flying-sheep / ggplot.multistats

Multiple Summary Statistics for Binned Stats/Geometries

ggplot.multistats

ggplot.multistats currently provides stat_summaries_hex and some helpers.

stat_summaries_hex is similar to ggplot2::stat_summary_hex, but allows specifying multiple stats using the funs parameter (see Example).

Installation

ggplot.multistats is on CRAN.

install.packages('ggplot.multistats')

You can also install the development version from GitHub:

# install.packages('devtools')
devtools::install_github('flying-sheep/ggplot.multistats')

Example

Specify a summary variable using the z aesthetic and specify a list of funs to provide after_stats for you:

library(ggplot2)
library(ggplot.multistats)

ggplot(iris, aes(Sepal.Width, Sepal.Length)) +
  stat_summaries_hex(
    aes(z = Petal.Width, fill = after_stat(median), alpha = after_stat(n)),
    funs = c('median', n = 'length'),
    bins = 5
  )

ggplot.multistats's Issues

Unable To Reproduce Example, z Aesthetic Is Dropped:

stat_summaries_hex seems to be exactly what I'm looking for!

However, I'm unable to reproduce the example and am wondering if I'm missing something or if an update has caused some incompatibility? When specifying both fill and alpha values with stats, only the alpha value appears to make an effect. If I remove the alpha parameter, then the hex bins are colored by n but only when alpha is not specified.

In addition, I'm receiving the warning message that the z aesthetic was dropped during statistical transformation - I assume that shouldn't happen, as the z aesthetic is what I'm trying to transform!

I have ggplot.multistats 1.0.0 and ggplot2 3.4.0 installed.

library(ggplot2)
library(ggplot.multistats)

ggplot(iris, aes(Sepal.Width, Sepal.Length)) +
  stat_summaries_hex(
    aes(z = Petal.Width, fill = stat(median), alpha = stat(n)),
    funs = c('median', n = 'length'),
    bins = 5
  )

Release v1.0.1

CRAN team is on vacation rn

Use with categorical variable

Hi, this is an excellent package.

Do you have any suggestion on what's the best way to represent a categorical variable as well?
E.g. if in any hexagon I wanted to see which of my factor levels is most represented as a colour from a qualitative palette.

Cheers!