R基础绘图篇 | 旭日图与冰柱图的绘制

写在前面

旭日图（sunbrust diagram），通常也被称为多层饼图（multi-level pie chart）或径向树图，通常会用来展示层级占比关系，通过一系列的圆环展示层次结构。冰柱图（icicle diagram）也叫分区层图（partition layer chart），也就是直角坐标系下的旭日图，他们都是展示层级占比关系的王者。

开始绘图

需要调用的R包有以下4个

代码语言：javascript复制library(ggraph)

library(igraph)

library(RColorBrewer)

library(dplyr)读取数据

代码语言：javascript复制#df<-read.csv('旭日图.csv',header=TRUE,stringsAsFactors=FALSE)

df<-read.csv(file.choose( ),header=TRUE,stringsAsFactors=FALSE)旭日图

分割角度均等平分

代码语言：javascript复制edges<- data.frame(rbind(

cbind(rep('origin',4),unique(as.character(df$Season))),

as.matrix(df[!duplicated(df[c('Season','Month')]),1:2]),

as.matrix(df[!duplicated(df[c('Month','Week')]),2:3])))

colnames(edges)<-c('from','to')

vertices0<-data.frame(name=unique(c(as.character(edges$from), as.character(edges$to))))

df_leaf<-df[,c('Week','Value','label')]

df_leaf$angle<-90-(1:nrow(df_leaf))/nrow(df_leaf)*360

df_leaf$angle<-ifelse(df_leaf$angle< -90, df_leaf$angle+180, df_leaf$angle)

vertices<-left_join(vertices0,df_leaf,by=c('name'='Week'))

graph <- graph_from_data_frame(edges, vertices = vertices)

# 分割角度均等平分

ggraph(graph, layout ='partition', circular = TRUE) +

geom_node_arc_bar(aes(filter =(depth<=2 & depth>0 )),fill='#FEE5D9',color='black',size=0.1)+#,size=size

geom_node_arc_bar(aes(filter =(depth==3 ),fill = Value),size=0.1)+

geom_node_text( aes(filter =(depth<=2 & depth>0 ),label=name,size=-depth),angle=0,colour="black") +

geom_node_text(aes(label=label,angle=angle),size=2,colour="black") +

scale_size(range=c(3,5))+

coord_fixed()+

scale_fill_distiller(palette='Reds')+

guides(size="none")+

theme_void()冰柱图

分割角度均等平分

代码语言：javascript复制ggraph(graph, layout ='partition') +

geom_node_tile(aes(filter =(depth<=2)),fill='#FEE5D9',color='black',size=0.1)+#,size=size

geom_node_tile(aes(filter =(depth==3 ),fill = Value),size=0.1)+

geom_node_text(aes(filter =(depth<=2 & depth>0),label=name,size=-depth),angle=90,colour="black") +

geom_node_text(aes(label=label),size=2,angle=90,colour="black")+

scale_size(range=c(3,4))+

scale_y_reverse()+

scale_fill_distiller(palette="Reds",na.value = "#FFF2EC")+

guides(size="none")+

theme_void()旭日图

分割角度与某个数值成比例

代码语言：javascript复制fake_circle<-c()

for (i in 1:nrow(df)){

fake_circle<-append(fake_circle,rep(df$Week[i],round(10*df$Value[i])))

}

edges<- data.frame(rbind(

cbind(rep('origin',4),unique(as.character(df$Season))),

as.matrix(df[!duplicated(df[c('Season','Month')]),1:2]),

as.matrix(df[!duplicated(df[c('Month','Week')]),2:3]),

cbind(fake_circle,as.character(1:length(fake_circle)))

))

colnames(edges)<-c('from','to')

vertices0<-data.frame(name=unique(c(as.character(edges$from), as.character(edges$to))))

df_leaf<-df[,c('Week','Value','label')]

df_leaf$angle<-90-(cumsum(df_leaf$Value)-df_leaf$Value/2)/sum(df_leaf$Value)*360

df_leaf$angle<-ifelse(df_leaf$angle< -90, df_leaf$angle+180, df_leaf$angle)

vertices<-left_join(vertices0,df_leaf,by=c('name'='Week'))

df_color<- data.frame(rbind(

as.matrix(df[!duplicated(df[c('Season','Season')]),c(1,1)]),

as.matrix(df[!duplicated(df[c('Season','Month')]),c(1,2)]),

as.matrix(df[!duplicated(df[c('Season','Week')]),c(1,3)])

))

colnames(df_color)<-c('Season','name')

vertices<-left_join(vertices,df_color,by='name')

graph <- graph_from_data_frame(edges, vertices = vertices)

#旭日图 分割角度与某个数值成比例

ggraph(graph, layout ='partition', circular = TRUE) +

geom_node_arc_bar(aes(filter =(depth<=3 & depth>0 ),fill = Season),size=0.1)+

geom_node_text(aes(filter =(depth<=2 & depth>0),label=name,size=-depth),angle=0,colour="black") +

geom_node_text(aes(label=label,angle=angle),size=2,colour="black")+

scale_size(range=c(3,4))+

coord_fixed()+

scale_fill_brewer(palette="Reds",direction=-1)+

guides(size="none",fill="none")+

theme_void()冰柱图

分割角度与某个数值成比例

代码语言：javascript复制#冰柱图 分割角度与某个数值成比例

ggraph(graph, layout ='partition')+

geom_node_tile(aes(filter =(depth<=3 ),fill = Season),size = 0.1)+

geom_node_text(aes(filter =(depth<=2 & depth>0),label=name,size=-depth),angle=90,colour="black") +

geom_node_text(aes(label=label),size=2,angle=90,colour="black")+

scale_size(range=c(3,4))+

scale_y_reverse()+

scale_fill_brewer(palette="Reds",na.value = "#FFF2EC")+

guides(size="none",fill="none")+

theme_void()