Usage Guide¶

dartwork-mpl bundles stylistic presets, curated colors/colormaps, and layout/font helpers so you get predictable results fast.

The Evolution of a Plot

Drag the slider to experience the dartwork-mpl pipeline step-by-step.

1. Base 2. Apply Style 3. Label & Format 4. Smart Layout

Choosing the right tools¶

Which preset?¶

Your context	Preset	Why
Academic paper / journal	`scientific`	Compact fonts (7.5 pt), all 4 spines, fits 3.5” column
Business report / memo	`report`	Clean look, no top/right spines, 8 pt fonts
Infographic / editorial	`minimal`	Tufte-style, no spines/ticks, data-ink focus
PowerPoint / Keynote	`presentation`	Large fonts (10.5 pt), readable when projected
Conference poster	`poster`	Largest fonts (12 pt), thick lines, 1-2m reading distance
Web docs / blog	`web`	Screen-optimized 11 pt, clean for HTML/Jupyter
Dark Jupyter / dark slides	`dark`	Inverted colors on `#1e1e1e` background
Any of the above + Korean	Add `-kr` suffix	Swaps to Paperlogy/Pretendard fonts

Which layout function?¶

Situation	Function	Notes
Most figures	`simple_layout(fig)`	Fast, consistent margins via L-BFGS-B optimizer
Long labels or multi-line titles	`auto_layout(fig)`	Detects and fixes text overflow automatically
Multi-panel with GridSpec	`simple_layout(fig, gs=gs)`	Respects your `hspace`/`wspace` settings
Need exact margin control	`simple_layout(fig, margins=(...))`	Specify margins in inches

Which color palette?¶

Style	Prefix	Best for
Open Color	`oc.*`	General purpose, well-balanced (recommended default)
Tailwind CSS	`tw.*`	Web-style designs, wider shade range (50–950)
Material Design	`md.*`	Google-style, vibrant
Ant Design	`an.*`	Enterprise UI aesthetics
Chakra UI	`ch.*`	Modern web app colors
Primer	`pr.*`	GitHub-style, subtle

Typical workflow¶

1. Pick a style

dm.style.use("scientific") sets fonts, line weights, spines, and tick styling in one call — 7 presets cover papers, reports, slides, posters, web, and dark mode. No manual rcParams guessing.

→ Styles and Presets

2. Add color

Use named colors like "oc.blue5" or "tw.emerald500" anywhere matplotlib accepts a color string. 900+ curated swatches from 6 design systems, plus perceptual OKLCH interpolation.

→ Colors and Colormaps

3. Layout & annotate

dm.simple_layout(fig) optimizes margins via L-BFGS-B — uniform spacing even with colorbars and long labels. dm.label_axes() adds panel labels automatically.

→ Layout and Typography

4. Export

dm.save_formats(fig, "output/fig", formats=("png", "svg")) writes multiple formats at once. Add validate=True to auto-detect overflow, overlap, and crowding before saving.

→ Save and Validation

5. Interactive UI

Launch a local web UI to tweak parameters with sliders, download plots, and generate reproducible Python scripts on the fly.

→ Interactive UI

Tip

Where do assets live? Style files are in asset/mplstyle/, colors in asset/color/, fonts in asset/font/, and figure constants in dm.SW, dm.DW. See Design Philosophy for the architecture behind these choices.