AM

Tutorials Blog Roll of Honour About

 

 

 

Selections in Altair

import pandas as pd
import altair as alt
from vega_datasets import data

df = data.iris()
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 150 entries, 0 to 149
Data columns (total 5 columns):
 #   Column       Non-Null Count  Dtype  
---  ------       --------------  -----  
 0   sepalLength  150 non-null    float64
 1   sepalWidth   150 non-null    float64
 2   petalLength  150 non-null    float64
 3   petalWidth   150 non-null    float64
 4   species      150 non-null    object 
dtypes: float64(4), object(1)
memory usage: 6.0+ KB

Selections

To make selections in Altair, we need three things:

1. A selection object;
2. A conditional object; something that changes according to the selection made, and
3. A chart.

There are two types of selection available; a selection interval, which is called a brush in some other plotting packages, and clickable selections. We'll start with the selection interval.

Selection Interval

Where the selection interval really comes into its own is in related subcharts, as the selection is repeated across the charts. An example makes that clearer.

selector = alt.selection_interval()

condition = alt.condition(selector,
                          'species',
                          alt.value('lightgray'))

petals = alt.Chart(df).mark_circle().encode(
    x='petalLength',
    y='petalWidth',
    color=condition,
    tooltip=df.columns.to_list())

sepals = alt.Chart(df).mark_circle().encode(
    x='sepalLength',
    y='sepalWidth',
    color=condition,
    tooltip=df.columns.to_list())

(petals | sepals).add_selection(selector)

Selection Single

A single selection is just as you'd expect - point and click.

selector = alt.selection_single()

condition = alt.condition(selector,
                          'species',
                          alt.value('lightgray'))

petals = alt.Chart(df).mark_circle().encode(
    x='petalLength',
    y='petalWidth',
    color=condition,
    tooltip=df.columns.to_list())

sepals = alt.Chart(df).mark_circle().encode(
    x='sepalLength',
    y='sepalWidth',
    color=condition,
    tooltip=df.columns.to_list())

(petals | sepals).add_selection(selector)

Single on mouseover

But because pointing and clicking can be tiresome, there's an on parameter we can set to mouseover.

selector = alt.selection_single(on="mouseover")

condition = alt.condition(selector,
                          'species',
                          alt.value('lightgray'))

petals = alt.Chart(df).mark_circle().encode(
    x='petalLength',
    y='petalWidth',
    color=condition,
    tooltip=df.columns.to_list())

sepals = alt.Chart(df).mark_circle().encode(
    x='sepalLength',
    y='sepalWidth',
    color=condition,
    tooltip=df.columns.to_list())

(petals | sepals).add_selection(selector)

Multiple Selections

To make multiple selections, we change our selector to .selection_multi(), set on="mouseover", and set toggle=False.

selector = alt.selection_multi(on="mouseover", toggle=False)

condition = alt.condition(selector,
                          'species',
                          alt.value('lightgray'))

petals = alt.Chart(df).mark_circle().encode(
    x='petalLength',
    y='petalWidth',
    color=condition,
    tooltip=df.columns.to_list())

sepals = alt.Chart(df).mark_circle().encode(
    x='sepalLength',
    y='sepalWidth',
    color=condition,
    tooltip=df.columns.to_list())

(petals | sepals).add_selection(selector)

Select by Fields

We can use the fields parameter to select by category.

selector = alt.selection_multi(fields=['species'])

condition = alt.condition(selector,
                          'species',
                          alt.value('lightgray'))

petals = alt.Chart(df).mark_circle().encode(
    x='petalLength',
    y='petalWidth',
    color=condition,
    tooltip=df.columns.to_list())

sepals = alt.Chart(df).mark_circle().encode(
    x='sepalLength',
    y='sepalWidth',
    color=condition,
    tooltip=df.columns.to_list())

(petals | sepals).add_selection(selector)

Selecting by Legend

Selecting by legend may be the most useful method of all. Again, the code is very similar to what it's been all along, but there are two points to note.

1. The second parameter of alt.condition takes an alt.Color() value, rather than just the column name. This means we can turn off the default legend.
2. We now have a fourth object, a legend object, to join our selector, condition, and chart objects.

selector = alt.selection_multi(fields=['species'])

condition = alt.condition(selector,
                          alt.Color('species:N', legend=None),
                          alt.value('lightgray'))

chart = alt.Chart(df).mark_circle().encode(
    x='petalLength',
    y='petalWidth',
    color=condition,
    tooltip=df.columns.to_list())

legend = alt.Chart(df).mark_circle().encode(
    y=alt.Y('species:N', axis=alt.Axis(orient='right')),
    color=condition)

(chart | legend).add_selection(selector)

The Original Video