I already have 'color' and 'fill' legends. How can I control the colour of a lin-CodePudding

I'm trying to product a bar plot and line plot on the same axes, with a rectangle to show where data may be incomplete and vertical lines showing important dates.

library(dplyr)
library(ggplot2)
library(scales)

dates <- c("2022-01-01","2022-01-08","2022-01-15")
count_variables <- c('pass','fail')
counts <- c(30,28,26,10,11,12)

data_bar <- expand.grid(date = dates, count_variable = count_variables, stringsAsFactors = F)
data_bar$count <- counts
data_bar$date <- as.Date(data_bar$date)

ave_marks <- c(75,72,68)

data_line <- data.frame(date = dates,
                        perc = ave_marks,
                        perc_variable = 'average mark')
data_line$date <- as.Date(data_line$date)


ggplot()  
  # Rectangle covering last few days
  geom_rect(aes(xmin=max(data_bar$date) 4, xmax=max(data_bar$date), ymin=-Inf, ymax=Inf, 
                fill="Time period subject\nto reporting delays"), colour=NA, alpha = .2)  
  # Bar plot
  geom_bar(data=data_bar, aes(fill=count_variable, y=count, x=date), position="dodge", stat="identity")  
  scale_fill_manual(labels = c("Pass","Fail","Time period subject\nto reporting delays"), 
                    values = c("green","red","#e4004620"))  
  # Line plot
  geom_line(data=data_line, aes(x=date, y=perc/3, color=perc_variable))  
  #scale_color_manual(labels = c("Pass rate"), values = c("blue"))  
  # Vertical lines
  geom_vline(aes(xintercept = as.Date("2022-01-05"),linetype = "Exam change 1 announced"),color='orange')  
  geom_vline(aes(xintercept = as.Date("2022-01-06"),linetype = "Exam change 1"),color='orange')  
  geom_vline(aes(xintercept = as.Date("2022-01-12"),linetype = "Exam change 2 announced"),color='black')  
  geom_vline(aes(xintercept = as.Date("2022-01-13"),linetype = "Exam change 2"),color='black')  
  scale_color_manual(labels = c("Pass rate","Exam change 1 announced","Exam change 1",
                                "Exam change 2 announced","Exam change 2"), 
                     values = c("blue","orange","orange","black","black"))  
  scale_linetype_manual(values=c("longdash","solid","longdash","solid"),
                        limits=c("Exam change 1 announced","Exam change 1",
                                 "Exam change 2 announced","Exam change 2"),
                        labels = c("Exam change 1 announced","Exam change 1",
                                   "Exam change 2 announced","Exam change 2"))  
  scale_y_continuous(sec.axis = sec_axis(~.*3, name = 'Percentage'),
                     labels = comma)

The code produces this chart.

The chart looks how I want it to, apart from the legend, which does not show the difference in colour between the lines that represent the two exam changes. How can I make the legend show 'Exam change 1 announced' and 'Exam change 1' in orange, rather than black?

CodePudding user response：

One option to achieve your desired result would be to override the aesthetics applied to your linetype legend using the override.aes argument of guide_legend which allows you to set the colors for each legend item:

library(dplyr)
library(ggplot2)
library(scales)

ggplot()  
  # Rectangle covering last few days
  geom_rect(aes(
    xmin = max(data_bar$date)   4, xmax = max(data_bar$date), ymin = -Inf, ymax = Inf,
    fill = "Time period subject\nto reporting delays"
  ), colour = NA, alpha = .2)  
  # Bar plot
  geom_bar(data = data_bar, aes(fill = count_variable, y = count, x = date), position = "dodge", stat = "identity")  
  scale_fill_manual(
    labels = c("Pass", "Fail", "Time period subject\nto reporting delays"),
    values = c("green", "red", "#e4004620")
  )  
  # Line plot
  geom_line(data = data_line, aes(x = date, y = perc / 3, color = perc_variable))  
  # scale_color_manual(labels = c("Pass rate"), values = c("blue"))  
  # Vertical lines
  geom_vline(aes(xintercept = as.Date("2022-01-05"), linetype = "Exam change 1 announced"), color = "orange")  
  geom_vline(aes(xintercept = as.Date("2022-01-06"), linetype = "Exam change 1"), color = "orange")  
  geom_vline(aes(xintercept = as.Date("2022-01-12"), linetype = "Exam change 2 announced"), color = "black")  
  geom_vline(aes(xintercept = as.Date("2022-01-13"), linetype = "Exam change 2"), color = "black")  
  scale_color_manual(
    labels = c(
      "Pass rate", "Exam change 1 announced", "Exam change 1",
      "Exam change 2 announced", "Exam change 2"
    ),
    values = c("blue", "orange", "orange", "black", "black")
  )  
  scale_linetype_manual(
    values = c("longdash", "solid", "longdash", "solid"),
    limits = c(
      "Exam change 1 announced", "Exam change 1",
      "Exam change 2 announced", "Exam change 2"
    ),
    labels = c(
      "Exam change 1 announced", "Exam change 1",
      "Exam change 2 announced", "Exam change 2"
    )
  )  
  scale_y_continuous(
    sec.axis = sec_axis(~ . * 3, name = "Percentage"),
    labels = comma
  )  
  guides(linetype = guide_legend(override.aes = list(color = c("orange", "orange", "black", "black"))))