Altair Interactivity

Altair’s biggest differentiator: charts are interactive by default in a browser/Jupyter, and that interactivity is genuinely built into the spec, not bolted on.

🖱️ Basic built-in interactivity

alt.Chart(df).mark_point().encode(x="a", y="b").interactive()   # enables pan + zoom, one call

💬 Tooltips

alt.Chart(df).mark_point().encode(
    x="a", y="b",
    tooltip=["a", "b", "category"]
)
alt.Chart(df).mark_point().encode(
    tooltip=[alt.Tooltip("value:Q", format=".2f", title="Value")]   # formatted tooltip
)

🎯 Selections (params) — the core interactivity mechanism

Modern Altair (5.x) uses alt.selection_point() / alt.selection_interval() + .add_params().

brush = alt.selection_interval()      # drag-select a rectangular region
 
alt.Chart(df).mark_point().encode(
    x="a", y="b",
    color=alt.condition(brush, "category:N", alt.value("lightgray"))   # highlight selected points
).add_params(brush)

click = alt.selection_point(fields=["category"])      # click to select a category
 
alt.Chart(df).mark_bar().encode(
    x="category:N", y="value:Q",
    opacity=alt.condition(click, alt.value(1), alt.value(0.3))
).add_params(click)

Selection type	Interaction
selection_point()	click a mark to select it
selection_point(fields=["col"])	click selects every mark sharing that field’s value
selection_interval()	drag a rectangular region (brush)
selection_interval(encodings=["x"])	brush limited to one axis

🎨 alt.condition — the “if selected, then” pattern

color=alt.condition(selection, "category:N", alt.value("lightgray"))

Reads as: if a mark is part of selection, color it by category; otherwise, gray it out.

alt.condition(predicate, if_true, if_false) is the single most important interactivity pattern in Altair — nearly every "highlight on click/hover" chart uses this exact shape.

🔗 Cross-filtering between two charts

brush = alt.selection_interval()
 
points = alt.Chart(df).mark_point().encode(
    x="a:Q", y="b:Q",
    color=alt.condition(brush, "category:N", alt.value("lightgray"))
).add_params(brush)
 
bars = alt.Chart(df).mark_bar().encode(
    x="category:N",
    y="count():Q"
).transform_filter(brush)      # bars update to reflect only the brushed selection
 
points | bars      # side-by-side, linked

Linking works because brush is a shared parameter — add_params() on one chart, transform_filter() on the other, referencing the same object.

🖱️ Hover-based selection

hover = alt.selection_point(on="mouseover", fields=["category"], nearest=True)
 
alt.Chart(df).mark_point().encode(
    x="a", y="b",
    size=alt.condition(hover, alt.value(200), alt.value(50))
).add_params(hover)

🎚️ Interactive parameter (e.g. a slider)

slider = alt.binding_range(min=0, max=100, step=1, name="Threshold ")
param = alt.param(bind=slider, value=50)
 
alt.Chart(df).mark_point().add_params(param).transform_filter(
    alt.datum.value > param
).encode(x="a", y="b")

🔗 Next

Altair Composition & Layouts · Altair Styling & Themes