A Julia package for visualizing data using deck.gl, a WebGL-powered framework for visual exploratory data analysis of large datasets.
using Pkg
Pkg.add(url="https://github.com/RallypointOne/DeckGL.jl")using DeckGL
# Sample data (any Tables.jl-compatible source works)
data = (
longitude = [-122.4, -122.5, -122.3, -122.45, -122.35],
latitude = [37.8, 37.7, 37.9, 37.75, 37.85],
size = [100, 200, 150, 180, 120]
)
# Create a scatterplot layer
layer = ScatterplotLayer(
data = data,
get_position = [:longitude, :latitude],
get_radius = :size,
get_fill_color = [255, 140, 0, 200]
)
# Create the deck, on a basemap
deck = Deck(
layer,
initial_view_state = ViewState(longitude=-122.4, latitude=37.8, zoom=11),
map_style = "https://basemaps.cartocdn.com/gl/positron-gl-style/style.json"
)
# Display in notebook or open in browser
open_html(deck)map_style is the URL of a MapLibre style. Carto's are free and
need no API key — swap positron for dark-matter or voyager. Leave map_style out
and the layers are drawn on a blank background. A basemap is fetched at view time, so it
needs network access.
- Tables.jl Integration: Works with DataFrames, NamedTuples, and any Tables.jl-compatible source
- Multiple Display Targets: Jupyter notebooks, VS Code, standalone HTML files
- GPU-Accelerated: Leverages deck.gl's WebGL rendering for large datasets
- Interactive: Pan, zoom, and rotate visualizations
Deck(layers; initial_view_state, map_style, controller)- Top-level visualization containerViewState(; longitude, latitude, zoom, pitch, bearing)- Camera configuration
All 32 deck.gl layers are available, generated from deck.gl's own prop tables. See the Layers page for a live example of each.
Core — one shape per record:
ScatterplotLayer, ArcLayer, LineLayer, PathLayer, PolygonLayer,
SolidPolygonLayer, TextLayer, IconLayer, BitmapLayer, ColumnLayer,
GridCellLayer, PointCloudLayer, GeoJsonLayer
Aggregation — bin records before drawing:
HexagonLayer, GridLayer, HeatmapLayer, ContourLayer, ScreenGridLayer
Geo — spatial indexes and tiled sources:
GreatCircleLayer, TripsLayer, TileLayer, MVTLayer, TerrainLayer,
Tile3DLayer, H3HexagonLayer, H3ClusterLayer, GeohashLayer, QuadkeyLayer,
S2Layer, A5Layer
Mesh — 3D geometry per record:
SimpleMeshLayer, ScenegraphLayer
Props may be written snake_case or camelCase, and anything left unset keeps
deck.gl's own default. DeckGL.props(ScatterplotLayer) lists what a layer accepts and
DeckGL.accessors(ScatterplotLayer) lists the props that can name a data column.
All 15 deck.gl widgets are available. See the Widgets page for a live example of each.
Camera — ZoomWidget, CompassWidget, GimbalWidget, ResetViewWidget,
ScrollbarWidget
Display — FullscreenWidget, ThemeWidget, ScreenshotWidget, LoadingWidget
Information — InfoWidget, PopupWidget, ContextMenuWidget
Custom controls — IconWidget, ToggleWidget, SelectorWidget
Deck(layer, widgets = [ZoomWidget(), CompassWidget(placement = "bottom-left")])to_js(deck)- The JavaScript that builds and mounts the visualizationto_html(deck)- Generate an HTML page (bundle=:localinlines deck.gl for offline use)save_html(deck, path)- Save as HTML fileopen_html(deck)- Open in default browser
Quick visualization helpers that return a Deck ready to display:
scatter(data, :lng, :lat; radius=:size, color=[255, 140, 0])
arcs(data, [:src_lng, :src_lat], [:dst_lng, :dst_lat])
lines(data, [:from_lng, :from_lat], [:to_lng, :to_lat])
paths(data, :path_column)
polygons(data, :polygon_column)
text(data, :lng, :lat, :label_column)
hexbin(data, :lng, :lat; radius=1000) # 3D hexagonal binning
heatmap(data, :lng, :lat; weight=:value) # Density heatmap
geojson(geojson_data) # GeoJSON visualizationMIT