Gallery

Every screenshot on this page is captured from a self-contained, runnable script in examples/ — each exposes a build() (returns the widget) and a main() (shows a window):

uv run python examples/01_hello.py

Showcase

Larger scenarios that combine several features around a realistic dataset.

Sensor telemetry

A rolling baseline, a 3σ tolerance Spread, flagged-anomaly Scatter, and an X-linked residual panel derived with an Expression — 26_telemetry_monitoring.py

Market analytics

Price Curve + 20/50-day moving averages + a Bollinger Spread, over an X-linked volume Bars panel — 27_market_analytics.py

Streaming telemetry

A live rolling feed (qv.stream), its 400k-point history datashaded, and a brush-driven detail panel wired through a Signal — 34_streaming_telemetry.py

The everyday figures

Step curve, stacked Area, horizontal Bars, Pie donut, Ecdf, filled Contour, SI ticks, a dual-axis pair, Quiver with a reference key, a boundary-level Mesh, Stem, and an annotated Heatmap — one grid, under the native toolbar — 35_everyday_figures.py

Linked dashboard

Three panels, shared X, brushing, dark theme — under sixty lines — dashboard_native.py

Big-data density map

2M categorized events datashaded into a categorical density map; hover reports the count under the cursor — 28_event_density_map.py

Gridded science

An xarray 2-D field as an Image map plus a 1-D cross-section Curve, and a 3-D cube overlaying 250 instance lines with a mean envelope — 29_climate_field.py · 30_xarray_sensor_lines.py

Mosaic layout

Layout.mosaic("AAB\nCCB", …) — spanning panes from an ASCII plan, track ratios, and a figure suptitle — 36_mosaic_layout.py

Named panes — the mosaic's list form names every subplot ("price", "volume", "depth"), link_x="col" keeps a column time-aligned, and the same names address the live render: view.pane("price").set_range(x=…) zooms programmatically (the linked pane follows), pane.export(...) writes one pane, view.on(..., pane="price") scopes events — 37_named_panes.py

Inset axes — overview * qv.Inset(zoom, rect=…, label="zoom", indicate=True): a child surface floating on its parent with the zoom window marked on the parent; the labeled inset is a pane (view.pane("zoom").set_range(…), pane-scoped events, per-pane export, state that survives backend switches) — 38_inset_zoom.py

Polar plots ([D119]) — polar as a transform: qv.polar(element) reinterprets x/y as (θ, r), qv.PolarGrid draws the circular chrome (degree or custom spoke labels — the radar case), qv.wedge() builds polar bars; the surface stays rectilinear, so pan/zoom/state/backends all keep working — 39_polar.py

Getting started

• 

  The smallest program — a scatter in a View, six lines.

  01_hello.py

• 

  Composition — overlay with *, lay out with +.

  02_composition.py

• 

  Switch engines at runtime — the same plot via pyqtgraph or matplotlib.

  03_backends.py

• 

  Themes & palettes — light/dark, colors, a registered custom palette.

  04_theming.py

Core concepts

• 

  Typed events — brush-select, pick, range.

  05_interaction.py

• 

  Accessors — bind channels to names, Expressions, callables, arrays.

  06_data_binding.py

• 

  Color & size encoding — color_by / size_by with automatic legends.

  12_color_mapping.py

• 

  Axis labels & titles — .opts(title=…, x=…, y=…) on any surface.

  31_axis_labels.py

• 

  The element vocabulary — the core element types in one grid.

  08_gallery.py

• 

  Two engines, one window — pyqtgraph beside matplotlib, one event stream.

  07_mixed_backends.py

• 

  The native escape hatch — view.native(id) returns the live pyqtgraph item.

  33_native_escape_hatch.py

Big data — Datashader

• 

  Millions of points — a density raster that re-aggregates to the viewport on zoom.

  09_datashader.py

• 

  Out-of-core — a lazy Dask DataFrame, never fully materialized.

  10_out_of_core.py

• 

  Backend-agnostic rasters — the same datashaded scatter on matplotlib.

  11_datashader_matplotlib.py

• 

  Hover a raster — HoverEvent.value reports the count under the cursor.

  25_raster_inspect.py

• 

  Raster legends & aggregation — a themed category blend with a legend, an agg="max" raster with a colorbar.

  32_datashader_legends.py

Reactive & adapters

• 

  Reactive crossfilter — brush one view; a Signal re-renders another.

  21_reactive_crossfilter.py

• 

  HoloViews, natively — a scatter * curve + bars tree, no browser.

  22_from_holoviews.py

• 

  DynamicMap — drive a kdim with a Qt slider; a Signal[Node] re-renders.

  23_from_holoviews_dynamicmap.py

• 

  hvplot one-liner — df.hvplot(kind="scatter") as a native Qt widget.

  24_from_hvplot.py

WebEngine

The same elements rendered as interactive Plotly in a QWebEngineView, plus RawFigure passthrough for existing Plotly / Bokeh / HoloViews figures. These need the webengine extra and a real display.

• 

  Plotly in a Qt window — typed events bridge back; swap to a native backend live.

  13_webengine.py

• 

  Multi-trace figures — an Overlay as Plotly traces; picks carry the series id.

  14_webengine_overlay.py

• 

  Mixed elements, one figure — Spread + Curve + Scatter together.

  15_webengine_elements.py

• 

  PNG export — view.handle.export("png", path).

  16_webengine_export.py

• 

  Plotly heatmap — a tabular Heatmap with a Viridis colorscale.

  17_webengine_heatmap.py

• 

  RawFigure passthrough — host a Plotly 3-D surface; events still bridge back.

  18_webengine_raw_figure.py

• 

  HoloViews via Bokeh — tap / box-select / range as typed qtviz events.

  19_webengine_holoviews.py

• 

  Native beside web — one window, one event stream, two engines.

  20_mixed_native_web.py

---

The full index with per-example notes and required extras lives in examples/README.md; regenerate every screenshot with uv run python tools/capture_screenshots.py.