Pivot Transform#

The pivot transform is, in short, a way to convert long-form data to wide-form data directly without any preprocessing (see Long-form vs. Wide-form Data for more information). Pivot transforms are useful for creating matrix or cross-tabulation data, acting as an inverse to the Fold Transform.

Here is an example, using Olympic medals data:

import altair as alt
import pandas as pd

df = pd.DataFrame.from_records([
    {"country": "Norway", "type": "gold", "count": 14},
    {"country": "Norway", "type": "silver", "count": 14},
    {"country": "Norway", "type": "bronze", "count": 11},
    {"country": "Germany", "type": "gold", "count": 14},
    {"country": "Germany", "type": "silver", "count": 10},
    {"country": "Germany", "type": "bronze", "count": 7},
    {"country": "Canada", "type": "gold", "count": 11},
    {"country": "Canada", "type": "silver", "count": 8},
    {"country": "Canada", "type": "bronze", "count": 10}
])

alt.Chart(df).transform_pivot(
    'type',
    groupby=['country'],
    value='count'
).mark_bar().encode(
    x='gold:Q',
    y='country:N',
)

The pivot transform, when combined with other elements of the Altair grammar, enables some very interesting chart types. For example, here we use pivot to create a single tooltip for values on multiple lines:

import altair as alt
from vega_datasets import data

source = data.stocks()
base = alt.Chart(source).encode(x='date:T')
columns = sorted(source.symbol.unique())
selection = alt.selection_point(
    fields=['date'], nearest=True, on='mouseover', empty='none', clear='mouseout'
)

lines = base.mark_line().encode(y='price:Q', color='symbol:N')
points = lines.mark_point().transform_filter(selection)

rule = base.transform_pivot(
    'symbol', value='price', groupby=['date']
).mark_rule().encode(
    opacity=alt.condition(selection, alt.value(0.3), alt.value(0)),
    tooltip=[alt.Tooltip(c, type='quantitative') for c in columns]
).add_parameter(selection)

lines + points + rule

Transform Options#

The transform_pivot() method is built on the PivotTransform class, which has the following options:

Property	Type	Description
groupby	array(`FieldName`)	The optional data fields to group by. If not specified, a single group containing all data objects will be used.
limit	`number`	An optional parameter indicating the maximum number of pivoted fields to generate. The default (`0`) applies no limit. The pivoted `pivot` names are sorted in ascending order prior to enforcing the limit. Default value: `0`
op	`AggregateOp`	The aggregation operation to apply to grouped `value` field values. Default value: `sum`
pivot	`FieldName`	The data field to pivot on. The unique values of this field become new field names in the output stream.
value	`FieldName`	The data field to populate pivoted fields. The aggregate values of this field become the values of the new pivoted fields.

LOESS Transform

Quantile Transform