In csdaw/ggbeeswarm2: Categorical Scatter (Violin Point) Plots

knitr::opts_chunk$set(
  fig.retina=2,
  fig.width=6,
  fig.height=4,
  fig.path = "man/figures/README-"
)

ggbeeswarm2: Beeswarm-style plots with ggplot2

Note

This is a fork of ggbeeswarm which is in an experimental and unfinished state. Some of changes made here including additional swarming and corral methods are now available in the development version of ggbeeswarm, which is probably what you should be using.

I plan on trying out re-writing these functions (again) so that the swarming is performed based on the plotting area and size of the plotting character as opposed to just the x/y range of the data (as in ggbeeswarm).

The position_beeswarm function has been rewritten and is compatible with R >= v4.0.0. Additionally, position_beeswarm works almost identically to the beeswarm function from the beeswarm package with all the additional methods implemented (including the new compactswarm).

For example:

set.seed(123)
distro <- list(runif = runif(100, min = -3, max = 3), 
               rnorm = rnorm(100))

# beeswarm
beeswarm::beeswarm(distro, 
                   col = 2:3, pch = 16,
                   method = "hex",
                   main = "title")

# ggbeeswarm2
library(dplyr)
library(ggbeeswarm2)

distro2 <- as.data.frame(rev(distro)) %>% 
  tidyr::pivot_longer(
    everything(), 
    values_to = "value", 
    names_to = "variable"
    )

 ggplot2::ggplot(distro2, aes(x = rev(variable), y = value)) + 
   geom_beeswarm(method = "hex") + 
   scale_y_continuous(limits = c(-3, 3))