ggplot2图表异常：图例显示但绘图无颜色/有颜色但无图例

我正在开发带双Y轴和图例的多变量ggplot2图表，参考相关解决方案后，目前得到两个版本的图表：

• 一个版本图例正常显示且颜色正确，但绘图元素没有颜色
• 另一个版本绘图元素有颜色，但图例不显示

我知道这和geom_point()中colour参数的位置有关，参考之前的回答：

colour= XYZ 应该放在 aes() 内部，而不是外部
geom_point(aes(data, colour=XYZ)) #------>显示图例
geom_point(aes(data),colour=XYZ) #------>不显示图例

数据集片段如下：

h1enraw <-structure(list(run = c(1738, 1739, 1740, 1741, 1742, 1743),
                         temp = c(19, 19, 19, 19, 21, 22),
                         avgbase = c(1386, 1386, 1389, 1389, 1352, 1336),
                         no2c = c(6.98, 6.96, 6.94, 6.99, 7.01, 7.01),
                         no3c = c(18.52, 17.6, 18.77, 19.81, 18.22, 18.60)), 
                    row.names = c(NA, 6L), class = "data.frame")

我理解为什么一个版本有图例另一个没有，但不清楚这个设置为何会影响绘图元素的颜色显示，恳请指点方向。

图例显示但绘图无颜色的代码 #

# 选择第二Y轴的缩放系数
scaleRight <- 40
ymax <-43

h1no2<-ggplot(h1enraw, aes(x=run)) +
  geom_path(aes(y=temp, group=1,colour="cornflowerblue"), size=0.9) +
  geom_point(aes(y=no2c, group=1,colour="red")) +
  geom_point(aes(y=no3c, group=1,colour="darkgreen")) +
  geom_line(aes(y=avgbase/scaleRight, group=1,colour="chocolate1"), size=0.9) +
  scale_y_continuous(breaks=seq(0,40,2), expand = expansion(mult = c(0,.05)),
                     sec.axis = sec_axis(~.*scaleRight, name = "Average Baseline (Transmitance Units)")) +
  coord_cartesian(ylim = c(0, ymax)) +
  theme_classic() +
  labs( y="Temperature (°C) / Concentration (ppm)", x="Run", title = "H1 - High Temperature Cycle") +
  theme(text = element_text(size=9),
        axis.text.x = element_text(angle=90, vjust=0.6),
        axis.text=element_text(size=12),
        axis.title=element_text(size=14,face="bold"),
        panel.border = element_rect(colour = "chocolate1", fill=NA, size=0.5),
        legend.position = "bottom", legend.title=element_text(size=10), legend.text = element_text(size=8),
        axis.title.y = element_text(size=10),
        plot.title = element_text(size=14, face="bold")) +
  theme(legend.title=element_blank()) + 
  scale_fill_manual(breaks=c("temp","no2c","no3c","avgbase"),
                    labels=c("Temperature", "NO2", "NO3", "Avg Baseline"),
                    values = c("temp"="cornflowerblue", "no2c"="red",
                               "no3c"="darkgreen", "avgbase"="chocolate1")) + 
  scale_color_manual(labels = c("Temperature", expression(NO[2]), expression(NO[3]), "Avg Baseline"),
                     values = c("temp"="cornflowerblue", "no2c"="red",
                                "no3c"="darkgreen", "avgbase"="chocolate1"),
                     breaks=c("temp","no2c","no3c","avgbase"))

（对应图表：图例显示但绘图元素无颜色）

绘图有颜色但无图例的代码 #

# 选择第二Y轴的缩放系数
scaleRight <- 40
ymax <-43

h2no2<-ggplot(h2enraw, aes(x=run)) +
  geom_path(aes(y=temp, group=1), colour="cornflowerblue", size=0.9) +
  geom_point(aes(y=no2c, group=1 ),colour="red") +
  geom_point(aes(y=no3c, group=1 ),colour="darkgreen") +
  geom_line(aes(y=avgbase/scaleRight, group=1 ),colour="chocolate1", size=0.9) +
  scale_y_continuous(breaks=seq(0,40,2), expand = expansion(mult = c(0,.05)),
                     sec.axis = sec_axis(~.*scaleRight, name = "Average Baseline (Transmitance Units)")) +
  coord_cartesian(ylim = c(0, ymax)) +
  theme_classic() +
  labs( y="Temperature (°C) / Concentration (ppm)", x="Run", title = "H2 - High Temperature Cycle") +
  theme(text = element_text(size=9),
        axis.text.x = element_text(angle=90, vjust=0.6),
        axis.text=element_text(size=12),
        axis.title=element_text(size=14,face="bold"),
        panel.border = element_rect(colour = "chocolate1", fill=NA, size=0.5),
        legend.position = "bottom", legend.title=element_text(size=10), legend.text = element_text(size=8),
        axis.title.y = element_text(size=10),
        plot.title = element_text(size=14, face="bold")) +
  scale_fill_manual(name="Legend",
                    breaks=c("temp","no2c","no3c","avgbase"),
                    labels=c("Temperature", "NO2", "NO3", "Avg Baseline"))

（对应图表：绘图元素有颜色但无图例）

为什么第一个版本绘图无颜色但图例正常？ #

当把colour="cornflowerblue"这类颜色值放在aes()内部时，ggplot2会将这些字符串视为分类变量的水平值，而非直接作为颜色代码。此时绘图元素的颜色由后续scale_color_manual决定，但scale_color_manual的breaks是"temp"、"no2c"这类变量名，和aes(colour=...)中传入的颜色字符串完全不匹配，导致ggplot找不到对应颜色映射，只能用默认灰色显示绘图元素。

图例能正常显示，是因为scale_color_manual定义了完整的标签和颜色映射，只是该映射与aes()内的颜色字符串无关联。

为什么第二个版本绘图有颜色但无图例？ #

当把colour="cornflowerblue"放在aes()外部时，这是直接给绘图元素设置固定颜色，不属于ggplot的“美学映射”范畴。ggplot仅当美学属性（如colour、fill）在aes()内部时，才会生成对应图例——外部设置的固定样式不会触发图例生成，因此看不到图例。

正确的解决方案：规范美学映射 #

要同时实现绘图元素有颜色+图例正常显示，需将colour映射到变量标识（而非直接写颜色值），再通过scale_color_manual绑定颜色：

方案1：整理为长数据（推荐，符合ggplot规范）

library(tidyr)
scaleRight <- 40
ymax <-43

# 转换为长格式数据
h1enraw_long <- h1enraw %>%
  mutate(avgbase_scaled = avgbase / scaleRight) %>%
  pivot_longer(cols = c(temp, no2c, no3c, avgbase_scaled),
               names_to = "variable",
               values_to = "value")

# 绘图
ggplot(h1enraw_long, aes(x=run, y=value, colour=variable)) +
  # 根据变量类型选择几何对象
  geom_line(data = subset(h1enraw_long, variable %in% c("temp", "avgbase_scaled")), size=0.9) +
  geom_point(data = subset(h1enraw_long, variable %in% c("no2c", "no3c"))) +
  scale_y_continuous(breaks=seq(0,40,2), expand = expansion(mult = c(0,.05)),
                     sec.axis = sec_axis(~.*scaleRight, name = "Average Baseline (Transmitance Units)")) +
  coord_cartesian(ylim = c(0, ymax)) +
  theme_classic() +
  labs(y="Temperature (°C) / Concentration (ppm)", x="Run", title = "H1 - High Temperature Cycle", colour="") +
  theme(text = element_text(size=9),
        axis.text.x = element_text(angle=90, vjust=0.6),
        axis.text=element_text(size=12),
        axis.title=element_text(size=14,face="bold"),
        panel.border = element_rect(colour = "chocolate1", fill=NA, size=0.5),
        legend.position = "bottom", legend.text = element_text(size=8),
        axis.title.y = element_text(size=10),
        plot.title = element_text(size=14, face="bold")) +
  scale_color_manual(labels = c("Temperature", expression(NO[2]), expression(NO[3]), "Avg Baseline"),
                     values = c("temp"="cornflowerblue", "no2c"="red",
                                "no3c"="darkgreen", "avgbase_scaled"="chocolate1"))

方案2：不整理数据，直接映射变量名

scaleRight <- 40
ymax <-43

ggplot(h1enraw, aes(x=run)) +
  geom_path(aes(y=temp, colour="temp"), size=0.9) +
  geom_point(aes(y=no2c, colour="no2c")) +
  geom_point(aes(y=no3c, colour="no3c")) +
  geom_line(aes(y=avgbase/scaleRight, colour="avgbase"), size=0.9) +
  scale_color_manual(labels = c("Temperature", expression(NO[2]), expression(NO[3]), "Avg Baseline"),
                     values = c("temp"="cornflowerblue", "no2c"="red",
                                "no3c"="darkgreen", "avgbase"="chocolate1"),
                     breaks=c("temp","no2c","no3c","avgbase")) +
  scale_y_continuous(breaks=seq(0,40,2), expand = expansion(mult = c(0,.05)),
                     sec.axis = sec_axis(~.*scaleRight, name = "Average Baseline (Transmitance Units)")) +
  coord_cartesian(ylim = c(0, ymax)) +
  theme_classic() +
  labs( y="Temperature (°C) / Concentration (ppm)", x="Run", title = "H1 - High Temperature Cycle", colour="") +
  theme(text = element_text(size=9),
        axis.text.x = element_text(angle=90, vjust=0.6),
        axis.text=element_text(size=12),
        axis.title=element_text(size=14,face="bold"),
        panel.border = element_rect(colour = "chocolate1", fill=NA, size=0.5),
        legend.position = "bottom", legend.text = element_text(size=8),
        axis.title.y = element_text(size=10),
        plot.title = element_text(size=14, face="bold"))

修改后，aes(colour="temp")中的"temp"作为分类键，与scale_color_manual的breaks对应，ggplot就能正确将颜色映射到绘图元素，同时生成对应图例。

内容的提问来源于stack exchange，提问作者MGB