Line graph with multiple factors in ggplot2

Question

I currently have a graph that looks okay but I am trying to convert it to a line graph. However, I cannot seem to get it to work and create multiple factors like I have previously. How can I convert my first plot to a line graph with 4 lines?

This is my original graph, with the code as follows (I know there are some weird mistakes):

ggplot(data = temp, aes(x = test, y = value, shape = factor(CASE.CONTROL))) + 
  stat_summary(geom = "point", fun.y = "mean", size = 5, aes(col = factor(gender)))+
  scale_shape_manual(name = "Case/Control", values = c(2,3))+
  scale_colour_manual(name = "Gender", values = c("red","blue"))

This is what my line graph looks like so far, but obviously there is plenty wrong with it. I need to have 4 lines, as in, male/case, female/case, male/control, female/control.

ggplot(data = temp, aes(x = test, y = value, group = factor(gender),group = factor(CASE.CONTROL))) + 
  stat_summary(geom = "line", fun.y = "mean", size = 2, aes(group = factor(CASE.CONTROL)))+
  stat_summary(geom = "line", fun.y = "mean", size = 2, aes(group = factor(gender)))+
  stat_summary(geom = "line", fun.y = "mean", size = 2, aes(col= factor(gender)))

This is what the line graph looks like so far but obviously there are major issues. I must be simply missing something simple here, how can I make a line graph that shows equivalent info from the first one?

You say obviously there are issues, but without knowing exactly what you want, it's unclear what those issues are. If you want lines through your data points, just use `geom_line` (call it *once*) and map color to whatever variable you're splitting lines by. It's hard to help beyond that without a [reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) — camille, Jan 21 '20 at 15:07
Share output of `dput(temp)`. [edit] your question and add the results of that in the body of your post. — M--, Jan 21 '20 at 15:08

score 3 · Answer 1 · answered Jan 21 '20 at 16:07

You can create a group for the lines first to make it easier:

set.seed(100)
temp = data.frame(
test=rep(1:5,each=3,times=4),
value=rnorm(60,rep(1:4,each=15),0.5),
CASE.CONTROL = rep(c("case","control"),each=30),
gender = rep(c("M","F"),each=15,times=4)
)

temp$line_group = with(temp,paste(gender,CASE.CONTROL))

I think it's quite confusing to have 4 colors again for your different groups, but try the 2 suggestions below.

only black lines:

ggplot(data = temp, aes(x = test, y = value, shape = factor(CASE.CONTROL))) + 
  stat_summary(geom = "point", fun.y = "mean", size = 5, aes(col = factor(gender)))+
  scale_shape_manual(name = "Case/Control", values = c(2,3))+
  scale_colour_manual(name = "Gender", values = c("red","blue")) +
  stat_summary(geom = "line", fun.y = "mean", size = 1, 
  aes(group = line_group),col="black")

Or a consistent color (I changed one of the case/control to circle so we can see it):

ggplot(data = temp, aes(x = test, y = value, shape =line_group,col=line_group)) + 
  stat_summary(geom = "point", fun.y = "mean", size = 5)+
  scale_shape_manual(values = c(2,1,2,1))+
  scale_colour_manual(values = c("red","red","blue","blue")) +
  stat_summary(geom = "line", fun.y = "mean", size = 1, 
  aes(col=line_group))

Line graph with multiple factors in ggplot2

1 Answers1