如何设置有序Labels为X轴并以Treatments为图例绘制产量图？

我需要分析以下农业产量数据并绘制统计图：

Labels  Treatments  Reps    Yield
T1  No fertilizer   1   8
T2  Ferti 1         1   12
T3  Ferti 2         1   14
T4  Ferti 3         1   16
T5  Ferti 4         1   18
T6  Ferti 5         1   16
T7  Ferti 6         1   24
T8  Ferti 7         1   22
T9  Ferti 8         1   26
T10 Ferti 9         1   30
T11 Ferti 10        1   32
T1  No fertilizer   2   8.2
T2  Ferti 1         2   12.8
T3  Ferti 2         2   13.8
T4  Ferti 3         2   16.6
T5  Ferti 4         2   18
T6  Ferti 5         2   16.6
T7  Ferti 6         2   25.2
T8  Ferti 7         2   22.5
T9  Ferti 8         2   26.4
T10 Ferti 9         2   31
T11 Ferti 10        2   33.2
T1  No fertilizer   3   7.8
T2  Ferti 1         3   12.4
T3  Ferti 2         3   14.1
T4  Ferti 3         3   16.2
T5  Ferti 4         3   18.8
T6  Ferti 5         3   16.9
T7  Ferti 6         3   26.4
T8  Ferti 7         3   22.5
T9  Ferti 8         3   27
T10 Ferti 9         3   32.3
T11 Ferti 10        3   34.5

需求：

• 执行Yield ~ Labels的ANOVA分析
• 绘制统计图时X轴按T1、T2……T11的顺序排列
• 以Treatments列作为图例（即T1对应"No fertilizer"，T2对应"Ferti 1"等）

遇到的问题：仅用Labels和Yield绘图时X轴顺序正确，但添加Treatments作为图例后X轴顺序混乱。

已尝试的R代码：

library(readxl)
library(multcompView)
library(dplyr)
library(ggplot2)

my_data <- read_excel(choose.files(), sheet = "Yield")

# 将Labels和Treatments列转为因子
my_data <- read_excel(file.choose(), sheet = 1)

my_data <- my_data |> mutate(Labels = factor(Labels, 
                         labels = c("T1", "T2", "T3", "T4", "T5", "T6", "T7", "T8", "T9", "T10", "T11"))) |> 
  mutate(Treatments = factor(Treatments, 
                             labels = c("No fertilizer", "Ferti 1","Ferti 2","Ferti 3","Ferti 4","Ferti 5",
                                        "Ferti 6","Ferti 7","Ferti 8","Ferti 9","Ferti 10")))

yield_aov <- aov(Yield ~ Labels, data = my_data)

yield_aov1 <- anova(yield_aov)

tukey_L <- TukeyHSD(yield_aov)

tukey_letters <- multcompLetters4(yield_aov, tukey_L)

table_yield <- my_data %>% group_by(Labels) %>% 
  summarise(Average = mean(Yield), .groups = 'drop')

myletters <- as.data.frame.list(tukey_letters$Labels)

table_yield$Letters <- myletters$Letters

table_yield <- arrange(table_yield, Labels)

table_yield %>% ggplot(aes(Labels, Average, fill = Labels)) +
  geom_bar(stat = "identity", colour = "black", width = 0.5) + 
  labs(x = "Treatments", 
       y = "Yield (q/ha)")

# 添加Treatments为图例后的绘图代码
table_yield %>% ggplot(aes(Labels, Average, fill = Labels)) +
  geom_bar(stat = "identity", colour = "black", width = 0.5) + 
  labs(x = "Treatments", 
       y = "Yield") + 
  scale_x_discrete(labels = unique(my_data$Treatments))

请问如何实现T1对应“No fertilizer”、T2对应“Ferti 1”……的映射，并保持X轴按T1到T11的顺序排列？

核心问题分析 #

使用scale_x_discrete(labels = unique(my_data$Treatments))时，unique(my_data$Treatments)的提取顺序无法保证和Labels的T1-T11严格对应，会导致X轴标签错位，同时破坏Labels的因子顺序，最终造成X轴混乱。

修正步骤 #

1. 强制Labels的因子顺序：明确指定Labels的水平为T1到T11，从底层保证X轴排序。
2. 保留Labels与Treatments的对应关系：通过分组提取每个Labels对应的唯一Treatments值，避免映射错位。
3. 精准替换X轴标签：用已匹配好顺序的Treatments值替换X轴显示标签，同时将Treatments绑定为图例变量。

完整修正代码 #

library(readxl)
library(multcompView)
library(dplyr)
library(ggplot2)

# 读取数据（合并重复读取的代码）
my_data <- read_excel(file.choose(), sheet = 1)

# 处理因子与对应关系：强制Labels顺序，绑定每个Labels对应的Treatments
my_data <- my_data |>
  mutate(Labels = factor(Labels, levels = paste0("T", 1:11))) |>  # 固定T1-T11的顺序
  group_by(Labels) |>
  mutate(Treatments = first(Treatments)) |>  # 确保每个Labels对应唯一的Treatments
  ungroup()

# ANOVA与显著性分析（原逻辑不变）
yield_aov <- aov(Yield ~ Labels, data = my_data)
yield_aov1 <- anova(yield_aov)
tukey_L <- TukeyHSD(yield_aov)
tukey_letters <- multcompLetters4(yield_aov, tukey_L)

# 汇总均值、Treatments与显著性字母
table_yield <- my_data %>% 
  group_by(Labels, Treatments) %>%  # 同时分组保留对应关系
  summarise(Average = mean(Yield), .groups = 'drop')

myletters <- as.data.frame.list(tukey_letters$Labels)
table_yield$Letters <- myletters$Letters

# 绘图：保证X轴顺序，绑定Treatments为图例
table_yield %>% 
  ggplot(aes(x = Labels, y = Average, fill = Treatments)) +  # fill绑定Treatments作为图例
  geom_bar(stat = "identity", colour = "black", width = 0.5) +
  geom_text(aes(label = Letters), vjust = -0.3, size = 4) +  # 添加显著性字母（可选）
  scale_x_discrete(labels = table_yield$Treatments) +  # 用匹配好的Treatments替换X轴标签
  labs(x = "Treatments", y = "Yield (q/ha)", fill = "Treatments") +  # 设置图例标题
  theme(axis.text.x = element_text(angle = 45, hjust = 1))  # 可选：旋转X轴标签避免重叠

关键说明 #

• 因子水平固定：factor(Labels, levels = paste0("T", 1:11))强制Labels按T1到T11排序，这是X轴顺序正确的核心保障。
• 对应关系绑定：分组时同时包含Labels和Treatments，确保每个Treatments都精准匹配对应的Labels，不会错位。
• 图例与标签同步：fill = Treatments直接将图例与处理名称关联，scale_x_discrete使用已排序的Treatments作为X轴标签，两者顺序完全一致。

内容的提问来源于stack exchange，提问作者Vinod