Ggplot2: Change legend title while controlling line types and colors

tidyverse
1

I need to be able to control line types and colors in a plot, but also to change the Legend title. I can do either, but I have not been able to figure out how to do both. First, set up the data.

lcols = cbbPalette[c(2,3,4,1)]
linetype = rep(c('solid', 'dashed'),2)
LegendTitle = "Treatment"
x=rep(1:10,4)
y=c(1*1:10, 0.5*1:10, 0.25*1:10, 0.125*1:10)
cohort = factor(rep(LETTERS[1:4], each = 10))
df1 <- data.frame(x=x, y = y, cohort = cohort)

I can control the line type and color as follows (The bold stuff is lines commented out):

p <- ggplot(df1, aes(x,y, group = cohort)) +
  geom_line(aes(linetype = cohort, col = cohort)) +
  scale_color_manual(values = lcols) +
  scale_linetype_manual(values = linetype) +
  theme(legend.position = c(.20, .65))# +
#  scale_linetype_discrete(name = LegendTitle) +
#  scale_color_discrete(name = LegendTitle) 
p

image

or I can change the legend title:

p <- ggplot(df1, aes(x,y, group = cohort)) +
  geom_line(aes(linetype = cohort, col = cohort)) +
#  scale_color_manual(values = lcols) +
#  scale_linetype_manual(values = linetype) +
  theme(legend.position = c(.20, .65)) +
  scale_linetype_discrete(name = LegendTitle) +
  scale_color_discrete(name = LegendTitle) 
p

image

But I can't do both together.

p <- ggplot(df1, aes(x,y, group = cohort)) +
  geom_line(aes(linetype = cohort, col = cohort)) +
  scale_color_manual(values = lcols) +
  scale_linetype_manual(values = linetype) +
  theme(legend.position = c(.20, .65)) +
  scale_linetype_discrete(name = LegendTitle) +
  scale_color_discrete(name = LegendTitle) 
> Scale for 'linetype' is already present. Adding another scale for
'linetype', which will replace the existing scale.
Scale for 'colour' is already present. Adding another scale for 'colour',
which will replace the existing scale.
p

There must be a more straightforward way to change the text of a legend than what I used in the second example above, but I have been unable to find it after working on this for half a day. Does anyone know how I can make this work? As always, thanks in advance for any help you can give me.
Larry Hunsicker

2

You're almost there !
You can use the name argument in each scale_xxx_manual to change the title of the legend and merge both legend in one

lcols = metagen::cbbPalette[c(2,3,4,1)]
linetype = rep(c('solid', 'dashed'),2)
LegendTitle = "Treatment"
x=rep(1:10,4)
y=c(1*1:10, 0.5*1:10, 0.25*1:10, 0.125*1:10)
cohort = factor(rep(LETTERS[1:4], each = 10))
df1 <- data.frame(x=x, y = y, cohort = cohort)

library(ggplot2)

p <- ggplot(df1, aes(x,y, group = cohort)) +
  geom_line(aes(linetype = cohort, col = cohort)) +
  scale_color_manual(name = LegendTitle, values = lcols) +
  scale_linetype_manual(name = LegendTitle, values = linetype) +
  theme(legend.position = c(.20, .65))
p

Created on 2018-09-26 by the reprex package (v0.2.1)

Is this what you want ?

5 Likes
3

Excellent!! I KNEW that there had to be a straightforward way. I had looked up the help for scale_color_manual() but had missed the "name=" option. Many thanks!!

4

Just an FYI, you can also accomplish this by specifying the names with the labs function.

lcols = metagen::cbbPalette[c(2,3,4,1)]
linetype = rep(c('solid', 'dashed'),2)
LegendTitle = "Treatment"
x=rep(1:10,4)
y=c(1*1:10, 0.5*1:10, 0.25*1:10, 0.125*1:10)
cohort = factor(rep(LETTERS[1:4], each = 10))
df1 <- data.frame(x=x, y = y, cohort = cohort)

library(ggplot2)

p <- ggplot(df1, aes(x,y, group = cohort)) +
  geom_line(aes(linetype = cohort, col = cohort)) +
  scale_color_manual(values = lcols) +
  scale_linetype_manual(values = linetype) +
  theme(legend.position = c(.20, .65)) + 
  labs(
    color = LegendTitle,
    linetype = LegendTitle
  )
p

Created on 2018-09-26 by the reprex package (v0.2.0).

1 Like