优化ggplot2系数图：减少Roads与Timing区域下方空白

问题说明 #

在使用ggplot2绘制的系数图中，需要减少Roads和Timing区域下方的空白，让图表布局更紧凑整洁。当前使用的代码如下：

plotdata <- structure(
  list(
    term = c("Category 1_a","Category 1_b","Category 1_c","Category 1_d","Category 2",
      "Category 3_a","Category 3_b"),
    estimate = c(-0.033882004,0.001508041,0.122957935,-0.033882004,0.001508041,-0.033882004,0.001508041),
    ymin = c(
      -0.13278953,-0.007547426,0.025116265,-0.13278953,-0.007547426,-0.13278953,-0.007547426), ymax = c(0.065025521,0.010563508,0.220799605,0.065025521,0.010563508,0.065025521,0.010563508)
  ), row.names = c(NA,-7L),
  spec = structure(list(cols = list(
      term = structure(list(), class = c("collector_character",
                                         "collector")),
      estimate = structure(list(), class = c("collector_double",
                                             "collector")),
      ymin = structure(list(), class = c("collector_double",
                                         "collector")),
      ymax = structure(list(), class = c("collector_double",
                                         "collector"))
    ),
    default = structure(list(), class = c("collector_guess",
                                          "collector")),
    delim = ","), class = "col_spec"),class = c("spec_tbl_df","tbl_df", "tbl", "data.frame"))

# Libraries
library(dplyr)
library(broom)
library(ggplot2)

# Plot
plotdata %>%
  mutate(category = rep(c("Pollution", "Roads", "Timing"), times = c(4, 1, 2))) %>%
  mutate(term = factor(term, levels = rev(term))) %>%
  ggplot(aes(x = term, y = estimate)) + 
  geom_hline(yintercept = 0, color="black", size = 0.5, linetype = "dashed") +
  geom_pointrange(aes(ymin = ymin, ymax = ymax), size = 0.5, 
                  fill="#0202fe", color="#b5b5b5", shape=21, stroke = 1) +
  labs(x = NULL, y = "Coefficient Estimate") +
  scale_x_discrete(expand = c(0.4, 0.1)) +
  coord_flip() +
  facet_grid(category~., switch = 'y', scales = 'free_y') +
  theme_classic(base_size = 13) +
  theme(strip.placement = 'outside',
        strip.background = element_blank(),
        strip.text.y.left = element_text(face = 'bold', angle = 0, vjust = 1, hjust = 1),
        panel.spacing.y = unit(0, 'mm'))

解决方案 #

问题核心是scale_x_discrete()中的expand参数设置过大，导致每个分面内的分类项上下产生过多空白。调整该参数为更小的值即可解决问题。

修改后的代码 #

plotdata %>%
  mutate(category = rep(c("Pollution", "Roads", "Timing"), times = c(4, 1, 2))) %>%
  mutate(term = factor(term, levels = rev(term))) %>%
  ggplot(aes(x = term, y = estimate)) + 
  geom_hline(yintercept = 0, color="black", size = 0.5, linetype = "dashed") +
  geom_pointrange(aes(ymin = ymin, ymax = ymax), size = 0.5, 
                  fill="#0202fe", color="#b5b5b5", shape=21, stroke = 1) +
  labs(x = NULL, y = "Coefficient Estimate") +
  # 调整expand参数，减少分类轴的空白
  scale_x_discrete(expand = c(0.1, 0.05)) +
  coord_flip() +
  facet_grid(category~., switch = 'y', scales = 'free_y') +
  theme_classic(base_size = 13) +
  theme(strip.placement = 'outside',
        strip.background = element_blank(),
        strip.text.y.left = element_text(face = 'bold', angle = 0, vjust = 1, hjust = 1),
        panel.spacing.y = unit(0, 'mm'))

关键修改说明 #

• 将scale_x_discrete(expand = c(0.4, 0.1))改为scale_x_discrete(expand = c(0.1, 0.05))：expand的两个数值分别控制轴两端的扩展比例，减小数值后，分类项与图表边缘的空白会显著减少，在Roads（仅1个分类项）和Timing（2个分类项）这类分类较少的区域效果尤为明显。
• 保留其他设置不变，确保图表的样式、分面布局等原有特性不受影响。

内容的提问来源于stack exchange，提问作者Neicooo