ggplot2绘制折线图添加误差栏时报错：找不到'mntd.obs.z'

尝试用ggplot2绘制带误差栏的折线图，运行代码时报错，提示第4层（geom_errorbar）计算美学属性时找不到对象mntd.obs.z。mean_data数据集包含所需的mean_mntd.obs.z和sd_mntd.obs.z变量，移除geom_errorbar图层后代码可正常运行。

原代码 #

mntd_loco<-ggplot(SES_MNTD_loco, aes(x = camp, y = mntd.obs.z, color = season, group = season)) +
  geom_point(alpha = 0.3, size= 3.5) +  # 添加带透明度的散点
  geom_line(data = mean_data, aes(y = mean_mntd.obs.z), linewidth= 1.5) +  # 添加均值折线
  geom_point(data = mean_data, aes(y = mean_mntd.obs.z), size = 5) +  # 添加均值点
  geom_errorbar(data = mean_data, aes(ymin = mean_mntd.obs.z-sd_mntd.obs.z, ymax= mean_mntd.obs.z + sd_mntd.obs.z ), linewidth= 1.5)+
  geom_hline(yintercept = 0, linetype = "longdash", color = "black", linewidth= 2) +
  labs(x = "海拔", y = "SES MNTD", color = "季节") +  # 设置坐标轴和图例标签
  scale_color_manual(values=c('#999999','#E69F00')) +
  theme_classic()

报错信息 #

Error in `geom_errorbar()`:
! Problem while computing aesthetics.
ℹ Error occurred in the 4th layer.
Caused by error in `FUN()`:
! object 'mntd.obs.z' not found

mean_data数据集 #

> mean_data
# A tibble: 8 × 4
# Groups:   camp [4]
  camp  season mean_mntd.obs.z sd_mntd.obs.z
  <fct> <chr>            <dbl>         <dbl>
1 DG    dry             -0.413         0.607
2 DG    wet             -0.230         0.346
3 PC    dry             -1.08          0.302
4 PC    wet             -0.974         0.417
5 CL    dry             -1.08          0.173
6 CL    wet             -1.07          0.235
7 MS    dry             -0.299         0.690
8 MS    wet             -0.382         0.61

生成mean_data的代码 #

mean_data <- SES_MNTD_loco %>%
  group_by(camp, season) %>%
  summarize(mean_mntd.obs.z = mean(mntd.obs.z),
            sd_mntd.obs.z = sd(mntd.obs.z)) 

SES_MNTD_loco数据集前6行 #

> head(SES_MNTD_loco)
         ntaxa mntd.obs mntd.rand.mean mntd.rand.sd mntd.obs.rank mntd.obs.z mntd.obs.p camp season
DG_dry.1    36 17.59933       20.40015     6.001859           355 -0.4666597 0.35464535   DG    dry
DG_dry.2    35 21.60363       20.52625     6.591346           643  0.1634539 0.64235764   DG    dry
DG_dry.3    43 13.62599       19.26860     5.472003           103 -1.0311768 0.10289710   DG    dry
DG_dry.4    35 23.39255       20.61614     7.202468           754  0.3854798 0.75324675   DG    dry
DG_dry.5    42 12.40775       19.34753     6.245307            64 -1.1111996 0.06393606   DG    dry
DG_dry.6    40 17.04743       19.79749     6.598712           389 -0.4167577 0.38861139   DG    dry

解决方案 #

报错核心原因是：ggplot2会继承全局设置的美学映射（aes(y = mntd.obs.z)），但mean_data中不存在mntd.obs.z变量，geom_errorbar继承该映射后，会尝试在mean_data里查找这个变量，从而触发报错。

有两种可行的解决方法：

• 方法一：显式指定geom_errorbar的y映射
  在geom_errorbar的aes中补充y = mean_mntd.obs.z，覆盖全局的y映射，其他美学属性（x、color、group）可正常继承：

  geom_errorbar(data = mean_data, aes(y = mean_mntd.obs.z, ymin = mean_mntd.obs.z-sd_mntd.obs.z, ymax= mean_mntd.obs.z + sd_mntd.obs.z ), linewidth= 1.5)
  

• 方法二：关闭美学映射继承
  在geom_errorbar中添加inherit.aes = FALSE，同时手动指定所需的所有映射（x、color、group）：

  geom_errorbar(data = mean_data, aes(x = camp, ymin = mean_mntd.obs.z-sd_mntd.obs.z, ymax= mean_mntd.obs.z + sd_mntd.obs.z, color = season, group = season), linewidth= 1.5, inherit.aes = FALSE)
  

推荐使用方法一，代码更简洁，仅需补充必要的y映射即可解决问题。

内容的提问来源于stack exchange，提问作者Guillermo Uceda