Color recipes

This vignette shows how various color-related tasks can be solved by the tools provided by unikn. (For a general introduction of the unikn colors, color palettes, and color functions, see the vignette on Colors and color functions. The vignette on text provides information on creating text boxes and decorations.)

The following recipes illustrate how plotting functions in R can use the colors and color palettes provided by the unikn package. Additional examples show how we can use the seecol(), usecol(), newpal(), and grepal() functions for solving color-related tasks.

Using colors in base R

Task: Using unikn colors in visualizations created by base R graphics.

When creating base R visualizations (e.g., using plot() for creating a scatterplot), use the usecol() function in combination with any color palette (e.g., pal_unikn) as the col argument:

my_col <- usecol(pal_unikn, alpha = .50)  # with transparency

plot(x = runif(99), y = runif(99), type = "p", 
     pch = 16, cex = 4,
     col = my_col,
     main = "99 transparent dots", axes = FALSE, xlab = NA, ylab = NA)

Using colors in ggplot2

Task: Using unikn colors in visualizations created by the ggplot2 package.

When using the ggplot() function of ggplot2 (e.g., for creating an area plot), use the usecol() function for defining a color palette (of the desired length and transparency) that is then provided as the values of the scale_color_manual() or scale_fill_manual() functions:

Wrap the desired color palette in the usecol() function of unikn.
Provide this palette as the values of the ggplot2 functions scale_color_manual() or scale_fill_manual().

# 0. Create data: ---- 

# Example based on https://www.r-graph-gallery.com/137-spring-shapes-data-art/
n <- 50
groups <- 1:n
df <- data.frame()

set.seed(3)

for (i in seq(1:30)){
  data = data.frame(matrix(0, n, 3))
  data[, 1] <- i
  data[, 2] <- sample(groups, nrow(data))
  data[, 3] <- prop.table(sample(c(rep(0, 100), c(1:n)), nrow(data)))
  df = rbind(df, data)}

names(df) <- c("X","Group","Y")
df$Group <- as.factor(df$Group)  
df <- df[c(2, 1, 3)]
df <- df[order(df$X, df$Group) , ]
rownames(df) <- NULL

my_data <- df

# 1. Colors: ---- 

# A. using RColorBrewer: 
# library(RColorBrewer)

# cur_col <- brewer.pal(11, "Paired") 
# cur_col <- colorRampPalette(cur_col)(n)
# cur_col <- cur_col[sample(c(1:length(cur_col)), size = length(cur_col))]  # randomize

# B. using unikn colors:
library(unikn)

# Mix a color gradient: 
cur_col <- usecol(c(Bordeaux, "white", Petrol), n = n)
# cur_col <- cur_col[sample(c(1:length(cur_col)), size = length(cur_col))]  # randomize

# 2. Plotting: ---- 

library(ggplot2)

ggplot(my_data, aes(x = X, y = Y, fill = Group)) + 
  geom_area() +
  scale_fill_manual(values = cur_col) +
  theme_void() +
  theme(legend.position = "none")

pal_unikn

Viewing and comparing color palettes

Task: Inspect (the colors and details of) a color palette.

For easily viewing an individual color palette (e.g., provided by the new HCL color palettes hcl.pals() of R), use the palette as the pal argument of the seecol() function:

col_pal <- hcl.colors(10, "Vik")  # some color palette

seecol(col_pal)  # see its colors and details

Note that the other arguments of the seecol() function — especially n and alpha — can modify the palette (in the same ways as the usecol() function).

Task: Compare (the colors and range of) multiple color palettes.

For comparing multiple color palettes (e.g., 20 random color palettes of hcl.pals()), provide them as a list in the pal argument of the seecol() function.

For instance, let’s create and compare a set of 20 random HCL palettes:

Generate random HCL palettes (and save them as a list):

# Parameters:
n_col <-  8   # colors in each palette
n_pal <- 20   # length(hcl.pals())
ptype <- NULL # c("diverging", "divergingx", "qualitative", "sequential")[4] 

s_pal <- sample(x = hcl.pals(type = ptype), size = n_pal, replace = FALSE)
s_pal <- sort(s_pal)

# Initialize:
t_lbl <- paste0("See ", n_pal, " random HCL color palettes (n = ", n_col, ")") 
l_pal <- vector("list", n_pal)  # lists of palettes

# Loop:
for (i in 1:n_pal){
  
  l_pal[[i]] <- hcl.colors(n = n_col, palette = s_pal[i])
  
}

Use seecol() for comparing multiple color palettes:

seecol(pal = l_pal, 
       pal_names = s_pal,
       col_brd = "white", lwd_brd = 1,
       main = t_lbl)

Finding similar colors

Task: Find a color that is similar to some given color.

Use the simcol() function for finding (and plotting/seeing) similar colors:

simcol("forestgreen")
#>    forestgreen    chartreuse4         green4      olivedrab     olivedrab4 
#>  "forestgreen"  "chartreuse4"       "green4"    "olivedrab"   "olivedrab4" 
#>     palegreen4   springgreen4 
#>   "palegreen4" "springgreen4"

The simcol() function compares a require target color (col_target) with a set of candidate colors (col_candidates, set to colors() by default). Color similarity is defined in terms of the colors’ pair-wise distances in RGB values. If this distance falls below some tolerance value(s) (tol), the corresponding col_candidates are shown and returned. Thus, increasing the tolerance value yields a wider range of colors:

simcol(col_target = "orange", col_candidates = pal_unikn_pref, tol = 150)
#>    orange     Pinky     Peach    Signal  Bordeaux 
#>  "orange" "#E0607E" "#FEA090" "#EFDC60" "#8E2043"

Specifying a vector of three tol values compares RGB dimensions in the order of their rank. This allows for more fine-grained similarity searches:

simcol(col_target = Seeblau, tol = c(20, 20, 80))
#>         Seeblau    lightskyblue   lightskyblue2         skyblue        skyblue1 
#>       "#59C7EB"  "lightskyblue" "lightskyblue2"       "skyblue"      "skyblue1" 
#>        skyblue2      steelblue1 
#>      "skyblue2"    "steelblue1"

Creating new color palettes

Task: Create a new color palette (with color names).

The general steps for creating an new color palette (with dedicated color names) are:

Choose some colors (and their R color names, or as RGB/HEX/HCL values, e.g., from sites like https://www.schemecolor.com).
Define the colors as an R vector (of type character).
Define optional color names as a second R vector (of type character).
Use the newpal() command to define the new color palette (as an R data frame or vector).

Once a new palette has been defined, it can be inspected (with seecol()) or further modified and used in visualizations (with usecol()):

# 1. Choose colors:
# Google logo colors (from <https://www.schemecolor.com/google-logo-colors.php>)

# 2. Define colors (as vector):
color_google <- c("#4285f4", "#34a853", "#fbbc05", "#ea4335")

# 3. Define color names (as vector):
names_google <- c("blueberry", "sea green", "selective yellow", "cinnabar")

# 4. Define color palette:
pal_google <- newpal(color_google, names_google)

# Inspect color palette:
seecol(pal_google, 
       col_brd = "white", lwd_brd = 8,
       main = "Colors of the Google logo")

The vignette on Institutional colors provides additional examples for defining color palettes.

Finding colors by name

Task: Find colors by matching their names to a pattern (as a regular expression).

Here are some additional examples for using the grepal() function for finding colors by their names (or a pattern):

# Get color palettes matching a pattern:
pal_1 <- grepal("orange", plot = FALSE)
pal_2 <- grepal("olive", plot = FALSE)
pal_3 <- grepal("white", plot = FALSE)

# See individual palettes:
# seecol(pal_1, main = "Hues of 'orange' colors()")
# seecol(pal_2, main = "Hues of 'olive' colors()")
seecol(pal_3, main = "Hues of 'white' colors()", col_bg = "grey90")

Providing a list of color palettes to the pal argument of the seecol() function allows comparing multiple color palettes:

# See multiple color palettes:
seecol(pal = list(pal_1, pal_2, pal_3), 
       pal_names = c("orange", "olive", "white"), 
       col_bg = "grey90")

The last example illustrates that (a) the pattern argument of grepal() can use regular expressions, and that (b) the x argument of grepal() works with vectors or data frames (e.g., the unikn color palettes).

Let’s compare different types of “blue”, “orange”, and “purple or violet” provided by the grDevices vector of named colors() with various shades of “blau”, “pink”, “peach” and “bordeaux” provided by unikn color palettes:

# Search colors() with specific patterns (in their names): 
blues    <- grepal("blue$", plot = FALSE)   # ending on "blue"
oranges  <- grepal("orange", plot = FALSE)  # containing "orange"
purpviol <- grepal("purple|violet", plot = FALSE)  # containing "purple" or "violet"

# Search unikn palettes for color names: 
blaus_1 <- grepal("blau", pal_unikn, plot = FALSE)
blaus_2 <- grepal("blau", pal_karpfenblau, plot = FALSE)
pinks <- grepal("pink", pal_pinky, plot = FALSE)
peach <- grepal("peach", pal_peach, plot = FALSE)
baux  <- grepal("bordeaux", pal_bordeaux, plot = FALSE)

# See multiple color palettes: 
seecol(list(blues, oranges, purpviol, 
            c(blaus_1, blaus_2, pinks, peach, baux)),
       pal_names = c("blues", "oranges", "purpviol", "unikn colors"),
       main ="Comparing custom color palettes")

Getting shades of a color

Task: Get different shades of a specific color.

The shades_of() function returns a vector of colors that are shades (i.e., typically lighter or darker versions) of an initial color .

By default, shades_of(n) yields n shades ranging from col_1 = "black" to col_n = "white (i.e., n shades of grey):

greys <- shades_of(10)
seecol(greys, main = "10 shades of grey")

By adjusting n, the initial color col_1, and/or the final color col_n, we can conveniently create simple color ranges:

seecol(shades_of(4, Seeblau, "black"), main = "4 shades of Seeblau")
seecol(shades_of(5, "forestgreen"), main = "5 shades of 'forestgreen'")

Using alpha in the range \([0; 1]\) allows adjusting the transparency of the color gradient:

wine_gold <- shades_of(4, Bordeaux, col_n = "gold", alpha = .75)
seecol(wine_gold, main = "5 transparent shades from Bordeaux to gold")

Creating color gradients

Task: Creating color gradients.

The shades_of() function is merely a convenient wrapper function for the usecol() function. Using the pal argument of usecol() allows defining more complex color gradients (e.g., by specifying more than two colors). For instance, to create a bi-polar (divergent) color palette, we could mix two main colors with “white” and then remove the middle color:

# Create bi-polar color gradient (and remove 'white' middle color):
divergent_gradient <- usecol(pal = c("firebrick", "white", "steelblue"), n = 7)[-4]

seecol(divergent_gradient, main = "A bi-polar color palette")

Type:	Version:	URL:
A. unikn (R package):	Release version	https://CRAN.R-project.org/package=unikn
	Development version	https://github.com/hneth/unikn/
B. Online documentation:	Release version	https://hneth.github.io/unikn/
	Development version	https://hneth.github.io/unikn/dev/

Color recipes

SPDS, uni.kn

2022 11 30