将Ann项目添加到ggplot2图例

Question

我想要实现的是，图例的标题为“年”，蓝色条显示为“ 2016”的小矩形，圆圈和三角形显示“ 2012”和“ 2008”的年。有什么想法可以实现吗？因为现在我只获得图例数据的图例。

可复制的示例：

# packages needed

library(tidyverse)
library(eurostat)
library(lubridate)

# downlaoding the dataset
ilc_di11 <- get_eurostat("ilc_di11", stringsAsFactors=FALSE)

# manipulating data
df <- ilc_di11 %>% filter(age=="TOTAL",
                         geo %in% c("EU28", "BE","BG","CZ","DK","DE","EE","IE","EL","ES","FR","HR","IT","CY","LV","LT","LU","HU","MT","NL","AT","PL","PT","RO","SI","SK","FI","SE","UK"),
                         sex=="T",
                         year(time) %in% c("2008", "2012", "2016"))%>%
        spread(time, values) %>%
        rename(`2008`=`2008-01-01`,
               `2012`=`2012-01-01`,
               `2016`=`2016-01-01`)%>%
        gather(var, values, 5:6)

# plotting data
ggplot(data=df)+
        geom_bar(aes(x=reorder(geo, `2016`), y=`2016`),
                 stat = "identity", 
                 position = "dodge",
                 fill="steelblue")+
        geom_point(aes(x=reorder(geo, `2016`), y=values, 
                       shape=var,
                       color=var),
                   size=3)+
        labs(title="Pajamų nelygybė nustatyta pagal S80/S20 santykį", 
             subtitle="Šaltinis: Eurostat (ilc_di11)", 
             x="Šalys", 
             y="Santykis")

Answer 1

有许多解决方案，其中之一就是将虚拟data.frame传递到单独的geom_point层。将此点的大小设置为0，这样它就不会显示在绘图中，而是为其指定颜色和形状。

ggplot(df, aes(shape = var, color = var)) +
    geom_bar(aes(reorder(geo, `2016`), `2016`), 
             stat = "identity", position = "dodge", fill = "steelblue", color = NA) +
    geom_point(aes(x, y),
               data.frame(x = Inf, y = Inf, var = "2016", var = "2016"),
               size = 0) +
    geom_point(aes(reorder(geo, `2016`), values, shape = var, color = var), 
               size = 3) +
    labs(title    = "Pajamų nelygybė nustatyta pagal S80/S20 santykį", 
         subtitle = "Šaltinis: Eurostat (ilc_di11)", 
         x        = "Šalys", 
         y        = "Santykis",
         color    = "Year",
         shape    = "Year")