Question

I’m attempting to plot a large scatter plot that varies along a large number of dimensions.

Here’s my starting plot:

 p <- ggplot(mtcars, aes(wt, mpg, shape=cyl, colour=gear, size=carb)) +
      geom_point()

Using the mtcars dataset, I’m simply including various shapes, colours, and sizes. Now what i’d like to add to this is some distinguishing mark around all points with carb==8. For example, add in a big black circle around these points.

Doing

p + geom_point() + geom_point(colour="black", shape=1, size=5)

is close in that it gets the circles as I want, but all points get circles, I want to instead limit this to just some of the points.

I would greatly appreciate any suggestions for this problem.

Answer 1

Instead of initially putting your variables within the ggplot function, try to map your variables individually for every layer you add.

p <- ggplot(data = mtcars)+
geom_point(aes(wt, mpg, shape=as.factor(cyl), colour=gear, size=carb))+
geom_point(aes(wt[carb==8], mpg[carb==8]), colour="black", shape=1, size=7)