Stack Transform#

The stack transform allows you to compute values associated with stacked versions of encodings. For example, consider this stacked bar chart:

import altair as alt
from vega_datasets import data

source = data.barley()

alt.Chart(source).mark_bar().encode(
    column='year:O',
    x='yield:Q',
    y='variety:N',
    color='site:N'
).properties(width=220)

Implicitly, this data is being grouped and stacked, but what if you would like to access those stacked values directly? We can construct that same chart manually using the stack transform:

import altair as alt
from vega_datasets import data

source = data.barley()

alt.Chart(source).transform_stack(
    stack='yield',
    as_=['yield_1', 'yield_2'],
    groupby=['year', 'variety'],
    sort=[alt.SortField('site', 'descending')]
).mark_bar().encode(
    column='year:O',
    x=alt.X('yield_1:Q', title='yield'),
    x2='yield_2:Q',
    y='variety:N',
    color='site:N',
    tooltip=['site', 'yield', 'variety']
).properties(width=220)

Notice that the bars are now explicitly drawn between values computed and specified within the x and x2 encodings.

Transform Options#

The transform_stack() method is built on the StackTransform class, which has the following options:

Property	Type	Description
as	anyOf(`FieldName`, array(`FieldName`))	Output field names. This can be either a string or an array of strings with two elements denoting the name for the fields for stack start and stack end respectively. If a single string(e.g., `"val"`) is provided, the end field will be `"val_end"`.
groupby	array(`FieldName`)	The data fields to group by.
offset	[‘zero’, ‘center’, ‘normalize’]	Mode for stacking marks. One of `"zero"` (default), `"center"`, or `"normalize"`. The `"zero"` offset will stack starting at `0`. The `"center"` offset will center the stacks. The `"normalize"` offset will compute percentage values for each stack point, with output values in the range `[0,1]`. Default value: `"zero"`
sort	array(`SortField`)	Field that determines the order of leaves in the stacked charts.
stack	`FieldName`	The field which is stacked.

Sample Transform

TimeUnit Transform