如何基于Specie列重新排序ggplot分组柱状图的显示顺序

我想用ggplot绘制分组柱状图，要求按Specie列排序：先展示sp1的基因型对应的柱状图，然后是sp2，最后是sp3；同时每个物种内的柱状图要按var值降序排列。目前已经画出图表，但顺序不符合要求，以下是我用的代码和数据集：

ggplot(data=data, aes(x=reorder(Genotype, -var) , y=var, fill=Treatment)) +
  geom_bar(stat="identity", position=position_dodge(width=0.85)) +
  scale_fill_manual(values = c('#4850CE','#D37A3E'),
                    labels = c("Control Temperature (CT)","Heat tress (HS)")) +
  geom_errorbar(aes(ymin = var, ymax = var+e,colour=Treatment),
                width=0.75, position = position_dodge(0.85))+
  scale_colour_manual(values=c('#4850CE','#D37A3E'))+
  theme_bw() + labs(x=NULL, y=expression("a. weight"^~g))+
  scale_y_continuous(breaks = seq(0, max(data$var+data$e), length.out = 6),
                     labels = number_format(accuracy = 0.01))+
  geom_text(aes(label = Genotype, y = 0),angle=90, hjust=0, size=3)+
  theme(panel.grid.minor = element_blank(),
        axis.text.x = element_blank(),
        axis.ticks.x = element_blank(),
        axis.title.y = element_text(colour = "black", size=8, face="bold"),
        axis.ticks.y = element_line(colour= "black"),
        axis.text.y =element_text(colour="black", size = 8),
        legend.position = c(0.8, 0.92), legend.background = element_blank(),
        legend.key.size = unit(4, 'mm'),
        legend.title = element_blank(),
        legend.text = element_text(size = 8,face="bold"))+
  guides(colour = FALSE)

数据集 #

要实现先按Specie排序（sp1→sp2→sp3），再在每个物种内按var降序排列Genotype，需要提前自定义Genotype的因子顺序，而不是仅用reorder函数。

步骤1：预处理数据，确定排序规则 #

先基于每个基因型的平均var值（覆盖CT和HS两个处理），结合物种顺序确定最终排序：

library(dplyr)

# 计算每个基因型的平均var值，按指定物种顺序+var降序排列
sorted_genotypes <- data %>%
  group_by(Genotype, Specie) %>%
  summarise(mean_var = mean(var), .groups = "drop") %>%
  arrange(factor(Specie, levels = c("sp1", "sp2", "sp3")), desc(mean_var)) %>%
  pull(Genotype)

# 将Genotype转换为因子，指定自定义顺序
data$Genotype <- factor(data$Genotype, levels = sorted_genotypes)

步骤2：修改ggplot代码 #

把原代码中x=reorder(Genotype, -var)替换为x=Genotype，因为已经提前设置好因子顺序：

ggplot(data=data, aes(x=Genotype, y=var, fill=Treatment)) +
  geom_bar(stat="identity", position=position_dodge(width=0.85)) +
  scale_fill_manual(values = c('#4850CE','#D37A3E'),
                    labels = c("Control Temperature (CT)","Heat stress (HS)")) + # 修正拼写错误tress→stress
  geom_errorbar(aes(ymin = var, ymax = var+e,colour=Treatment),
                width=0.75, position = position_dodge(0.85))+
  scale_colour_manual(values=c('#4850CE','#D37A3E'))+
  theme_bw() + labs(x=NULL, y=expression("a. weight"^~g))+
  scale_y_continuous(breaks = seq(0, max(data$var+data$e), length.out = 6),
                     labels = scales::number_format(accuracy = 0.01))+ # 加scales::前缀避免函数找不到
  geom_text(aes(label = Genotype, y = 0),angle=90, hjust=0, size=3)+
  theme(panel.grid.minor = element_blank(),
        axis.text.x = element_blank(),
        axis.ticks.x = element_blank(),
        axis.title.y = element_text(colour = "black", size=8, face="bold"),
        axis.ticks.y = element_line(colour= "black"),
        axis.text.y =element_text(colour="black", size = 8),
        legend.position = c(0.8, 0.92), legend.background = element_blank(),
        legend.key.size = unit(4, 'mm'),
        legend.title = element_blank(),
        legend.text = element_text(size = 8,face="bold"))+
  guides(colour = FALSE)

说明 #

• 用平均var值排序，确保每个基因型的整体水平作为排序依据，而非单个处理的数值
• 通过factor(Specie, levels = c("sp1", "sp2", "sp3"))强制指定物种的展示顺序
• 因子顺序设置完成后，ggplot会自动按照该顺序绘制柱状图

内容的提问来源于stack exchange，提问作者Ikram Bashir