如何在ggplot2柱状图中按孵化质量从小到大排列X轴

你的柱状图X轴（孵化前体重区间）排列混乱，核心原因是last_rank列大概率为字符类型，ggplot默认按字符字典序排序，而非数值逻辑顺序。以下是两种可行的修复方案：

方案1：提前将last_rank转为有序因子（推荐） #

在数据处理流程中，先提取每个区间的起始数值，按数值大小排序后将last_rank设为有序因子，确保绘图时X轴按正确逻辑排列：

H <- read.csv("H.csv")
library("ggplot2")
library(dplyr)
library(tidyr)

# 提取区间起始值，排序后生成因子水平列表
sorted_ranks <- H %>%
  drop_na(last_rank) %>%
  distinct(last_rank) %>%
  mutate(start_num = as.numeric(sub("-.*", "", last_rank))) %>%
  arrange(start_num) %>%
  pull(last_rank)

H %>%
  drop_na(last_rank) %>%
  filter(chick_outcome %in% c("Released", "Died")) %>%
  count(last_rank, chick_outcome) %>%
  group_by(last_rank) %>%
  mutate(n = n/sum(n) * 100) %>%
  # 将last_rank转为有序因子，绑定排序后的水平
  mutate(last_rank = factor(last_rank, levels = sorted_ranks, ordered = TRUE)) %>%
  ggplot() +
  aes(last_rank, n, fill = chick_outcome, label = paste0(round(n,2), "%")) +
  geom_col(position = 'stack') +  # 注：geom_col是geom_bar(stat='identity')的简写，二者保留一个即可
  theme(axis.text.x = element_text(angle = 45, hjust = 0.9)) +
  geom_text(position = position_stack(0.5)) +
  labs(title = "孵化前体重对雏鸟存活结果的影响",
       x = "孵化前体重区间(g)",
       y = "结果占比(%)") +
  scale_fill_manual(values = c("red", "green"),
                    name = "雏鸟结果") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5, size = 17, family = "Times new roman"))

方案2：在ggplot的aes中直接指定X轴顺序 #

如果不想修改数据类型，可在aes中用reorder()函数直接指定排序逻辑（按区间起始数值排序）：

H <- read.csv("H.csv")
library("ggplot2")
library(dplyr)
library(tidyr)

H %>%
  drop_na(last_rank) %>%
  filter(chick_outcome %in% c("Released", "Died")) %>%
  count(last_rank, chick_outcome) %>%
  group_by(last_rank) %>%
  mutate(n = n/sum(n) * 100) %>%
  ggplot() +
  # 用reorder指定last_rank按区间起始数值排序
  aes(x = reorder(last_rank, as.numeric(sub("-.*", "", last_rank))), 
      y = n, fill = chick_outcome, label = paste0(round(n,2), "%")) +
  geom_col(position = 'stack') +
  theme(axis.text.x = element_text(angle = 45, hjust = 0.9)) +
  geom_text(position = position_stack(0.5)) +
  labs(title = "孵化前体重对雏鸟存活结果的影响",
       x = "孵化前体重区间(g)",
       y = "结果占比(%)") +
  scale_fill_manual(values = c("red", "green"),
                    name = "雏鸟结果") +
  theme_classic() +
  theme(plot.title = element_text(hjust = 0.5, size = 17, family = "Times new roman"))

额外优化提示 #

原代码中geom_bar(position = 'stack', stat = 'identity')与geom_col()功能完全重复，保留其中一个即可。

内容的提问来源于stack exchange，提问作者Jana Rau