❝本节来介绍一种新的方法来进行分面注释,下面通过1个案例来进行展示
加载R包
代码语言：javascript复制library(tidyverse)
library(ggh4x)
library(ggsignif)
library(ggsci)
library(grid)
定义函数
代码语言：javascript复制annotation_custom2 <- function (grob, xmin = -Inf, xmax = Inf, ymin = -Inf, ymax = Inf, data)
{
layer(data = data, stat = StatIdentity, position = PositionIdentity,
geom = ggplot2:::GeomCustomAnn,
inherit.aes = TRUE, params = list(grob = grob,
xmin = xmin, xmax = xmax,
ymin = ymin, ymax = ymax))
}
定义文本
代码语言：javascript复制grob <- grobTree(textGrob("Scatter plot", x=0.1, y=0.95, hjust=0,
gp=gpar(col="red", fontsize=13, fontface="italic")))
### 定义线段数据
data.segm<-data.frame(x=5,xend=6,y=3,yend=3,name="Sepal.Width")
plot <- ggplot() geom_segment(data=data.segm,color="red",
aes(x=x,y=y,yend=yend,xend=xend),inherit.aes=FALSE) theme_void()
单分面注释
代码语言：javascript复制mtcars%>% ggplot(aes(mpg,disp))
geom_point() facet_grid(vs~am)
annotation_custom2(grob,data = mtcars %>% filter(vs == 1,am==1))
❝看到此处，那么可以联想到使用「annotation_custom2」函数还可以跨分面添加注释信息，下面来通过一个例子来进行展示，当然细节很多，需要各位观众老爷细细体会 ❞
跨分面注释
代码语言：javascript复制p1 <- iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>%
ggplot(aes(Species,value,fill=Species))
stat_boxplot(geom="errorbar",
position=position_dodge(width=0.8),width=0.2)
geom_boxplot(position=position_dodge(width =0.8))
pal <- c("#E64B35FF","#4DBBD5FF","#00A087FF","#F39B7FFF","#3C5488FF","#91D1C2FF")
### 构建线段数据
data.segm<-data.frame(x=0.5,xend=2,y=1,yend=1)
### 绘制线段
plot2 <- ggplot() geom_segment(data=data.segm,color="black",
aes(x=x,y=y,yend=yend,xend=xend),inherit.aes=FALSE)
theme_void()
### 构建显著性数据
grob2 <- grobTree(textGrob("***", x=1.8, y=2.1,hjust=0,
gp=gpar(col="red", fontsize=18)))
数据可视化
代码语言：javascript复制 p1 facet_nested(.~type name,drop=T,scale="free",space="free",switch="y",
strip =strip_nested(background_x =elem_list_rect(fill =pal),by_layer_x = F))
theme(panel.spacing.x = unit(0, "cm"))
annotation_custom2(grob=ggplotGrob(plot2),data =iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>% filter(name=="Petal.Length"),
ymin =2, ymax=2, xmin=0.8, xmax=Inf)
annotation_custom2(grob=ggplotGrob(plot2),data =iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>% filter(name=="Petal.Width"),
ymin =2, ymax=2, xmin=0, xmax=1.1)
annotation_custom2(grob=grob2,data =iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>% filter(name=="Petal.Length"),
ymin =2.1, ymax=2.1, xmin=1.8, xmax=2)
annotation_custom2(grob=ggplotGrob(plot2),data =iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>% filter(name=="Petal.Length"),
ymin =8, ymax=8, xmin=Inf,xmax=Inf)
annotation_custom2(grob=ggplotGrob(plot2),data =iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>% filter(name=="Petal.Width"),
ymin =8, ymax=8, xmin=-0.8,xmax=Inf)
annotation_custom2(grob=ggplotGrob(plot2),data =iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>% filter(name=="Sepal.Length"),
ymin =8, ymax=8, xmin=-0.8,xmax=3.3)
annotation_custom2(grob=grob2,data =iris %>% pivot_longer(-Species) %>%
mutate(type=name) %>%
separate(type,into=c("type","type2",sep=".")) %>%
select(-.,-type2) %>% filter(name=="Petal.Width"),
ymin =8.05, ymax=8.05, xmin=1.8, xmax=2)
labs(x=NULL,y=NULL) scale_fill_brewer()
theme_minimal()
theme(axis.text.x = element_blank(),
panel.spacing.x = unit(0, "cm"),
legend.position = "npn")
❝使用「annotation_custom2」函数进行跨分面添加注释相对于之前的强行画图的确好很多，但是若需要注释的数据过多也比较繁琐，此外还有不少细节需要优化；那么本节介绍到此结束，「喜欢的观众老爷欢迎分享转发，每天都想着呈现一些实用内容给各位」，本文数据均使用R内置数据集,
参考链接如下https://rdrr.io/github/LoiseauN/dimensionality/src/R/annotation_custom2.R ❞
