How to control line chart colors in plotly when lines are added via shiny::selectInput()

[petzi53]

I want to display the life expectancy for various countries over time. The user can choose from multiple shiny::selectInput().

My problem is that the colors for the different countries change with the number of chosen countries. For instance, if only Afghanistan is selected, the line is orange, but if Afghanistan and Albania are chosen, the color for Afghanistan changes to light blue. This is distracting and confusing.

(The only way for me to prevent the confusing color changes was to add an actionButton() together with the use of eventReactive() to delay reactions until a user clicks the action button. However, in this case, the UI is more complex and less intuitive.

suppressWarnings(suppressPackageStartupMessages({
    library(shiny)
    library(bslib)
    library(dplyr)
    library(plotly)
}))

gapminder <- gapminder::gapminder

ui <- page_sidebar(
    titlePanel("Life Expectations (1952-2007)"),
    sidebar = sidebar(
        selectInput(
            inputId = "country",
            label = "Country",
            choices = unique(gapminder$country),
            multiple = TRUE
        )
    ),
    card(plotlyOutput("p"))
)

server <- function(input, output, session) {

    countries <- reactive({
        req(input$country)
        gapminder |>
            select(year, lifeExp, country) |>
            filter(country %in% input$country) |>
            arrange(year) |>
            na.omit() |>
            droplevels()
    })

    output$p <- renderPlotly({
        req(countries())
        plotly::plot_ly(
            data = countries(),
            x = ~year,
            y = ~lifeExp,
            color = ~country,
            colors = RColorBrewer::brewer.pal(12, "Paired"),
            type = 'scatter',
            mode = 'lines+markers',
            marker = list(size = 10)
        )
    })
}

shinyApp(ui, server)
#> 
#> Listening on http://127.0.0.1:6810

^{Created on 2025-09-06 with reprex v2.1.1}

I found similar answered questions in StackOverflow, but they didn't help me:

• Same color assigned to same level of factor in R plotly: The number of elements for the line chart is fixed (invariable), and it is therefore easy to provide every line with a pre-defined color.
• Setting different line colors on the chart in R Shiny: This example suffers from the same problem as my question. Although it uses the same colors for lines of both menus, adding another line in one of the dropdown menus changes the color of the previously chosen lines.

So my question is:
Is there a way to use a multiple selectInput() function to add countries and to adjust, e.g., to fix the colors for the countries already chosen, regardless of the number of countries selected interactively?

[petzi53]

I found a solution myself. (I just noted that there was a comment to my question in StackOverflow by Tim G which provided a similar solution!)

The tricky part for me was that I needed a named color vector, as shown in the second example of Custom Color Scale.

What follows is the repetition of my Gapminder example with four lines added and one line changed.

suppressWarnings(suppressPackageStartupMessages({
    library(shiny)
    library(bslib)
    library(dplyr)
    library(plotly)
}))

gapminder <- gapminder::gapminder

ui <- page_sidebar(
    titlePanel("Life Expectations (1952-2007)"),
    sidebar = sidebar(
        selectInput(
            inputId = "country",
            label = "Country",
            choices = unique(gapminder$country),
            multiple = TRUE
        )
    ),
    card(plotlyOutput("p"))
)

server <- function(input, output, session) {

    pal = RColorBrewer::brewer.pal(12, "Paired")  # added

    countries <- reactive({
        req(input$country)
        gapminder |>
            select(year, lifeExp, country) |>
            filter(country %in% input$country) |>
            arrange(year) |>
            na.omit() |>
            droplevels()
    })

    output$p <- renderPlotly({
        req(countries())
        l <- length(input$country)           # added
        length(pal) <- l                     # added
        pal <- setNames(pal, input$country)  # added
        plotly::plot_ly(
            data = countries(),
            x = ~year,
            y = ~lifeExp,
            color = ~country,
            colors = pal,                     # changed
            type = 'scatter',
            mode = 'lines+markers',
            marker = list(size = 10)
        )
    })
}

shinyApp(ui, server)
#> 
#> Listening on http://127.0.0.1:5510

^{Created on 2025-09-08 with reprex v2.1.1}

[ismirsehregal]

Here the according SO crosspost can be found.