我们可以添加legend.box = "horizontal并为每个ggplot上的图例设置正确的顺序。

然后，在width上添加align = "v"和ggarrange

p1 <- ggplot(coex1,aes(x=X,y=Y,shape=letters,col=value,col=value)) +
  geom_jitter(width=0.05) +
  scale_color_gradient(low="blue",high="red") +
  theme(legend.box = "horizontal")+
  guides(color = guide_legend(order=1),size = guide_legend(order=2),shape = guide_legend(order=3))


p2 <- ggplot(coex2,size=value)) +
  geom_jitter(width=0.05) +
  scale_color_gradient(low="orange",high="yellow") +
  theme(legend.box = "horizontal")+
  guides(color = guide_legend(order=1),shape = guide_legend(order=3))



p3 <- ggplot(coex3,size=value)) +
  geom_jitter(width=0.05) +
  scale_color_gradient(low="green",high="cyan") +
  theme(legend.box = "horizontal")+
  guides(color = guide_legend(order=1),shape = guide_legend(order=3))


ggarrange(p1,p2,p3 + rremove("x.text"),labels = c("a","b","c"),ncol = 1,nrow = 3,legend = "right",widths = c(2,2,3),align = "v"
          )

使用这种方法会使您的生活变得非常困难。使用ggplot出色的分组功能，然后就不需要绘图组合包了。

您可以按美学和/或构面分组。见下文。关键是使数据变长，另请参见下文。了解它还可以如何大大减少您的代码。

我还介绍了如何将颜色和尺寸组合成一个图例。 see this thread

library(tidyverse)

mydat <- data.frame(Samples,X,Y,AA,BB,CC) %>% 
  pivot_longer(names_to = "letters",values_to = "value",cols = AA:CC) %>%
  group_by(letters) %>%
  mutate(scaled_val = scale(value)) %>%
  ungroup()

ggplot(mydat,aes(X,col = scaled_val,size = scaled_val)) +
  geom_jitter(width = 0.05) +
  scale_color_gradient(low = "blue",high = "red") +
  facet_grid(~letters) +
  guides(color=guide_legend(),size = guide_legend())
#> Warning: Removed 15 rows containing missing values (geom_point).

^{由reprex package（v0.3.0）于2020-03-30创建}

数据

Samples <- c(1,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17)
X <- c(1.16,1.16,0.96,0.67,0.78,0.55,0.3,0.26,0.26)
Y <- c(75.45,75.45,86.66,103.36,NA,107.53,128.49,128.49)
AA <- c(0.003437318,0.005842468,0.005573348,0.006074338,0.002537367,0.006583666,0.006015314,0.010983784,0.009116288,0.010872489,0.010924257,0.009359167,0.009068434,0.00601658,0.017616501,0.014813675,0.018048576)
BB <- c(0.007614672,0.007632451,0.007066506,0.007524053,0.008337992,0.012520277,0.012249,0.011351902,0.01263021,0.009969673,0.008850031,0.007290232,0.00724349,0.007161781,0.004299581,0.004896156,0.005970637)
CC <- c(0.002133046,0.00168291,0.001580502,0.001491037,0.001295399,0.001644785,0.001738881,0.001496376,0.00140218,0.001247361,0.001364975,0.001209774,0.000933038,0.002034014,0.000665552,0.000855588,0.000878233)