如何按指定表型分组并按频率高低重新排序条形图

问题核心 #

你的代码未达到预期效果的原因是：仅调整了Phenotype的因子顺序，但x轴的Genotype未按「指定表型分组+组内频率降序」的规则设置为有序因子，导致条形图的排序不符合要求。

修改后的完整代码 #

library(tidyverse)
library(ggpubr)

# 示例数据集
df1 <- data.frame(Sample = c("PT1", "PT2", "PT3", "PT4", "PT5", "PT6", "PT7", "PT8", "PT9", "PT10"),
                  Genotype = c("AA/BB", "AA/CC", "BB/CC", "AA/DD", "AA/AA", "BB/CC", "BB/DD", "CC/DD", "BB/BB", "AA/DD"),
                  Phenotype = c("Red", "Blue", "Pink", "Blue", "Blue", "Pink", "Pink", "White", "Red", "Blue"))

# 数据预处理：统计频率+设置有序因子
df2 <- df1 %>%
  select(Genotype, Phenotype) %>%
  # 统计每个基因型-表型组合的频率
  count(Phenotype, Genotype, name = "Freq") %>%
  filter(Freq > 0) %>%
  # 先按指定表型顺序排序，再在每个表型内按频率降序排序
  arrange(factor(Phenotype, levels = c("Blue", "Red", "Pink", "White")), desc(Freq)) %>%
  # 将Genotype转换为有序因子，顺序沿用arrange后的结果
  mutate(Genotype = fct_inorder(Genotype),
         # 设置Phenotype的因子顺序，保证填充色与分组对应
         Phenotype = fct_relevel(Phenotype, c("Blue", "Red", "Pink", "White")))

# 绘制条形图
ggplot(df2, aes(x = str_wrap(Genotype, 5), y = Freq, fill = Phenotype)) +
  geom_bar(stat = "identity", position = "dodge", color = "black") +
  theme_pubr() +
  theme(axis.text.x = element_text(angle = 270, vjust = 1, hjust = 1, size = 7),
        axis.text.y = element_text(size = 8)) +
  xlab("Genotypes by order of phenotype and freq") +
  scale_fill_manual(values = c("blue", "red", "pink", "white"))

关键修改说明 #

1. 替换频率统计方式：用count()替代table()+转置的操作，更符合tidyverse语法逻辑，避免数据结构转换带来的混乱。
2. 设置基因型的有序顺序：通过arrange()先按指定表型顺序排序，再在每个表型组内按频率降序排列，最后用fct_inorder()将Genotype转换为有序因子，确保x轴条形的排列顺序符合预期。
3. 保留表型的因子顺序：确保填充颜色与指定的表型顺序一一对应，避免颜色错位。

内容的提问来源于stack exchange，提问作者vanish007