R语言Plotly子图共享X轴时，如何按首个子图排序X轴？

核心问题是开启shareX = TRUE后，Plotly会统一重置X轴的排序规则，覆盖单个子图的categoryorder设置。解决思路是手动指定所有子图的X轴类别顺序，或者在合并子图后统一设置全局X轴的排序规则。

方法一：提取首个子图的排序后，统一给所有子图设置 #

1. 先单独生成第一个子图，提取其按Y值降序后的X类别顺序
2. 给所有子图的X轴设置categoryarray为这个提取到的顺序，同时指定categoryorder = "array"

示例代码（R语言）：

library(plotly)

# 模拟测试数据
df <- data.frame(
  category = c("B", "A", "C", "D"),
  val1 = c(10, 20, 5, 15),
  val2 = c(8, 12, 6, 9),
  val3 = c(5, 18, 3, 10),
  val4 = c(12, 7, 9, 14)
)

# 生成第一个子图并提取排序后的X类别数组
p1_temp <- plot_ly(df, x = ~category, y = ~val1, type = "bar") %>%
  layout(xaxis = list(categoryorder = "total descending"))
sorted_cats <- p1_temp$x$layout$xaxis$categoryarray

# 给所有子图配置统一的X轴排序规则
p1 <- plot_ly(df, x = ~category, y = ~val1, type = "bar") %>%
  layout(xaxis = list(categoryarray = sorted_cats, categoryorder = "array"))
p2 <- plot_ly(df, x = ~category, y = ~val2, type = "bar") %>%
  layout(xaxis = list(categoryarray = sorted_cats, categoryorder = "array"))
p3 <- plot_ly(df, x = ~category, y = ~val3, type = "bar") %>%
  layout(xaxis = list(categoryarray = sorted_cats, categoryorder = "array"))
p4 <- plot_ly(df, x = ~category, y = ~val4, type = "bar") %>%
  layout(xaxis = list(categoryarray = sorted_cats, categoryorder = "array"))

# 合并子图并开启X轴共享
subplot(p1, p2, p3, p4, shareX = TRUE)

方法二：合并子图后直接修改全局X轴布局 #

不需要逐个修改子图，合并后直接在全局布局里指定X轴的排序规则：

# 先生成所有基础子图
p1 <- plot_ly(df, x = ~category, y = ~val1, type = "bar")
p2 <- plot_ly(df, x = ~category, y = ~val2, type = "bar")
p3 <- plot_ly(df, x = ~category, y = ~val3, type = "bar")
p4 <- plot_ly(df, x = ~category, y = ~val4, type = "bar")

# 直接根据首个子图的Y值生成降序的X类别顺序
sorted_cats <- df$category[order(-df$val1)]

# 合并子图并设置全局X轴排序
subplot(p1, p2, p3, p4, shareX = TRUE) %>%
  layout(xaxis = list(
    categoryarray = sorted_cats,
    categoryorder = "array"
  ))

两种方法本质都是用固定的类别数组强制锁定X轴顺序，避免共享X轴时Plotly自动重置排序规则。

内容的提问来源于stack exchange，提问作者Pawels