camminady_altair_theme

A minimal Altair theme: muted gray text, dashed grid lines, no frame around the plot, and horizontal axis titles at the end of each axis. It comes in a light (camminady) and a dark (camminady_dark) variant.

Installation

uv add camminady-altair-theme
# or
pip install camminady-altair-theme

Requires Python 3.11+ and Altair 6.

Usage

import altair as alt

alt.theme.enable("camminady")  # or "camminady_dark"

There is no need to import the package: it registers both themes through Altair's theme entry points. Importing camminady_altair_theme registers the themes but does not enable them (before 0.2.0, importing the package enabled the theme).

To use the theme for a single chart only:

with alt.theme.enable("camminady"):
    chart.save("chart.png")

Customising

All colours, font sizes and default dimensions are fields of Config. Build your own variant from it:

import altair as alt
from camminady_altair_theme import Config, build_theme


@alt.theme.register("camminady_small", enable=True)
def camminady_small() -> alt.theme.ThemeConfig:
    return build_theme(Config(width=400, height=250))

Individual charts can still override anything with chart.configure_*(...).

Colour-blind safety

The theme keeps Vega's default tableau10 category palette. Its third (red) and fifth (green) colours are nearly indistinguishable for people with deuteranopia, so for five or more categories use the bundled Okabe-Ito palette:

from camminady_altair_theme import OKABE_ITO

chart.encode(color=alt.Color("group:N").scale(range=OKABE_ITO))

Layout notes

• Single views are 800 × 400 px by default, for continuous and discrete axes.
• The y-axis title sits horizontally above the y axis and the x-axis title at the right end of the x axis. Legends (also with orient="top") and facet headers are placed so that they do not overlap these titles.
• X-axis labels are horizontal, also for categorical axes. For many or long category names, rotate them (alt.Axis(labelAngle=-45)) or use a horizontal bar chart.
• With an x axis at the top (orient="top"), the y-axis title overlaps its labels; move it up with alt.Y(..., axis=alt.Axis(titleY=-45)).

Examples

A simple chart.

And a faceted version.

Development

uv sync
uv run pytest
uv run ruff check . && uv run ruff format --check .
uv run ty check
uv run python create_charts.py  # regenerate the images above