Create a concentric circle legend for a ggplot bubble chart

Question

I am trying to recreate this visualization of a bubble chart using ggplot2 (I have found the code for doing this in R, but not with the ggplot2 package). This is what I have so far. There are some other errors with my code at the moment, but I want to have the legend show concentric circles for size, versus circles shown in rows. Thanks for your help!

Original visualization:

My reproduction:

My (simplified) code:

crime <-
read.csv("http://datasets.flowingdata.com/crimeRatesByState2005.tsv",
       header=TRUE, sep="\t")
ggplot(crime,
       mapping= aes(x=murder, y=burglary))+
    geom_point(aes(size=population), color="red")+
    geom_text(aes(label=state.name), show.legend=FALSE, size=3)+ 
    theme(legend.position = c(0.9, 0.2))

Dorweiler

Answer 1

Here's an approach where we build the legend as imagined from scratch.

1) This part slightly tweaks your base chart.

Thank you for including the source data. I missed that earlier and have edited this answer to use it. I switched to a different point shape so that we can specify both outside border (color) as well as interior fill.

gg <- ggplot(crime,
       mapping= aes(x=murder, y=burglary))+
  geom_point(aes(size=population), shape = 21, color="white", fill = "red")+

  ggrepel::geom_text_repel(aes(label = state.name),
                           size = 3, segment.color = NA,
                           point.padding = unit(0.1, "lines")) +
  theme_classic() +

  # This scales area to size (not radius), specifies max size, and hides legend
  scale_size_area(max_size = 20, guide = FALSE)

2) Here I make another table to use for the concentric legend circles

library(dplyr); library(ggplot2)
legend_bubbles <- data.frame(
  label = c("3", "20", "40m"),
  size  = c(3E6, 20E6, 40E6)
) %>%
mutate(radius = sqrt(size / pi))   

3) This section adds the legend bubbles, text, and title.

It's not ideal, since different print sizes will require placement tweaks. But it seems like it'd get complicated to get into the underlying grobs with ggplot_build to extract and use those sizing adjustments...

gg + geom_point(data = legend_bubbles,
             #  The "radius/50" was trial and error. Better way?
             aes(x = 8.5, y = 250 + radius/50, size = size),
             shape = 21, color = "black", fill = NA) +
  geom_text(data = legend_bubbles, size = 3,
            aes(x = 8.5, y = 275 + 2 * radius/50, label = label)) +
  annotate("text", x = 8.5, y = 450, label = "Population", fontface = "bold")

Mesosphere

Thank you for your help! This is a good work around. Is there no way of displaying like that natively; we have to create a new dataframe and superimpose it? – Dorweiler

Not that I'm aware of, but I'd love to know if someone has built one already. Seems like a useful ggplot addon. – Mesosphere

I´m in search for a native solution as well. I found one here riatelab.github.io/cartography . Its not ggplot, but I share it anyway. – Hightail