R语言堆叠图：调整类别顺序并移除图例中的'a'

一、调整堆叠顺序（让'Mandatory School'在柱子底部） #

核心是控制学历分类的顺序，不同绘图工具的处理方式如下：

1. ggplot2 场景 #

把学历变量转换成有序因子，将'Mandatory School'设为第一个水平（ggplot默认按因子水平从下到上堆叠）：

# 假设数据框为df，学历列名为education
df$education <- factor(df$education,
                       # 按学历从低到高排序，最高学历放最后
                       levels = c("Mandatory School", "High School", "College", "Graduate"),
                       ordered = TRUE)

后续用geom_col()绘图时，'Mandatory School'会自动处于柱子底部。

2. 基础绘图（barplot）场景 #

调整数据矩阵的行顺序，将'Mandatory School'对应的行放在第一行（base R的barplot默认按行顺序从下到上堆叠）：

# 假设数据为矩阵df_mat，行是学历、列是分组
df_mat <- df_mat[c("Mandatory School", "High School", "College", "Graduate"), ]

二、移除图例中多余的"a"字符 #

分两种常见情况处理：

1. 因子水平自带"a" #

先检查学历变量的水平：

levels(df$education)

如果输出包含"aMandatory School"这类带a的内容，直接修改因子标签：

df$education <- factor(df$education,
                       levels = c("aMandatory School", "aHigh School", ...), # 原始带a的水平
                       labels = c("Mandatory School", "High School", ...)) # 去掉a的新标签

2. 图例标题错误显示为"a" #

这通常是因为你在aes()里误写了fill=a(education)，改成fill=education后，用labs()设置正确的图例标题：

ggplot(df, aes(x=分组列名, y=数值列名, fill=education)) +
  geom_col() +
  labs(fill = "学历") # 替换为你需要的标题

完整示例（ggplot2） #

library(ggplot2)

# 模拟数据集
df <- data.frame(
  group = rep(c("城市", "农村"), each=4),
  education = rep(c("aMandatory School", "High School", "College", "Graduate"), 2),
  count = c(35,25,20,20, 45,20,15,20)
)

# 处理因子顺序和标签
df$education <- factor(df$education,
                       levels = c("aMandatory School", "High School", "College", "Graduate"),
                       labels = c("Mandatory School", "High School", "College", "Graduate"),
                       ordered = TRUE)

# 绘图
ggplot(df, aes(x=group, y=count, fill=education)) +
  geom_col() +
  labs(title = "不同群体学历分布", x="群体", y="人数", fill="学历") +
  theme_classic()

内容的提问来源于stack exchange，提问作者TFT