Powered by AppSignal & Oban Pro

Hexagonal bins: geom_hex

notebooks/geom_hex.livemd

Hexagonal bins: geom_hex

Mix.install([
  {:giggly, path: Path.join(__DIR__, "..")},
  {:kino, "~> 0.19"}
  ])

alias Giggly.Plot

diamonds = Giggly.Datasets.diamonds()

render = fn plot ->
  plot
  |> Giggly.to_svg()
  |> IO.iodata_to_binary()
  |> Kino.Image.new(:svg)
end

About this notebook

This notebook follows the examples on ggplot2's reference page for geom_hex(), which covers stat_bin_hex() too, in the order the page gives them. Each example shows ggplot2's code first and then Giggly's. An example Giggly cannot draw yet keeps its ggplot2 code and names what it is waiting for, so the page is also a list of what remains.

geom_hex() is a heatmap of counts in two dimensions: its stat, stat_bin_hex(), divides the plane into regular hexagons and counts the rows that fall in each, and the count is mapped to the hexagon's fill. Hexagons avoid the artefacts the strict alignment of square bins (geom_bin_2d()) can produce. The size of the bins is set by their number or their width, for x and y together or each on its own. Giggly draws hexagonal bins with Giggly.Plot.geom_hex/3, and Giggly.Plot.stat_bin_hex/3 adds the same layer; bins: takes one number of bins or a pair, x's first, and binwidth: one width or a pair. A pair of numbers of bins divides each axis's extent by its own, as ggplot2's documentation says, where ggplot2 4.0.3 divides x's extent by both. A plot is described with Giggly.Plot and drawn with Giggly.to_svg/2, which returns an SVG document. The setup cell aliases Giggly.Plot, so the cells write Plot.new/2, and defines render, which draws a plot and shows it below the cell with Kino.Image. The cells draw Giggly.Datasets.diamonds/0, ggplot2's diamonds: the prices and other attributes of almost 54,000 diamonds, which the setup cell binds to diamonds, ggplot2's name for it.

Counting in hexagons

d <- ggplot(diamonds, aes(carat, price))
d + geom_hex()

Each hexagon's count is mapped to its fill, so the plot has a legend for the counts.

d = Plot.new(diamonds, %{x: :carat, y: :price})

d
|> Plot.geom_hex()
|> render.()

The number of bins

# You can control the size of the bins by specifying the number of
# bins in each direction:
d + geom_hex(bins = 10)
d + geom_hex(bins = 30)

Thirty bins along each axis is the default, so the second plot is the first again.

d
|> Plot.geom_hex(bins: 10)
|> render.()
d
|> Plot.geom_hex(bins: 30)
|> render.()

The width of the bins

# Or by specifying the width of the bins
d + geom_hex(binwidth = c(1, 1000))
d + geom_hex(binwidth = c(.1, 500))

A pair of widths gives the width along x first, then along y, which Giggly writes as a tuple, {1, 1000} for c(1, 1000).

d
|> Plot.geom_hex(binwidth: {1, 1000})
|> render.()
d
|> Plot.geom_hex(binwidth: {0.1, 500})
|> render.()