Stacked Density Estimates

To plot a stacked graph of estimates, use a shared extent and a fixed number of subdivision steps to ensure that the points for each area align well. Density estimates of body mass for each penguin species are plotted in a stacked method. In addition, setting counts to true multiplies the densities by the number of data points in each group, preserving proportional differences.

Method syntax

import altair as alt
from altair.datasets import data

source = data.penguins()

alt.Chart(source).transform_density(
    density='Body Mass (g)',
    groupby=['Species'],
    extent= [2500, 6500],
    counts = True,
    steps=200
).mark_area().encode(
    alt.X('value:Q').title('Body Mass (g)'),
    alt.Y('density:Q', stack='zero'),
    alt.Color('Species:N')
).properties(
    width=400,
    height=80,
    title='Distribution of Body Mass of Penguins'
)

Attribute syntax

import altair as alt
from altair.datasets import data

source = data.penguins()

alt.Chart(source).transform_density(
    density='Body Mass (g)',
    groupby=['Species'],
    extent= [2500, 6500],
    counts = True,
    steps=200
).mark_area().encode(
    alt.X('value:Q').title('Body Mass (g)'),
    alt.Y('density:Q', stack='zero'),
    alt.Color('Species:N')
).properties(
    width=400,
    height=80,
    title='Distribution of Body Mass of Penguins'
)