

R
ggplot2
bslib
shiny
Author

Philippe PERET

Published

June 4, 2025

Plots generated with the ggplot2 library can be customized to improve data visualization & user experience.

Code 
library(ggplot2)

# -- generate data
data <- data.frame(year = 2020:2025,
                   value = c(100, 110, 112, 107, 109, 115))

# -- build plot
p <- ggplot(data,
       aes(x = year,
           y = value)) +
  
  geom_line(
    color = "#2596be") +
  
  geom_point(
    size = 4,
    shape = 21,
    color = "#2596be",
    fill = "#eab676")

# -- print
p

But keeping the color arguments aligned across the app can be tedious as you need to copy / paste the color code many times. It also makes your code less reusable in other projects as it will need to be updated in many places.

Pass colors as arguments of the plot function

Whenever I make a new plot, I always encapsulate the plot layer building into a dedicated function that will be called and return a ggplot2 object (most time it will feed an output with renderPlot()). It makes it much easier to maintain the plot code as you can use it directly from the console for fine tuning, but also to reuse that code for another project.

As you will pass the data as an argument to the plot function, a first approach is to define a variable (typically a named vector) that carries the color codes at the app level and pass it as a parameter of dedicated plot functions.

Code 
# -- define color variable
colors <- c("primary" = "#2596be", "secondary" = "#eab676")

# -- define plot function with theme argument
basic_plot <- function(data, theme){
  
  # -- build & return plot
  ggplot(data,
         aes(x = year,
             y = value)) +
    
    geom_line(
      color = theme['primary']) +
    
    geom_point(
      size = 4,
      shape = 21,
      color = theme['primary'],
      fill = theme['secondary'])
  
}

# -- call function with colors
p <- basic_plot(data, theme = colors)

# -- print
p

With this, colors can be tuned easily across the app and plot functions can be reused in a different project without much update since you will just update the color variable at the app level.

Code 
# -- update color variable
colors <- c("primary" = "#5d25be", "secondary" = "#ffff9b")

# -- call function with colors
p <- basic_plot(data, theme = colors)

# -- print
p

Pass theme colors to the plot function

In case your Shiny app is using bslib to build the ui, then it’s even possible to reuse the global app theme to feed this color variable and share it to all your plots.

To do so, you need - from the server side - to get the current theme that was applied on the ui side.
And bslib has a bs_current_theme() function for that:

Code 
# -- Dependencies
library(shiny)
library(bslib)

# -- Define ui
ui <- page_fillable(
  title = "Demo",
  
  # -- Set theme
  theme = bs_theme(bootswatch = "minty"),
  
  # -- Display theme object structure
  verbatimTextOutput("theme_str")
  
)


# -- Define server
server <- function(input, output) {
  
  # -- Get theme
  theme <- bs_current_theme()
  
  # -- Display theme object structure
  output$theme_str <- renderPrint(str(theme))

}

shinyApp(ui, server)

If you have a look into that theme object, you will soon understand that diving into it to get the theme colors won’t be an easy task. Indeed bs_current_theme() returns a bs_theme() object that is itself defined as a sass::sass_bundle() (list-like) object.

Theme object structure

Once again, bslib has made it easy for you and provides a handy bs_get_variables() function to get the colors out of that theme bundle without effort:

Code 
# -- Define ui
ui <- page_fillable(
  title = "Demo",
  
  # -- Set theme
  theme = bs_theme(bootswatch = "minty"),
  
  # -- Display theme object structure
  uiOutput("theme_colors")
  
)


# -- Define server
server <- function(input, output) {
  
  # -- Get theme
  theme <- bs_current_theme()
  
  # -- Get colors from the theme
  colors <- bs_get_variables(theme, varnames = c("primary", "secondary"))
  
  # -- Display theme object structure
  output$theme_colors <- renderUI(
    layout_column_wrap(
    card(
      card_header(
        class = "bg-primary"),
      p("Primary color =", colors['primary'])
    ),
    
    card(
      card_header(
        class = "bg-secondary"),
      p("Secondary color =", colors['secondary'])
    )))
  
}

shinyApp(ui, server)

Now you just need to pass this variable to the plot function we customized in the first step and get your plot match with the global app theme.

Code 
# -- Define ui
ui <- page_fillable(
  title = "Demo",
  
  # -- Set theme
  theme = bs_theme(bootswatch = "flatly"),
  
  # -- Display theme object structure
  uiOutput("theme_colors"),
  plotOutput("plot")
  
)


# -- Define server
server <- function(input, output) {
  
  # -- Get theme
  theme <- bs_current_theme()
  
  # -- Get colors from the theme
  colors <- bs_get_variables(theme, varnames = c("primary", "secondary"))
  
  # -- Display theme object structure
  output$theme_colors <- renderUI(
    layout_column_wrap(
    card(
      card_header(
        class = "bg-primary"),
      p("Primary color =", colors['primary'])
    ),
    
    card(
      card_header(
        class = "bg-secondary"),
      p("Secondary color =", colors['secondary'])
    )))
  
  # -- Define plot output
  output$plot <- renderPlot(basic_plot(data, theme = colors))
  
}

shinyApp(ui, server)
Theme minty Theme flatly

A basic example is provided in the GitHub repository listed below in the references & materials section.

References & materials