29

ggplot2 配色

色彩方案与调色板

Tidyverse 篇
R 
library(tidyverse)
library(palmerpenguins)

为了让图更好看，需要在画图中使用配色，但如果从颜色的色相、色度、明亮度三个属性（Hue-Chroma-Luminance ）开始学，感觉这样要学的东西太多了 [R表达式]. 事实上，大神们已经为我们准备好了很多好看的模板，我们可以偷懒直接拿来用[R表达式].

我个人比较喜欢colorspace中的配色，今天我们就讲讲如何使用这个宏包！

R 
library(colorspace)

colorspace 宏包提供了三种类型的配色模板:

图片
  • Qualitative: 分类，用于呈现分类信息，比如不同种类用不同的颜色，颜色之间一般对比鲜明。
  • Sequential: 序列，用于呈现有序/连续的数值信息，比如为了展示某地区黑人比例，比例越高颜色越深，比例越低颜色越浅。
  • Diverging: 分歧，用于呈现有序/连续的数值信息，这些数值围绕着一个中心值，比中心值越大的方向用一种渐变色，比中心值越小用另一种渐变色。

三种类型对应着三个函数 qualitative_hcl(), sequential_hcl(), 和 diverging_hcl().

配色模板

根据你需要颜色的三大种类，先找适合的模板palettes

R 
hcl_palettes(plot = TRUE)

或者指定某个种类，显示所有的模板

R 
hcl_palettes("qualitative", plot = TRUE)
hcl_palettes("sequential (single-hue)", n = 7, plot = TRUE)
hcl_palettes("sequential (multi-hue)", n = 7, plot = TRUE)
hcl_palettes("diverging", n = 7, plot = TRUE)

如果看中某个模板的颜色，可以获取它的值，比如sequential种类下的 "Peach"模板，

R 
#colorspace::diverging_hcl(n = 7, "Dark 2")
#colorspace::qualitative_hcl(4, palette = "myset")
colorspace::sequential_hcl(n = 7, palette = "Peach")

当然，在用之前，我们先检验下

R 
colorspace::sequential_hcl(n = 7, palette = "Peach") %>% 
  colorspace::swatchplot()

在ggplot2中使用

在ggplot2中可以免去手工操作，而直接使用。事实上， colorspace 模板使用起来很方便，有统一格式scale_<aesthetic>_<datatype>_<colorscale>(),

  • 这里 <aesthetic> 是指定映射 (fill, color, colour),
  • 这里 <datatype> 是表明数据变量的类型 (discrete, continuous, binned),
  • 这里 colorscale 是声明颜色标度类型 (qualitative, sequential, diverging, divergingx).
Scale functionAestheticData typePalette type
scale_color_discrete_qualitative()colordiscretequalitative
scale_fill_continuous_sequential()fillcontinuoussequential
scale_colour_continuous_divergingx()colourcontinuousdiverging

使用案例1

ggplot2默认

R 
penguins %>% 
   ggplot(aes(bill_length_mm, fill = species)) +
   geom_density(alpha = 0.6)

手动修改

R 
penguins %>% 
   ggplot(aes(bill_length_mm, fill = species)) +
   geom_density(alpha = 0.6) +
   scale_fill_manual(
     breaks = c("Adelie", "Chinstrap", "Gentoo"), 
     values = c("darkorange", "purple", "cyan4")
  )

使用colorspace模板配色

R 
penguins %>% 
   ggplot(aes(bill_length_mm, fill = species)) +
   geom_density(alpha = 0.6) +
   scale_fill_discrete_qualitative(palette = "cold")
R 
penguins %>% 
   ggplot(aes(bill_length_mm, fill = species)) +
   geom_density(alpha = 0.6) +
   scale_fill_discrete_qualitative(palette = "cold", nmax = 4, order = 2:4)
R 
sequential_hcl(palette = "Purples 3", n = 6) %>% 
  swatchplot()

penguins %>% 
   ggplot(aes(bill_length_mm, fill = species)) +
   geom_density(alpha = 0.6) +
   scale_fill_discrete_sequential(palette = "Purples 3", 
                                  nmax = 6, 
                                  rev = FALSE, 
                                  order = 3:5)

使用案例2

R 
temps_months <- read_csv("./demo_data/tempnormals.csv") %>%
  group_by(location, month_name) %>%
  summarize(mean = mean(temperature)) %>%
  mutate(
    month = factor(
      month_name,
      levels = c("Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec")
    ),
    location = factor(
      location, levels = c("Death Valley", "Houston", "San Diego", "Chicago")
    )
  ) %>%
  select(-month_name)

temps_months
R 
temps_months %>% 
  ggplot(aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic()
R 
temps_months %>% 
  ggplot(aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic() +
  scale_fill_gradient()
R 
temps_months %>% 
  ggplot(aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic() +
  scale_fill_viridis_c()
R 
temps_months %>% 
  ggplot(aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic() +
  scale_fill_viridis_c(option = "B", begin = 0.15)
R 
temps_months %>% 
  ggplot(aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic() +
  scale_fill_continuous_sequential(palette = "YlGnBu", rev = FALSE)
R 
temps_months %>% 
  ggplot(aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic() +
  scale_fill_continuous_sequential(palette = "Viridis", rev = FALSE)
R 
temps_months %>% 
  ggplot( aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic() +
  scale_fill_continuous_sequential(palette = "Inferno", begin = 0.15, rev = FALSE)
R 
temps_months %>% 
  ggplot(aes(x = month, y = location, fill = mean)) +
  geom_tile(width = 0.95, height = 0.95) +
  coord_fixed(expand = FALSE) +
  theme_classic() +
  scale_fill_continuous_sequential(palette = "Plasma", begin = 0.35, rev = FALSE)

配色

这里有个小小的提示：

  • 尽可能避免使用"red", "green", "blue", "cyan", "magenta", "yellow"颜色
  • 使用相对柔和的颜色"firebrick", "springgreen4", "blue3", "turquoise3", "darkorchid2", "gold2"，会让人觉得舒服

可以对比下

R 
colorspace::swatchplot(c("red", "green", "blue", "cyan", "magenta", "yellow"))
colorspace::swatchplot(c("firebrick", "springgreen4", "blue3", "turquoise3", "darkorchid2", "gold2"))

color-wheel