plotperfect

A lightweight matplotlib style library for producing publication-ready benchmark plots with consistent styling, colorblind-safe palettes, and zero configuration.

Installation

pip install git+https://github.com/NASA-IMPACT/plotperfect.git

For local development:

pip install -e ./plotperfect

Gallery

Grouped Bar	Single Bar
Line Plot	Scatter Plot
Stacked Bar	Horizontal Bar

Quick Start

import plotperfect as S

S.apply_style()
fig, ax = S.new_figure()

ax.bar(["A", "B", "C"], [3, 7, 5], color=S.COLOR_PRIMARY)
S.style_ax(ax, title="Demo", xlabel="Category", ylabel="Value")
S.save(fig, "demo.png")

See example.py for complete examples of every plot type shown in the gallery.

Style Guide

Color Palette

Ten colorblind-safe colors based on the seaborn "colorblind" palette. Access individual colors via S.COLORS["blue"] or use S.PALETTE (an ordered list) for multi-series plots.

Semantic aliases:

Alias	Color	Hex
COLOR_PRIMARY	blue	#2A79DB
COLOR_SECONDARY	grey	#949494
COLOR_TERTIARY	green	#029E73

Typography

All text is rendered in DM Sans (bundled with the package, auto-registered at import).

Element	Size	Weight
Title	14 pt	Bold
Axis labels	14 pt	Bold
Tick labels	12 pt	Regular
Legend	14 pt	Regular
Annotations	11 pt	Regular

Layout

Property	Default
Figure size	10 x 7 in
DPI	300
Bar width	0.35
Bar edge width	1.5

Colorblind Accessibility

Every plot should be distinguishable without color:

• Bar plots — apply S.HATCHES (white hatch patterns) to differentiate series
• Line / Scatter plots — apply S.MARKERS (distinct shapes per series)

Axes

style_ax() enforces a clean, consistent look:

• Bold title and axis labels
• Horizontal grid lines behind data
• No vertical grid
• Top and right spines removed
• Bottom and left spines in #333333

Legend

add_legend() places the legend below the chart (bottom-center, outside the axes) in a single row with no title.

Annotations

Function	Default format	Notes
annotate_bars()	{val:.2f}	Pass total=N for percentage format
annotate_line()	{val:.2f}	Configurable offset and format

Both accept a custom fmt string with {val}, {pct}, and {total} placeholders.

API Reference

Function	Description
apply_style()	Set global matplotlib rcParams. Call once per script.
new_figure()	Return a styled (fig, ax) pair.
style_ax()	Apply grid, spines, labels, and limits to an axes.
annotate_bars()	Add value labels above each bar.
annotate_line()	Add value labels at each data point on a line.
add_legend()	Place a bottom-center legend outside the axes.
save()	Save the figure at 300 DPI with tight bounding box.

License

MIT