为ggplot密度图指定自定义线条颜色并调整图例样式

1. 模拟可复现数据 #

先构造包含9个模型的示例数据集，方便后续测试：

library(ggplot2)
set.seed(123) # 确保结果可复现

# 生成9个模型的模拟数据
model_data <- data.frame(
  value = c(rnorm(500, mean = 0, sd = 1),
            rnorm(500, mean = 0.5, sd = 1.2),
            rnorm(500, mean = -0.3, sd = 0.8),
            rnorm(500, mean = 1, sd = 0.9),
            rnorm(500, mean = -1, sd = 1.1),
            rnorm(500, mean = 0.2, sd = 0.7),
            rnorm(500, mean = -0.7, sd = 1.3),
            rnorm(500, mean = 0.8, sd = 0.6),
            rnorm(500, mean = -0.5, sd = 0.9)),
  model = rep(paste0("Model_", 1:9), each = 500)
)

2. 定义自定义颜色向量 #

指定9种区分度较高的对应颜色，你可以根据需求自行替换：

line_col <- c("#E69F00", "#56B4E9", "#009E73", "#F0E442", "#0072B2", 
              "#D55E00", "#CC79A7", "#999999", "#000000")

3. 绘制仅带线条的密度图并调整图例 #

使用geom_density并设置fill = NA只保留线条，同时通过guides()修改图例符号为线条：

ggplot(model_data, aes(x = value, color = model)) +
  geom_density(fill = NA, size = 1) + # 仅显示密度线条，去掉填充
  scale_color_manual(values = line_col) + # 应用自定义颜色向量
  guides(color = guide_legend(override.aes = list(linetype = 1, shape = NA))) + # 图例符号改为线条
  theme_minimal() +
  labs(x = "Value", y = "Density", title = "Density Plot of 9 Models (Lines Only)")

关键说明 #

• geom_density(fill = NA)：移除直方图填充色，仅保留密度线条轮廓
• scale_color_manual(values = line_col)：将自定义颜色映射到每个模型
• guides(color = guide_legend(override.aes = list(linetype = 1, shape = NA)))：强制图例显示线条符号，替换默认的填充方块；shape = NA隐藏默认点符号，linetype = 1设置为实线

如果你的原始数据是宽格式（每个模型为单独列），可先通过tidyr::pivot_longer转换为长格式，再套用上述代码绘制。

内容的提问来源于stack exchange，提问作者climsaver