代码之家 › 专栏 › 技术社区 › John Conor

为面板组设置单独的配色方案+ggplot2+facet_wrap

facet visualization ggplot2 dataframe r

John Conor · 技术社区 · 3 年前

我正在尝试使用ggplot 2为面板组分配单独的颜色方案。在下面的例子中,我的目标是A和B将具有相同的托盘,C将具有不同的托盘,D和E将具有相同托盘。

# Number of IDs
n_ids <- 10 

# Generate IDs
ids <- paste0("ID_", 1:n_ids)

# Number of periods per ID
n_periods <- 3

# Generate periods
periods <- rep(c("P1", "P2", "P3"), each = n_ids)

# Number of categories 
n_cats <- 5 

# Generate categories
categories <- rep(LETTERS[1:n_cats], times = n_ids * n_periods)

# Generate values
values <- rnorm(n_ids * n_periods * n_cats, 100, 15)

# Assemble into dataframe
df <- data.frame(ID = ids, 
                 Period = periods,
                 Category = categories,
                 Value = values)

# Plot
p <- ggplot(df, aes(x=Period, y=Value, fill=Period)) +
  geom_boxplot() +
  geom_jitter(width=0.2,alpha=0.4, size=1) +
  facet_wrap(~Category, ncol=6, scales='free') +
  labs(title="Repetitive Behavior Score", x="Group", y="Value")

# Color vector
colors <- c("#1F78B4", "#E69F00", "#33A02C", "#FB9A99")

# Map colors to categories
color_map <- c(
  "A" = "#1F78B4",
  "B" = "#1F78B4",
  
  "C" = "#E69F00",
  
  "D" = "#33A02C",
  "E" = "#33A02C"
)

# Set fill scale
p + 
  scale_fill_manual(values = color_map)

但他们都只是恢复了灰色。。。以前有人试过这个吗?我知道这是可以做到的 scale_fill 和 scale_color 但我找不到镶嵌面板的例子。

2 回复 | 直到 3 年前

Jon Spring 3 年前

另一种方法可以是将情节与拼凑相结合:

library(ggplot2)
library(patchwork)
plot_func <- function(df1, title = "Repetitive Behavior Score") {
  ggplot(df1, aes(x=Period, y=Value, fill=Period)) +
  geom_boxplot() +
  geom_jitter(width=0.2,alpha=0.4, size=1) +
  facet_wrap(~Category, ncol=6, scales='free') +
  labs(title=title, x="Group", y="Value") 
}

plot_func(subset(df, Category %in% c("A", "B"))) + 
  plot_func(subset(df, Category %in% c("C")), title = "") +
  scale_fill_viridis_d() +
  plot_func(subset(df, Category %in% c("D", "E")), title = "") + 
  scale_fill_brewer(palette = "Set3") +
  plot_layout(widths = c(2,1,2))

Allan Cameron 3 年前

您可以将填充美学映射到 interaction 你的x轴和面变量。缺点是需要手动指定颜色。

ggplot(df, aes(Period, Value, fill = interaction(Period, Category))) +
  geom_boxplot() +
  geom_jitter(width = 0.2, alpha = 0.4, size = 1) +
  facet_wrap(~Category, ncol = 6, scales = 'free') +
  labs(title = "Repetitive Behavior Score", x = "Group", y =" Value") +
  scale_fill_manual(values = c(rep(c("red4", "green4", "blue4"), 2),
                               c("hotpink", "orange", "dodgerblue"),
                               rep(c("purple3", "beige", "gray80"), 2)),
                    guide = guide_none()) +
  jtools::theme_apa()

John Conor 3 年前

事实上,这很容易,我所需要做的就是改变 fill 在里面 ggplot(df, aes(x=Period, y=Value, fill=Category)) 到 Category 。见下文:

# Number of IDs
n_ids <- 10 

# Generate IDs
ids <- paste0("ID_", 1:n_ids)

# Number of periods per ID
n_periods <- 3

# Generate periods
periods <- rep(c("P1", "P2", "P3"), each = n_ids)

# Number of categories 
n_cats <- 5 

# Generate categories
categories <- rep(LETTERS[1:n_cats], times = n_ids * n_periods)

# Generate values
values <- rnorm(n_ids * n_periods * n_cats, 100, 15)

# Assemble into dataframe
df <- data.frame(ID = ids, 
                 Period = periods,
                 Category = categories,
                 Value = values)

# Plot
p <- ggplot(df, aes(x=Period, y=Value, fill=Category)) +
  geom_boxplot() +
  geom_jitter(width=0.2,alpha=0.4, size=1) +
  facet_wrap(~Category, ncol=6, scales='free') +
  labs(title="Repetitive Behavior Score", x="Group", y="Value") +
  theme_apa() 

# Color vector
colors <- c("#1F78B4", "#E69F00", "#33A02C", "#FB9A99")

# Map colors to categories
color_map <- c(
  "A" = "#1F78B4",
  "B" = "#1F78B4",
  
  "C" = "#E69F00",
  
  "D" = "#33A02C",
  "E" = "#33A02C"
)

# Set fill scale
p + scale_fill_manual(values = color_map)