Borealis
Borealis

Reputation: 8480

Add legend to ggplot line graph based on linetype

I am trying to add a legend to a ggplot graph that uses solid and dashed lines.

require(ggplot2)

DATE <- c("2020-10-14", "2020-10-15", "2020-10-16", "2020-10-17", "2020-10-18")
TMAX <- c(47, 45, 43, 40, 4)
TMIN <- c(35, 34, 28, 26, 29)

df <- data.frame(DATE, TMAX, TMIN)

ggplot(data = df, aes(x = DATE, y = TMIN, group = 1)) +
  geom_path(linetype = 1, size = 1.5) +
  labs(x = "Date",
       y = "Temp (F)") +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1)) + 
  geom_path(data = df, linetype = 2, size = 1.5, aes(x = DATE, y=TMAX)) 

enter image description here

A similar question states that I should include linetype within aes, although this does not yield a legend. For example:

ggplot(data = df, aes(x = DATE, y = TMIN, group = 1, linetype = 1)) +
  geom_path(size = 1.5) +
  labs(x = "Date",
       y = "Temp (F)") +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1)) + 
  geom_path(data = df, size = 1.5, aes(x = DATE, y=TMAX, linetype=2)) 

Here is the error:

Error: A continuous variable can not be mapped to linetype

How can I add a legend to the figure showing both my solid line and dashed line?

Upvotes: 1

Views: 235

Answers (2)

Duck
Duck

Reputation: 39613

Try this reshaping data to long, and then using a variable in the linetype statement. In that way you can obtain the legend. Here the code:

require(ggplot2)
require(tidyr)
require(dplyr)
#Data
DATE <- c("2020-10-14", "2020-10-15", "2020-10-16", "2020-10-17", "2020-10-18")
TMAX <- c(47, 45, 43, 40, 4)
TMIN <- c(35, 34, 28, 26, 29)
df <- data.frame(DATE, TMAX, TMIN)
#Plot
df %>% pivot_longer(-DATE) %>%
  ggplot(aes(x = DATE, y = value, group = name,linetype=name)) +
  geom_path(size = 1.5) +
  labs(x = "Date",
       y = "Temp (F)") +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))+
  scale_linetype_manual(values=c(2,1))+labs(linetype='Var')+
  guides(linetype = guide_legend(override.aes = list(size = 0.5)))

Output:

enter image description here

Upvotes: 2

Allan Cameron
Allan Cameron

Reputation: 174278

The alternative to reshaping if you only have two lines is to put the linetype as a character assignment inside aes

ggplot(data = df, aes(x = DATE, y = TMIN, group = 1)) +
  geom_path(aes(linetype = "TMIN"), size = 1.5) +
  geom_path(aes(y = TMAX, linetype = "TMAX"), size = 1.5) +
  labs(x = "Date", y = "Temp (F)") +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1))

enter image description here

Upvotes: 3

Related Questions