Introduction

Plotly has some truly awesome default plot and graph colors. These are also extensively customizable.

In this tutorial, we will have a look at the different ways to customize the colors and also how to use the RColorBrewer package to change plot colors.

Creating simulated data

In the code chunk below, we create a string vector containing a list of \(200\) cities values from a sample space of seven unique data points. Try and spot the odd one out!

set.seed(123)
cities <- sample(c("NYC", "Boston", "LA", "Seattle", "SF", "Miami", "Cape Town"),
                 200,
                 replace = TRUE)

The table() function will display a list of this nominal categorical variable and a count of the sample space elements.

table(cities)
## cities
##    Boston Cape Town        LA     Miami       NYC   Seattle        SF 
##        30        30        34        22        22        28        34

The names() function will return only the elements in the sample space of our nominal categorical variable. We can store this in an appropriately named computer variable.

city.names <- names(table(cities))
city.names
## [1] "Boston"    "Cape Town" "LA"        "Miami"     "NYC"       "Seattle"  
## [7] "SF"

The as.numeric() function will return only the counts. We will store this information too.

city.counts <- as.numeric(table(cities))
city.counts
## [1] 30 30 34 22 22 28 34

Color choices

Default color bar chart

These two vectors will allow us to create a simple bar chart, ideal for nominal categorical variables.

p1 <- plot_ly() %>% 
  add_trace(type = "bar",
            x = city.names,
            y = city.counts) %>% 
  layout(title = "Cities",
         xaxis = list(title = "City",
                      zeroline = FALSE),
         yaxis = list(title = "Count",
                      zeroline = FALSE))
p1

Default color bar chart

rgb and rgba

We see the typical deep sky blue default color. By adding the marker argument, we can stipulate the bar and the liner (edge) color. In the first example, we specify the colors by rgb (red, green, and blue) value. Each value can range from \(0\) to \(255\); that is minimum to maximum intensity. Mixing the values allows us to create many, many colors (that is \(16777216\) to be precise). In the code chunk below, we go for a gray fill and a very dark gray border of \(2\) pixels.

p2 <- plot_ly() %>% 
  add_trace(type = "bar",
            x = city.names,
            y = city.counts,
            marker = list(color = "rgb(195, 195, 195)",
                            line = list(color = "rgb(20, 20, 20)",
                                        width = 2))) %>% 
  layout(title = "Cities",
         xaxis = list(title = "City",
                      zeroline = FALSE),
         yaxis = list(title = "Count",
                      zeroline = FALSE))
p2

Using rgb color values

The rgba value allows us to add an opacity value in the range \(0\) to \(1\), the former being fully transparent and the latter indicating fully opaque. In the code chunk below, we go for a transparent red fill.

p3 <- plot_ly() %>% 
  add_trace(type = "bar",
            x = city.names,
            y = city.counts,
            marker = list(color = "rgba(255, 0, 0, 0.6)",
                            line = list(color = "rgb(20, 20, 20)",
                                        width = 2))) %>% 
  layout(title = "Cities",
         xaxis = list(title = "City",
                      zeroline = FALSE),
         yaxis = list(title = "Count",
                      zeroline = FALSE))
p3

Adding opacity

Individual rgb colors

Since we know that there are seven bars, we can specify individual colors for each. In the code below, LA and SF are colored red and the rest are silver. All have a transparency level of \(0.7\).

p4 <- plot_ly() %>% 
  add_trace(type = "bar",
            x = city.names,
            y = city.counts,
            marker = list(color = c("rgba(192, 192, 192, 0.7)",
                                    "rgba(192, 192, 192, 0.7)",
                                    "rgba(255, 0, 0, 0.7)",
                                    "rgba(192, 192, 192, 0.7)",
                                    "rgba(192, 192, 192, 0.7)",
                                    "rgba(192, 192, 192, 0.7)",
                                    "rgba(255, 0, 0, 0.7)"))) %>% 
  layout(title = "Cities",
         xaxis = list(title = "City",
                      zeroline = FALSE),
         yaxis = list(title = "Count",
                      zeroline = FALSE))
p4

Individual colors

Colors by name

Instead of rgb and rgba values, we can also use official HTML names. The opacity needs to be a vector of values. The color names are available at https://www.w3.org/TR/css-color-3/#svg-color

p5 <- plot_ly() %>% 
  add_trace(type = "bar",
            x = city.names,
            y = city.counts,
            marker = list(color = c("silver",
                                    "silver",
                                    "red",
                                    "silver",
                                    "silver",
                                    "silver",
                                    "red"),
                          opacity = rep(0.7, 7))) %>% 
  layout(title = "Cities",
         xaxis = list(title = "City",
                      zeroline = FALSE),
         yaxis = list(title = "Count",
                      zeroline = FALSE))
p5

Named color values

Hexademical colors

The hex, hsl, and hsv value sets will also work. In the code chunk below is an example of hexadecimal values.

p6 <- plot_ly() %>% 
  add_trace(type = "bar",
            x = city.names,
            y = city.counts,
            marker = list(color = c("#C0C0C0",
                                    "#C0C0C0",
                                    "#FF0000",
                                    "#C0C0C0",
                                    "#C0C0C0",
                                    "#C0C0C0",
                                    "#FF0000"),
                          opacity = rep(0.7, 7))) %>% 
  layout(title = "Cities",
         xaxis = list(title = "City",
                      zeroline = FALSE),
         yaxis = list(title = "Count",
                      zeroline = FALSE))
p6

Hexadecimal color values

Using named color palettes

In the code chunk below we create a tibble object to serve as data for the rest of this tutorial. It is saved in the df computer variable, which we print out to the screen as a datatable from the DT package.

df <- tibble(
  City = cities,
  Sale = round(rnorm(length(cities),
                      mean = 100,
                      sd = 10),
                digits = 1)
  )
datatable(df)

A box-and-whisker chart with default colors

A box-and-whisker chart is ideal to visualize the spread of a numerical variable among one or more sample space elements of a categorical variable.

p7 <- plot_ly() %>% 
  add_trace(data = df,
            y = ~Sale,
            color = ~City,
            type = "box") %>% 
  layout(title = "Sales per city",
         xaxis = list(title = "City"))
p7

Default color box plot

Using color palettes

Plotly color palettes

There are several color sets available in Plotly. These include Set1 Set2 Set3 Pastel1 Pastel2 Paired Dark2 Accent qualitative palettes, many sequential light to dark palettes, and even more dark-to-light-to-dark palettes. Below are examples of Dark2 and BrBG.

p8 <- plot_ly() %>% 
  add_trace(data = df,
            y = ~Sale,
            color = ~City,
            colors = "Dark2",
            type = "box") %>% 
  layout(title = "Sales per city",
         xaxis = list(title = "City"))
p8

Dark2 pallete 

p9 <- plot_ly() %>% 
  add_trace(data = df,
            y = ~Sale,
            color = ~City,
            colors = "BrBG",
            type = "box") %>% 
  layout(title = "Sales per city",
         xaxis = list(title = "City"))
p9

BrBG palette

RColorBrewer color palettes

The RColorBrewer package makes the ColorBrewer colors available in R. From the R documentation page for this package we read the following: There are 3 types of palettes, sequential, diverging, and qualitative. 1. Sequential palettes are suited to ordered data that progress from low to high. Lightness steps dominate the look of these schemes, with light colors for low data values to dark colors for high data values. 2. Diverging palettes put equal emphasis on mid-range critical values and extremes at both ends of the data range. The critical class or break in the middle of the legend is emphasized with light colors and low and high extremes are emphasized with dark colors that have contrasting hues. 3. Qualitative palettes do not imply magnitude differences between legend classes, and hues are used to create the primary visual differences between classes. Qualitative schemes are best suited to representing nominal or categorical data. The sequential palettes names are Blues BuGn BuPu GnBu Greens Greys Oranges OrRd PuBu PuBuGn PuRd Purples RdPu Reds YlGn YlGnBu YlOrBr YlOrRd All the sequential palettes are available in variations from 3 different values up to 9 different values. The diverging palettes are BrBG PiYG PRGn PuOr RdBu RdGy RdYlBu RdYlGn Spectral All the diverging palettes are available in variations from 3 different values up to 11 different values. For qualitative palettes, the lowest number of distinct values available always is 3, but the largest number is different for different palettes. It is given together with the palette names in the following table. lr{ Accent 8 Dark2 8 Paired 12 Pastel1 9 Pastel2 8 Set1 9 Set2 8 Set3 12 }

The ColorBrewer website is at http://colorbrewer2.org In the code chunk below we use the Paired color palette.

p10 <- plot_ly() %>% 
  add_trace(data = df,
            y = ~Sale,
            color = ~City,
            colors = brewer.pal(length(names(table(cities))),
                                "Paired"),
            type = "box") %>% 
  layout(title = "Sales per city",
         xaxis = list(title = "City"))
p10

RColorBrewer Paired palette

The number of colors vary between \(8\) and \(12\). The RColorBrewer package allows for many more colors with the colorRampPalette() function. The code chunk below shows a sine curve with \(41\) colored markers.

x <- seq(from = -2,
         to = 2,
         b = 0.1)
y <- sin(x)

p11 <- plot_ly() %>% 
  add_trace(type = "scatter",
            x = ~x,
            y = ~y,
            mode = "markers",
            marker = list(size = 10,
                          color = colorRampPalette(brewer.pal(10,"Spectral"))(41))) %>% 
  layout(title = "Multicolored sine curve",
         xaxis = list(title = "x-axis"),
         yaxis = list(title = "y-axis"))
p11

Fourty-one colors

Enjoy your colored plots and graphs with Plotly for R.