Combined line & bar geoms: How to generate proper legend?

心已入冬 提交于 2019-12-30 02:21:06

问题


The legend for d2 looks fine; for d1, I would like to show just the hoizontal line against a white/transparent backgounnd.

df = data.frame(
  Date = c("2012-11-30", "2012-12-03", "2012-12-04"),
  d1 = c(9, 5, 11),
  d2 = c(4, 6, 3)
)
ggplot(df, aes(Date)) + 
  geom_bar(aes(y = d2, color = "d2"), stat="identity", fill = "red") +
  geom_line(aes(y = d1, group = 1, color = "d1"))  +
  scale_colour_manual("", values=c("d1" = "blue", "d2" = "red")) 


回答1:


It is not an elegant solution but at least it gives some result.

I added aes(fill="d2") in geom_bar() and removed fill="red". Then I added separate scales for line and for bars. Then in theme() I removed grey background from legend entry.

To ensure that d1 in legend is shown before d2 in scale_colour_manual(" ") there should be extra space between quotes ("longer" name).

To keep legend keys in one line, legend.box="horizontal" added to theme()

  ggplot(df, aes(Date)) + 
    geom_bar(aes(y = d2,fill="d2"), stat="identity") +
    geom_line(aes(y = d1, group = 1, color = "d1")) +
    scale_colour_manual(" ", values=c("d1" = "blue", "d2" = "red"))+
    scale_fill_manual("",values="red")+
    theme(legend.key=element_blank(),
          legend.title=element_blank(),
          legend.box="horizontal")




回答2:


# Bar graph: Notice the placement of fill argument in aes()  
  geom_bar(aes(y=prop.P*100, fill="Seropositive"), stat = "identity",
           position = "dodge", width = 0.5)+

# This line defines the legend for the bar
  scale_fill_manual(name="", values = c("Seropositive"="steelblue3"))+

# Adding errorbar
  geom_errorbar(aes(ymin = ciLow*100, ymax = ciHigh*100), width = 0.2,
                position = position_dodge(width=0.8))+

# Adding the line to the graph. Used the color argument within aes() to define custom colour
  geom_line(aes(y=mmr1_cov*100, group=1, color="MMR1 Coverage"), size=1)+

# Adding points represent the plotted data
  geom_point(aes(x=age, y=mmr1_cov*100), color="red", size=2)+

# Here adding the legend for the line graph and define the colour
  scale_color_manual(name="", values=c("MMR1 Coverage"="red"))+

# Label the axis
  labs(x="Age (Months)", y="Percentage")+

# Place the legend bottom of the graph
  theme(legend.position = "bottom")

Section of the graph displaying the results



来源:https://stackoverflow.com/questions/13789458/combined-line-bar-geoms-how-to-generate-proper-legend

易学教程内所有资源均来自网络或用户发布的内容,如有违反法律规定的内容欢迎反馈
该文章没有解决你所遇到的问题?点击提问,说说你的问题,让更多的人一起探讨吧!