R语言ggplot2条形图如何自定义设置不同类别条形的填充颜色

原代码直接将fill、color美学映射全局绑定到model字段，ggplot会为每个全局唯一的模型值分配固定颜色，导致无法实现「非B模型跨分类保持同色、仅B模型在不同分类下使用差异化配色」的规则，且提前定义的色板未正确接入比例尺。

核心调整逻辑：

• 预处理阶段为每一行数据预计算对应的主题色：非B模型按模型名绑定固定色，B模型按所属分类分配差异化颜色
• 正负条形的边框色统一绑定预计算的主题色，负值条形固定填充白色
• 使用identity类比例尺直接读取预定义的颜色值，避免ggplot自动分配颜色

library(tidyverse)
library(tidytext)

df <- data.frame(model  = c("A","B","C","D","B","C"),
                 category = c("origin", "origin","origin","abroad","abroad","abroad"),
                 pos = c(40,50,45,100,105,80),
                 neg = c(-10,-5,-4,-16,-7,-2))

# 长宽转换+预计算每个条形对应的主题色
df <- df %>% 
  pivot_longer(cols=c('pos','neg'), names_to = 'sign') %>% 
  mutate(
    sign = factor(sign, levels = c("neg", "pos")),
    # 按规则分配颜色
    bar_color = case_when(
      model == "A" ~ "chocolate",
      model == "C" ~ "yellow",
      model == "D" ~ "steelblue3",
      model == "B" & category == "origin" ~ "deeppink4",
      model == "B" & category == "abroad" ~ "steelblue3"
    )
  )

ggplot() +
  # 绘制正值条形：填充、边框均用预定义色
  geom_col(aes(value, tidytext::reorder_within(model, value, category),
               fill = bar_color, color = bar_color), 
           data = df[df$sign == "pos", ],
           position = "stack") +
  # 绘制负值条形：填充固定为白色，边框和同组正值条形同色
  geom_col(aes(value, tidytext::reorder_within(model, value, category),
               color = bar_color), 
           data = df[df$sign == "neg", ], 
           fill = "white",
           position = "stack") +
  # 直接使用预定义颜色值，不自动分配配色
  scale_fill_identity() +
  scale_color_identity() +
  tidytext::scale_y_reordered() +
  facet_grid(category ~ ., switch = "y",scales = "free_y") +
  theme(axis.text.x = element_text(angle = 90),
        strip.background = element_rect(fill = "white"),
        strip.placement = "outside",
        strip.text.y.left = element_text(angle = 0),
        panel.spacing = unit(0, "lines")) + 
  theme(legend.position="none") + 
  labs( title = "BarPlot",
        subtitle = "changes",
        y = " ")

• A、C、D模型跨分类保持统一配色：A为巧克力色、C为黄色、D为钢蓝色
• B模型按分类差异化配色：origin类下为深粉色、abroad类下为钢蓝色
• 所有负值条形填充为白色，边框颜色与对应组正值条形完全一致

内容的提问来源于stack exchange，提问作者user18937485