ggplot2添加年度索赔总额文本报错：bands2对象未找到

错误原因 #

你在ggplot的全局美学映射里设置了fill = bands2, color = bands2，但第二个geom_text使用的yearly_totals数据集里并没有bands2这一列。由于geom_text默认会继承全局的美学映射，程序试图在yearly_totals里查找bands2，自然找不到就报错了。

修复方案 #

有两种简单的修复方式：

方案1：禁用美学继承 #

给第二个geom_text添加inherit.aes = FALSE参数，让它不继承全局的fill和color映射：

yearly_totals <- df_claims %>%
  group_by(AccidentYear) %>%
  summarise(total = n())

gg_df_claims <- df_claims %>%
  ggplot(aes(AccidentYear, fill = bands2, color = bands2)) +
  geom_bar(position = position_stack(reverse = TRUE), stat = "count") +
  geom_text(
    aes(label = after_stat(count), y = after_stat(count)),
    stat = "count",
    position = position_stack(reverse = TRUE, vjust = 0.5),
    size = 2,
    color = "white"
  ) +
  geom_text(
    data = yearly_totals,
    aes(label = total, y = total),
    inherit.aes = FALSE,  # 关键：禁用全局美学继承
    position = position_nudge(y = 0.3),  # 可选：调整位置避免和柱子文本重叠
    size = 2,
    color = "black",
    fontface = "bold"
  ) +
  scale_fill_manual(values = v_greens, labels = labels2) +
  scale_color_manual(values = v_greens) +
  xlab(NULL) +
  ylab("Count of Claims") +
  theme_pubclean() +
  theme(
    panel.grid.major.y = element_line(color = "gray", linetype = "dashed"),
    panel.grid.minor.y = element_blank()
  )

方案2：覆盖全局美学映射 #

在第二个geom_text的aes里显式设置fill和color为NULL，覆盖全局的映射：

# ... 前面代码保持不变
geom_text(
  data = yearly_totals,
  aes(label = total, y = total, fill = NULL, color = NULL),  # 覆盖全局美学
  position = position_nudge(y = 0.3),
  size = 2,
  color = "black",
  fontface = "bold"
)
# ... 后面代码保持不变

额外提示 #

建议给年度总额文本添加position_nudge(y = 0.3)这类位置调整，避免和柱子顶部的小计文本重叠，让图表可读性更强。

内容的提问来源于stack exchange，提问作者Florian