Plotting

Three plotting backends share the same API and accept the same arguments. Pick whichever matches your output target.

SVG (built-in, no dependencies)

from pedotri.plot import render_svg, TextureDiagram

svg = render_svg("USDA", title="USDA Soil Texture Triangle")

# Save to file
TextureDiagram("GEPPA", locale="fr").save("geppa.svg")

# Display inline in Jupyter
TextureDiagram(
    "USDA",
    points=[(40, 25), (15, 60)],
    point_labels=["A", "B"],
    title="Two field samples",
)

The SVG renderer is self-contained — no matplotlib, no plotly, just string-built SVG. Files are small and embed cleanly in HTML, Markdown, GitHub READMEs, and Jupyter notebooks (via _repr_svg_).

Matplotlib (pip install pedotri[matplotlib])

from pedotri.plot import render_mpl

fig = render_mpl(
    "USDA",
    points=[(40, 25)],
    point_labels=["sample"],
    title="USDA",
    figsize=(8, 7),
)
fig.savefig("usda.pdf")

Returns a matplotlib.figure.Figure you can further annotate, compose into a multi-panel figure, or export to any matplotlib-supported format.

Plotly (pip install pedotri[plotly])

from pedotri.plot import render_plotly

fig = render_plotly(
    "USDA",
    points=[(40, 25)],
    point_labels=["sample"],
)
fig.write_html("usda.html")   # interactive standalone page

The Plotly backend uses Plotly's native Scatterternary, so you get hover tooltips per class, zoom, pan, and the ability to export to interactive HTML.

Common arguments

Argument	Default	Meaning
classification	required	Classification key ("USDA", "GEPPA", …) or Classification instance.
points	None	Iterable of (sand, clay) pairs for 2-D, or 1-tuples / scalars for 1-D.
point_labels	None	Labels rendered alongside each marker.
locale	classification default	Locale tag for class names in the legend / tooltips.
title	None	Title above the diagram.
show_legend	True	Toggle the legend. (SVG / matplotlib.)
show_grid	True	10 % gridlines inside the triangle. (SVG / matplotlib.)

1-D classifications

Classifications with a single axis (e.g. KACHINSKY) render as a horizontal banded bar rather than a triangle in all three backends — class regions are coloured by group, tick labels mark each interval boundary, and point overlays are drawn as markers along the bar.