Colour¶

Stay in ink until colour distinguishes series, then take it from a colour-blind-safe scheme.

Colour is yours to choose; vanzelfsprekend only chooses the greys. Three carry the roles: DATA_INK for the marks, TEXT_INK for labels and titles, LINE_INK for the frame, darkest for the data and lightest for the furniture. apply installs the neutral ink cycle in place of matplotlib's default, so a lone series stays near-black and nothing turns colourful on its own. A cycle you set yourself, before or after the call, is left alone, and restore gives matplotlib's back.

Opt into colour once it tells series apart, with a scheme cycle installed before you draw. Keep the lines solid and directly labelled: a dash pattern or a legend would say the same thing twice.

ax.set_prop_cycle(vzs.palettes.cycle("muted"))

The schemes are Paul Tol's, colour-blind-safe and ordered by his technote's picking sequence, so in each row below the colours to reach for first are on the left. A single colour goes anywhere matplotlib takes one, by name: tol:orange is the vibrant default, and tol:muted.rose or tol:bright.blue reaches into a scheme. The tol: prefix keeps these names apart from matplotlib's own in its global registry; cycle takes the bare scheme name, "muted", since it already lives in vanzelfsprekend's namespace. vzs.palettes.SCHEMES enumerates them in code.

bright

• blue
• red
• green
• yellow
• cyan
• purple
• grey

high_contrast

• blue
• yellow
• red

vibrant

• orange
• blue
• cyan
• magenta
• red
• teal
• grey

muted

• rose
• indigo
• sand
• green
• cyan
• wine
• teal
• olive
• purple
• pale_grey

medium_contrast

• light_blue
• dark_blue
• light_yellow
• dark_red
• dark_yellow
• light_red

pale

• blue
• cyan
• green
• yellow
• red
• grey

dark

• blue
• cyan
• green
• yellow
• red
• grey

light

• light_blue
• orange
• light_yellow
• pink
• light_cyan
• mint
• pear
• olive
• pale_grey

Hand a finished plot to apply and it trims the frame and greys the furniture, touching nothing you drew. When you draw with restraint in mind from the start, three knobs each own one thing: apply, or range_frame and mute as separate steps, the frame and its ink; cycle the colour; the vanzelfsprekend style the marks' geometry, lighter lines, smaller marks, quieter titles. None overlaps, so you compose the ones you want, and apply is the last word:

with plt.style.context("vanzelfsprekend"):
    fig, ax = plt.subplots()
    ax.set_prop_cycle(vzs.palettes.cycle("muted"))
    ax.plot(...)
    vzs.apply(ax, frame="data")
    fig.savefig("figure.png")

Keep the drawing, apply and save inside the style.context: matplotlib reads these defaults when it renders, not when you call plot, so leaving the context early drops them before the figure is drawn.