R语言gganimate中geom_text()柱状图标签显示异常如何解决？

异常原因 #

• 柱状图使用position = "dodge"做分列展示时，对应的数值标注geom_text未配置相同的避让规则，导致文本和柱子错位，视觉上表现为显示异常。
• 数据转换为长表后，单站点对应两条上下车记录，折线载荷标注stop_charge存在重复值，也会影响标注的渲染逻辑。

修复代码 #

核心修改点为：给柱状图对应的geom_text添加和geom_col一致的position_dodge参数、调整文本偏移位置避免和柱子重叠、单独指定原宽表数据做载荷标注避免重复值问题。

library(tidyr)
library(dplyr)
library(ggplot2)
library(gganimate)

tramway_df <- data.frame(
  stop_order = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17),
  stop_id = c(9943, 5041, 5361, 2902, 2903, 5872, 5802, 5803, 5804, 5805, 
               5710, 5711, 5712, 1643, 5714, 521, 529),
  stop_name = c("DA VINCI", "DA VINCI", "BORDET STATION", "VAN CUTSEM", "FONSON", "PAIX", 
            "TILLEUL", "HELMET", "FOYER SCHAERBEEK.", "WAELHEM", "VERBOEKHOVEN", "PAVILLON",
            "RUBENS", "LIEDTS", "THOMAS", "GARE DU NORD", "ROGIER"),
  stop_in = c(27, 102, 153, 141, 190, 317, 379, 438, 328, 214, 278, 298, 154, 144, 162, 185, 0),
  stop_out = c(0, 7, 3, 10, 22, 37, 81, 122, 143, 168, 291, 230, 149, 301, 412, 707, 821),
  stop_charge = c(27, 121, 271, 402, 571, 851, 1149, 1465, 1650, 1696, 1682, 1745, 1750, 1592, 1342, 821, 0)
)

plot <- tramway_df %>% 
  gather(movement, value, stop_in:stop_out) %>% 
  mutate(stop_id = factor(stop_id, levels = unique(stop_id)),
         value = ifelse(movement == "stop_out", -value, value)) %>% 
  ggplot(aes(x = stop_id)) +
  geom_col(aes(y = value, fill = movement), position = position_dodge(width = 1), width = 1) +
  # 柱状图标注：添加和geom_col一致的避让规则，调整文本位置避免重叠
  geom_text(aes(label = abs(value), y = value + ifelse(value>0, 15, -15), group = movement), 
            position = position_dodge(width = 1), vjust = 0.5) +
  geom_point(aes(y = stop_charge), color = "green") + 
  geom_line(aes(y = stop_charge, group=1), color = "green") + 
  # 载荷标注：单独指定原宽表数据，避免长表重复值导致的渲染异常
  geom_text(data = tramway_df, aes(label = stop_charge, y = stop_charge + 20), hjust = 0.5, vjust = 0, color = "darkgreen") +
  scale_x_discrete(breaks = tramway_df$stop_id,
                   labels = tramway_df$stop_name) +
  theme_bw() +
  theme(axis.text.x = element_text(angle = 60, hjust = 1, vjust = 1))

animation <- plot + transition_reveal(stop_order)

animate(animation, height = 500, width = 800)

其他标注数值的可选方案 #

• 用geom_label代替geom_text，自带白色背景填充，不会被下方的柱子或折线遮挡，动态场景下辨识度更高，使用时同样需要配置和柱状图一致的position_dodge参数。
• 用ggtext::geom_richtext，支持自定义文本样式、背景边框，适配gganimate的过渡效果更好，适合需要高亮标注的场景。
• 若标注逻辑复杂，可以提前生成每帧对应的标注专属数据框，在调用geom_text时单独指定data参数，不受主绘图数据结构限制。

内容的提问来源于stack exchange，提问作者salwa