R语言：如何为5个量表4个类别的数据绘制堆叠条形图

• 语法错误：倒数第二行的管道符拼写错误，应为%>%而非%
• 逻辑不匹配：你提供的示例数据中各量表列的值已经是Cat1-Cat4字符串类型，无法用数值比较规则做分类判定，会直接报错
• 因子水平错误：仅有5个量表，但你给Scale设置了6个因子水平，还包含空值，会导致匹配后出现缺失值

数据预处理 #

library(tidyverse)

# 替换为你的真实数据即可，此处保留示例数据供测试
df <- tribble(
    ~ID, ~scale1,  ~scale2, ~scale3, ~scale4, ~scale5,
    "1",  "Cat1",    "Cat2",    "Cat4",   "Cat1",  "Cat2",
    "2",  "Cat3",    "Cat3",    "Cat1",   "Cat2",  "Cat3",
    "3",  "Cat4",    "Cat1",    "Cat2",   "Cat3",  "Cat2",
    "4",  "Cat1",    "Cat4",    "Cat2",   "Cat4",  "Cat4",
    "5",  "Cat2",    "Cat1",    "Cat3",   "Cat1",  "Cat1")

df_clean <- df %>%
  select(ID, scale1:scale5) %>%
  pivot_longer(cols = scale1:scale5, names_to = "Scale", values_to = "Category") %>%
  mutate(
    # 直接设置分类的因子顺序
    Category = factor(Category, levels = c("Cat4","Cat3","Cat2","Cat1")),
    # 按需修改量表的显示名称，5个量表对应5个标签即可
    Scale = factor(Scale, levels = paste0("scale", 1:5),
                   labels = c("数学","物理","计算机","体育","英语"))
  )

绘制堆叠条形图 #

ggplot(df_clean, aes(x = Scale, fill = Category)) +
  # 计数堆叠条形图，position设为"fill"可改为百分比堆叠图
  geom_bar(position = "stack") +
  labs(x = "量表名称", y = "样本数量", fill = "分类") +
  theme_bw()

如果你的真实原始数据中，量表列存储的是1-4的数值而非Cat开头的字符串，可改用以下预处理逻辑：

df_clean <- df %>%
  select(ID, scale1:scale5) %>%
  pivot_longer(cols = scale1:scale5, names_to = "Scale", values_to = "Value") %>%
  mutate(
    Value = as.numeric(Value),
    Category = case_when(
      Value >=4 ~ "Cat4",
      Value >=3 & Value <4 ~ "Cat3",
      Value >=2 & Value <3 ~ "Cat2",
      Value <2 ~ "Cat1",
      TRUE ~ "None"
    ),
    Category = factor(Category, levels = c("Cat4","Cat3","Cat2","Cat1")),
    Scale = factor(Scale, levels = paste0("scale", 1:5),
                   labels = c("数学","物理","计算机","体育","英语"))
  )

内容的提问来源于stack exchange，提问作者Hani