在条形图上垂直对齐 geom_text 图层

Question

我目前将 geom_text 设置为条形图的中心...

library(dplyr)
library(ggplot2)

enroll_bar <- enroll_cohort %>%
    filter(chrt_grad != 2013) %>%
    mutate(college_enrolled = factor(college_enrolled),
           chrt_grad = factor(chrt_grad)) %>%
    mutate(label_height = cumsum(n)) %>%
    ggplot() +
    geom_col(mapping = aes(x = chrt_grad,
                           y = n,
                           fill = fct_rev(college_enrolled))) +
    geom_text(aes(x = chrt_grad, y = label_height, label = n), color = "white",
              position = position_stack(vjust = 0.5)) +
    scale_y_continuous(expand = expansion(mult = c(0, 0.1))) +
    labs(x = NULL, y = NULL) +
    scale_fill_manual(labels = c("Enrolled", "Not Enrolled"),
                      values = c("#00aeff", "#005488"))

如何垂直对齐文本，以便将 geom_text 图层设置在每个栏的相同 y 轴位置？另外，有没有办法做到这一点，无论 y 轴的比例如何，文本始终对齐，因为这是一个参数化报告，y 轴值随每个报告而变化？

Answer 1

如果您希望标签的高度统一，您可以在 geom_text() 外的 y 中指定一个 y 和您要查找的高度。

我使用了数据集diamonds。列太长了 non-uniform，白色文本不可见，所以我将顶部标签更改为黑色。

因为底部组的最低值真的很低，所以我将其设置为 100。顶部组设置为 5000。本质上，我创建了一个向量 -- 100, 5000, 100, 5000, 100... 和依此类推，使用 rep()（重复）。

数据设置不同，因为您没有将其包含在您的问题中。不过，这应该无关紧要。

图书馆（tidyverse）

diamonds %>% 
  filter(color %in% c("E", "I")) %>% 
  ggplot(aes(x = cut,
             fill = color)) + 
  geom_bar() +
  geom_text(aes(label = ..count..),  # use the count
            y = rep(c(100, 5000),    # btm label y = 100, top label y = 5K
                    times = 5),
            stat = 'count',
            color = rep(c("white", "black"), # btm label white, top black
                        times = 5),
            position = position_stack(vjust = .5)) +
  scale_y_continuous(expand = expansion(mult = c(0, 0.1))) +
  labs(x = NULL, y = NULL) +
  scale_fill_manual(labels = c("E", "I"),
                    values = c("#00aeff", "#005488")) +
  theme_bw()

Answer 2

您可以使用 if_else（或 case_when 用于 >2 个组）为每个组设置统一的标签高度。对于单个图，您可以简单地设置一个值，例如 label_height = if_else(college_enrolled == "Enrolled", 20000, 3000)。要使多个绘图的相对高度一致，您可以将 label_height 设置为 y-axis 范围的比例：

library(tidyverse)

# make a fake dataset
enroll_cohort <- expand_grid(
    chrt_grad = factor(2014:2021),
    college_enrolled = factor(c("Enrolled", "Not Enrolled")),
  ) %>%
  mutate(
    n = sample(18000:26000, 16),
    n = if_else(college_enrolled == "Enrolled", n, as.integer(n / 3))
  )

enroll_bar <- enroll_cohort %>% 
  group_by(chrt_grad) %>%            # find each bar's height by summing up `n`
  mutate(bar_height = sum(n)) %>%    # within each year
  ungroup() %>%
  mutate(label_height = if_else(
    college_enrolled == "Enrolled", 
    max(bar_height) * .6,            # axis height is max() of bar heights;
    max(bar_height) * .1             # set label_height as % of axis height
  )) %>%
  ggplot() +
  geom_col(aes(x = chrt_grad, y = n, fill = college_enrolled), color = NA) +
  geom_text(
    aes(x = chrt_grad, y = label_height, label = n), 
    color = "white"
  ) +
  scale_y_continuous(expand = expansion(mult = c(0, 0.1))) +
  labs(x = NULL, y = NULL) +
  scale_fill_manual(values = c("#00aeff", "#005488"))

如果我们生成另一个具有不同 n 值范围的数据集——例如，~1200 - ~2000——文本标签保持在相同的相对位置：

在条形图上垂直对齐 geom_text 图层

Aligning a geom_text layer vertically on a bar chart

r

ggplot2

geom-text