Color regression line depending on a condition / wrapping caption

Question

I am a R-noob and need help with designing a plot. I have the plot mostly done however I cannot change the design of one of the two regression lines. I would like to have one of the lines to be dotted so that the plot is understandable when printed in B&W. Also, I would like to describe the plot in the caption in accordance with APA style. However, the caption is too long. How can I get the caption to use several lines instead of one way too long line? This is what I have so far:

P1plotV1.1 <- ggplot(subPD, 
                     aes(x = subPD$Digit, y = subPD$Phoflu_tot, shape = Group, color=Group)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE) +
  labs(title="Primary analysis") + 
  labs(x="Digit span") + 
  labs(y="Phonemic fluency") +
  labs(caption="test")

Which gives me the plot attached. I would be happy about any suggestion on how I can change one of the regression lines to a dotted line.

Don't use `$` inside `aes`. Can't you just add `lty = Group` inside `aes`? You can line breaks in the caption with `\n`. — Axeman, Feb 24 '17 at 13:59
Provide your data so that your plot is reproducible (see [how to make a reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example/5965451#5965451)) or use a default dataset such as iris or cars. — Paul Rougieux, Feb 24 '17 at 15:25

Paul Rougieux · Answer 1 · 2017-02-24T15:51:26.487

Use the linetype aesthetic with a manual scale scale_linetype_manual to specify which groups should have a dotted or continuous line. For example using the iris dataset.

library(ggplot2)
ggplot(iris, 
       aes(x = Sepal.Length, y = Petal.Length, 
           shape = Species, color=Species, 
           linetype = Species)) + # linetype defines the dotted lines
    geom_point() +
    # Choose the linetype, in the order of the Species factor 
    # value 1 produces a continuous line, 
    # other values give various large or finely dotted lines
    scale_linetype_manual(values = c(1,2,3)) + 
    geom_smooth(method = "lm", se = FALSE) +
    labs(title="Primary analysis with a very long title \n and a new line if you prefer",
         subtitle = "Subtitle")

Thank you so much, that solves it! :) – Lukas Preis Feb 26 '17 at 10:08 — Lukas Preis, Feb 26 '17 at 10:08

Color regression line depending on a condition / wrapping caption

1 Answers1