Add marginal density plot based on subset of plotted data

Question

I have a line plot p generated from a long/melted dataframe DF.

I would now like to add a marginal density plot along the y axis to show the distribution of score at t5. I have just come across ggExtra::ggMarginal, but the tutorials I've found only seem to work with scatter plots and not with subsets of the main plot's data.

---

Here's the structure of DF:

> str(DF)
'data.frame':   600 obs. of  3 variables:
 $ ID      : int  1 2 3 4 5 6 7 8 9 10 ...
 $ variable: Factor w/ 6 levels "t0","t1","t2",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ value   : num  0.737 0.33 0.526 0.737 0.357 ...

And here's my line plot p:

---

I tried this...

# the code that produces the line plot above
p <- ggplot(DF, aes(x = variable, y = value, color = ID, group = ID))+
  geom_line() +
  theme_bw() +
  ylim(0, 1) +
  labs(x = "Time", y = "Score", color = "ID", subtitle = "")+
  theme(legend.position = "none")

#trying to add marginal density plot based on subsetted t5 scores
p <- ggMarginal(p, data = DF[which(DF$variable == "t5"),], x = value, type = "density")

But it throws this error: Error: No geom_point layer was found in your scatter plot

Does ggMarginal only work with scatter plots? How might I add a marginal density plot of score at t5 to the right side y axis?

Answer 1

I'll go ahead and answer my own question following @ricoderks comment.

It seems the solution is to just plot an invisible scatter plot. Here's the code that worked:

p <- ggplot(DF, aes(x = variable, y = value, color = ID, group = ID))+
  geom_line() +
  geom_point(data = DF[which(DF$variable == "t5"),], alpha = 0)+
  theme_bw() +
  ylim(0, 1) +
  labs(x = "Time", y = "Score", color = "ID", subtitle = "")+
  theme(legend.position = "none")
p <- ggMarginal(p, type = "density", margins = "y", size = 6, color = "steelblue4")