Examples

Prefer reading in the browser? The Manual walks through the core workflow with the live map embedded after each step, and the Tutorial covers everything below on one page. The notebooks here are the same material, runnable.

A set of small, focused, runnable notebooks in notebooks/. Each teaches one topic and is self-contained — open it in Jupyter and run top to bottom. They use the bundled sample data (notebooks/data/sodermalm_edges.gpkg, ~4,600 real road edges) and write any output to notebooks/output/ (git-ignored).

Notebook	What you'll learn
01 · Quickstart	Load edges → render_edges → .save(). Base maps and palettes (highsat/carto/mono).
02 · Data-driven styling	Colour roads by your own data — categorical (color_by + colors) and numeric (color_by + cmap + width_by) with legends.
03 · Filtering & highlighting	highway_types, include/exclude, and a neon selected= highlight.
04 · Outputs for the web	The canonical JSON spec (to_spec), to_geojson, save_spec/load_spec, to_html, to_iframe, save.
05 · Embedding with roadstyle.js	The decoupled path: Python bakes a JSON spec; the bundled roadstyle.js draws it in your own page with a custom sidebar. Generates a self-contained output/web/ folder.
06 · Customizing the look	Base maps + switcher, and custom colour vocabularies via color_by + colors.
07 · Large datasets (lonboard)	The same API on the GPU/WebGL backend for big edge sets.
08 · Loading data	Every input render_edges/to_spec accept: GeoDataFrame, file path, GeoJSON (+ spec), pyarrow Table, DuckDB (from_duckdb).
09 · Overlays	Extra layers (zones / POIs / lines) with Overlay: placement, popups, the Layers toggle, settings-driven styling, querying from JS.

Start at 01 if you're new; jump to 05 if your goal is embedding a roadstyle map in a website.

Example scripts

Self-contained scripts in examples/ — run any of them to write an HTML file you can open. They use the same bundled sample edges (with a couple of seeded columns where the demo needs data).

Script	What it shows
recolor_web_backend.py	Switchable data-driven colouring on the web backend (color_options + a Colour by dropdown), plus a custom panel wired through window.rsSetColorField / the rs:colorchange event.
overlays_web.py	Extra overlay layers — synthetic TAZ zones (under the roads) + POI circles (on top), both clickable, with a Layers toggle.
recolor_custom_panel.py	The same switchable colouring on the roadstyle.js spec page, driven by a custom addPanel + the setColorField / event-bus API.

The generated pages used in the Manual are built by docs/build_maps.py.