hexbin

Gridded & matrix

Minimal call

import random
rng = random.Random(13)
xs = [rng.gauss(0, 1) for _ in range(3000)]
df = {"x": xs, "y": [x + rng.gauss(0, 1) for x in xs]}

c = pt.chart(df, aes(x="x", y="y"), xlabel="x", ylabel="y")
c.add_hexbin(gridsize=20)

Worked example

import random

import plotlet as pt
from plotlet import aes

rng = random.Random(13)
xs = [rng.gauss(0, 1) + rng.gauss(0, 0.4) for _ in range(3000)]
df = {"x": xs, "y": [x + rng.gauss(0, 1) for x in xs]}

hexes = pt.chart(df, aes(x="x", y="y"),
                 title="3000 points, binned", xlabel="x", ylabel="y",
                 data_width=300, data_height=260)
hexes.add_hexbin(gridsize=22)

c = hexes | pt.legend(hexes)

Docstring

Hexagonal density binning for crowded scatter; cells colored by point count.

For dense 2-D scatter, hex binning beats overplotting. Each hexagonal cell
counts the points that fall in it; cells are colored by count via a sequential
colormap. Used wherever a c.add_scatter() would devolve into a black blob.

Hexagons are sized and binned in pixel space so they always appear as regular
hexagons that tile without gaps regardless of the canvas aspect ratio.

API: c.add_hexbin(aes(x="col", y="col"))

Styling kwargs:
  gridsize=30    number of hex columns across the canvas width
  cmap='viridis' colormap name for coloring cells by count
  vmin=0         colormap domain lower bound
  vmax=None      colormap domain upper bound (defaults to max count)