Real-World Dashboards¶

This gallery showcases complete, production-ready dashboard examples that combine multiple dartwork-mpl features — composite layouts, semantic colors, hierarchical typography, and auto-layout — on realistic sample data.

Each example focuses on one chart in one figure so that the techniques are easy to isolate, remix, and drop into a larger multi-panel report.

Example Domain ¶

All examples in this gallery render a consistent synthetic dataset: a small fleet of environmental sensors reporting temperature, humidity, and power draw at several sites over time. The choice of domain is incidental — the visual patterns (dual-axis trends, stacked components, threshold overlays, grouped comparisons, summary cards) transfer directly to any time series or categorical data.

Techniques Demonstrated ¶

Composite dashboards: plt.figure + GridSpec for multi-panel layouts
Dual-axis time series: bar + line on shared x-axis with a secondary y-axis
Stacked area charts: component mix over time
Threshold-annotated trends: reference lines and shaded bands
Grouped bar comparisons: side-by-side categories across periods
Text-only summary cards: axis(“off”) + ax.text for KPI tiles

Conventions ¶

All dashboards follow dartwork-mpl conventions:

Preset styles: dm.style.use("report") (or "report-kr" for Korean copy) is applied before any figure is created.
Relative typography: titles and labels use dm.fs(n) so that a preset change rescales the whole figure uniformly.
Named colors: charts pull from the oc.* and tw.* palettes instead of raw hex strings.
Auto-layout: dm.simple_layout(fig) is called before plt.show so that margins adapt to actual content rather than static values.
Single chart per file: each example renders exactly one figure. To assemble a multi-panel dashboard, compose the individual snippets inside a single GridSpec.