如何调整ggplot水平图变量轴顺序，将回归参考类别置于顶部？

要让“Standard Job (ref)”显示在图的顶部、“Urban”显示在底部，只需要反转变量因子的水平顺序即可。因为coord_flip()会将x轴转为y轴，而ggplot默认会按照因子水平的顺序从下到上排列y轴标签，把定义的correct_order反转后赋值给因子水平，就能实现需求。

修改后的完整代码如下：

# 创建比值比数据框
odds_ratio_data <- data.frame(
  Variables = c("Standard Job (ref)", "Precarious job", "Out of labor force", "Unemployed", 
                "Age", "Female", "Non Black and Non Hispanic (ref)", "Black", "Hispanic", 
                "Never Married (ref)", "Married/Cohabiting", "Divorced/Widowed", "Number of children", 
                "Less than High school (ref)", "High school/GED", "College or above", 
                "Both bio parent (ref)", "One bio parent", "Bio and step parent", 
                "Northeast", "North central", "South", "West", "Urban"),
  OddsRatio = c(1, 1.377127764, 1.150273799, 1.616074402, 0.93239382, 0.923116346, 1, 
                1.161834243, 1.733253018, 1, 0.121237966, 0.60653066, 0.826959134, 1, 
                0.93239382, 0.57694981, 1, 0.66365025, 0.486752256, 1, 0.711770323, 0.818730753, 
                0.835270211, 0.711770323),
  LowerCI_OR = c(1, 1.173510871, 1.040810774, 1.404947591, 0.913931185, 0.860707976, 1, 
                 1.051271096, 1.419067549, 1, 0.095369162, 0.516851334, 0.778800783, 1, 
                 0.810584246, 0.491644197, 1, 0.548811636, 0.386741023, 1, 0.637628152, 
                 0.733446956, 0.748263568, 0.637628152),
  UpperCI_OR = c(1, 1.616074402, 1.27124915, 1.858928042, 0.951229425, 0.990049834, 1, 
                 1.284025417, 2.117000017, 1, 0.154123662, 0.711770323, 0.878095431, 1, 
                 1.072508181, 0.677056874, 1, 0.802518798, 0.612626394, 1, 0.786627861, 
                 0.913931185, 0.93239382, 0.786627861)
)

# 定义变量顺序，反转后让"Standard Job (ref)"在顶部，"Urban"在底部
correct_order <- c("Standard Job (ref)", "Precarious job", "Out of labor force", "Unemployed", 
                "Age", "Female", "Non Black and Non Hispanic (ref)", "Black", "Hispanic", 
                "Never Married (ref)", "Married/Cohabiting", "Divorced/Widowed", "Number of children", 
                "Less than High school (ref)", "High school/GED", "College or above", 
                "Both bio parent (ref)", "One bio parent", "Bio and step parent", 
                "Northeast", "North central", "South", "West", "Urban")

# 反转因子水平顺序
odds_ratio_data$Variables <- factor(odds_ratio_data$Variables, levels = rev(correct_order))

# 绘制水平比值比图
ggplot(odds_ratio_data, aes(x = Variables, y = OddsRatio)) +
  geom_point(color = "blue", size = 3) +  # 绘制比值比点
  geom_errorbar(aes(ymin = LowerCI_OR, ymax = UpperCI_OR), width = 0.2) +  # 置信区间误差棒
  coord_flip() +  # 翻转坐标轴为水平图
  theme_minimal() +
  labs(title = "预测NLSY07（2005-2017）年轻成年人同住情况的逻辑回归模型比值比",
       x = "变量",
       y = "比值比") +
  theme(plot.title = element_text(hjust = 0.5),  # 标题居中
        axis.text.y = element_text(size = 10))  # 调整y轴标签大小

关键修改点 #

• 在将Variables转为因子时，使用rev(correct_order)替代原来的correct_order，反转因子水平顺序，配合coord_flip()实现标签从上到下的需求。
• 同步将图表的标题和轴标签翻译成中文，适配中文使用场景。

内容的提问来源于stack exchange，提问作者Quan Mai