ggplot2中如何按TYPE统一匹配图例、颜色、线型与形状？

以下是修改后的代码，完全满足你的四个需求：

data %>%
  ggplot(aes(x = IND_TEST_SCORE, y = SCORE,
             color = TYPE, linetype = TYPE, shape = TYPE)) +
  geom_point(alpha = 0.1) +  # 移除硬编码的color参数，让颜色遵循TYPE映射
  stat_smooth(formula = y ~ x, method = lm, se = TRUE) +
  # 统一三个scale的name，让图例合并
  scale_shape_manual(name = "Type:", values = c(16, 17)) +
  scale_linetype_manual(name = "Type:", values = c("solid", "dashed")) +
  scale_color_manual(name = "Type:",
                     values = c(A = "yellow", B = "cadetblue2"),
                     labels = c(A = "TYPE A", B = "TYPE B")) +
  labs(x = "IND_TEST_SCORE",
       y = "SCORE",
       title = "Effect of IND_TEST_SCORE and TYPE on SCORE") +
  theme_bw()

关键修改说明 #

1. 匹配scale_color_manual与TYPE变量：移除geom_point()中的color = "red"硬编码，点的颜色会按照aes(color = TYPE)的设置，自动匹配scale_color_manual中定义的A（黄色）、B（cadetblue2）颜色。
2. 图例匹配线型定义：scale_linetype_manual明确指定TYPE对应的线型（A为实线，B为虚线），结合统一的name设置，线型会整合到合并后的图例中。
3. 统一形状、线型与颜色图例：给scale_shape_manual、scale_linetype_manual、scale_color_manual设置相同的name = "Type:"，ggplot2会自动将三个美学映射的图例合并为一个，每个类别同时显示对应的形状、线型和颜色。
4. 移除重复图例：通过统一三个scale的name，避免生成多个独立图例，最终只保留一个整合后的图例。

额外建议：先过滤掉NA/NaN值，让回归拟合更准确：

data %>%
  filter(!is.na(IND_TEST_SCORE) & !is.na(SCORE) & !is.nan(SCORE)) %>%
  ggplot(aes(x = IND_TEST_SCORE, y = SCORE,
             color = TYPE, linetype = TYPE, shape = TYPE)) +
  geom_point(alpha = 0.1) +
  stat_smooth(formula = y ~ x, method = lm, se = TRUE) +
  scale_shape_manual(name = "Type:", values = c(16, 17)) +
  scale_linetype_manual(name = "Type:", values = c("solid", "dashed")) +
  scale_color_manual(name = "Type:",
                     values = c(A = "yellow", B = "cadetblue2"),
                     labels = c(A = "TYPE A", B = "TYPE B")) +
  labs(x = "IND_TEST_SCORE",
       y = "SCORE",
       title = "Effect of IND_TEST_SCORE and TYPE on SCORE") +
  theme_bw()

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