Figurestead

Experimental scientific figures for Python: line, scatter, histogram, strip-summary and heatmap figures with explicit labels and themes.

Python 0.9.0a5

Requires Python >=3.10, Matplotlib >=3.7, NumPy >=1.24 and Pillow >=10. In a fresh virtual environment:

python -m pip install figurestead==0.9.0a5

from figurestead import PlotSpec, line, save_figure

figure, axes = line(
    [0, 1, 2], [[0, 1, 0]],
    theme="lavender_fog_notebook",
    spec=PlotSpec(title="Synthetic response", xlabel="Time (s)",
                  ylabel="Response (a.u.)", note="Synthetic example; not measurements."),
)
save_figure(figure, "response.png", dpi=150)
save_figure(figure, "response.svg")

The destination directory must exist. save_figure stages and replaces a single output file only after a successful save. Rendering errors propagate and leave an existing destination unchanged. For streams or intentionally external SVG images, use Matplotlib's figure.savefig and manage those outputs directly.

This version repairs note and sample-count placement, warns when explicit histogram edges exclude observations, and fits the optional signature only where there is measured spare space. Callers supplying axes own their margins; an insufficient source-note footer raises sizing guidance. See the frozen layout and export contract for the behavior incorporated into this version.

Numeric NaN/infinity category identities reject before string coercion; literal string labels remain valid. Appended labeled lines retain truthful accumulated legends; anonymous scatter/strip layers do not acquire scientific labels.

Inspect exports at their intended size. Shared Python/browser semantics do not imply identical pixels or matching renderer coverage. Reference themes have a bounded three-series line designation, not universal accessibility or physical print qualification.

Line behavior and limits · Source · Issue tracker

MIT licensed.