Getting Started

Basic plots

AlgebraOfVega re-exports AlgebraOfGraphics, so you use the same data, mapping, and visual functions:

using AlgebraOfVega

# Scatter plot
spec = data(df) * mapping(:horsepower, :mpg, color=:origin) * visual(Scatter)
draw(spec)

Config

Use config() to set Vega-Lite properties like width, height, title, and custom encodings:

spec = data(df) * mapping(:x, :y) * visual(Lines) *
    config(width=600, height=400, title="Time Series")

Config's encoding is deep-merged with auto-generated encodings, so you can add properties like aggregation without losing the generated field/type:

# Add mean aggregation to y
config(encoding=Dict("y" => Dict("aggregate" => "mean")))

Composition

Use + to layer and * to combine:

# Scatter + regression line
spec = data(df) * mapping(:x, :y) * (visual(Scatter) + linear())

# Multiple layers with shared data
spec = data(df) * mapping(:x, :y, color=:group) *
    (visual(Scatter, opacity=0.5) + smooth())

Statistical analyses

AoG analyses are translated to native Vega-Lite transforms:

density()      # kernel density → area mark
histogram()    # binning → bar mark
linear()       # OLS regression → line mark
smooth()       # LOESS smoothing → line mark
frequency()    # count aggregation → bar mark
expectation()  # mean aggregation → bar mark

Add confidence intervals:

linear(interval=:confidence)  # regression line + CI band
smooth(interval=:confidence)  # LOESS line + CI band

Faceting

Use AoG's col and row mappings for faceted layouts:

# Column facets
spec = data(df) * mapping(:x, :y, col=:species) * visual(Scatter)

# Row + column grid
spec = data(df) * mapping(:x, :y, row=:sex, col=:species) * visual(Scatter)

Output formats

draw(spec)         # HTMX Node (for web apps using HTMXObjects)
to_html(spec)      # standalone HTML string with embedded scripts
to_vegalite(spec)  # Vega-Lite JSON as Dict
to_json(spec)      # Vega-Lite JSON string

Mark types

AoG visual	Vega-Lite mark
Scatter	point
Lines	line
ScatterLines	line (point: true)
BarPlot	bar
Heatmap	rect
BoxPlot	boxplot
Band	area
HLines / VLines	rule
Stairs	line (step-after)
Errorbars / Rangebars	errorbar