How to draw a pie chart in Altair
Altair, a Python-based declarative statistical visualization library, offers a streamlined approach to crafting diverse visualizations. It is designed to simplify the creation of a wide range of visualizations, making it easy for users to generate expressive and informative charts without the need for low-level plotting code. Altair is part of the broader ecosystem of tools for data science and visualization in Python.
The pie chart in Altair
A pie chart, often called a circular statistical graphic, visually segments a whole into proportional slices to showcase numerical proportions. Each segment within the pie chart symbolizes a portion of the total dataset, with the magnitude of each slice corresponding to its respective quantity. The entirety of the circle encapsulates the entirety of the data, representing a complete dataset.
Pie charts are commonly used to show the composition of a whole, especially when there are a small number of categories. They effectively display relative proportions and provide a quick visual overview of how parts contribute to the whole.
Creating a pie chart in Altair is similar to creating other charts. Here are the steps to draw a pie chart using Altair:
Importing Altair: We import the Altair library in our Python code. We need a dataset to visualize. We can use pandas or other data manipulation libraries to load our data.
import altair as altimport pandas as pd# Load your data into a pandas DataFramedata = pd.read_csv('your_data.csv')
Creating a pie chart: We use the Altair API to create a pie chart. We specify the data source, encoding channels, and mark type (in this case,
arc).
chart = alt.Chart(data).mark_arc().encode(color='Category:N',theta='Value:Q',tooltip=['Category', 'Value']).properties(width=300,height=300)
Customizing the chart: We can add various customizations to our pie charts, such as labels, titles, tooltips, and color schemes.
Example
Let's create a meaningful example using a hypothetical dataset representing the distribution of expenses in a monthly budget. This pie chart will show the proportions of spending in different expense categories
import altair as altimport pandas as pdimport os# Hypothetical data: Monthly budget expensesdata = pd.DataFrame({'Category': ['Housing', 'Transportation', 'Food', 'Entertainment', 'Utilities'],'Expense': [1200, 400, 300, 200, 150]})# Create a pie chartchart = alt.Chart(data).mark_arc().encode(color='Category:N',theta='Expense:Q',tooltip=['Category', 'Expense']).properties(title='Monthly Budget Expenses',width=400,height=400)# Display the chartchart.save('chart.html')os.system('cat chart.html')
Explanation
Lines 1–3: We import Altair and other necessary libraries.
Lines 5–9: We create a pandas DataFrame named
datawithCategoryandExpensecolumns.Lines 11–20: We initialize an Altair chart with data where:
Categoryvariable (NominalN) determines the color of arcsValuevariable (QuantitativeQ) sets the angle of arcstooltipdisplays both category and value informationChart dimensions are set to 400x400 pixels
N and Q are shorthand notations for data types in Altair, which represent nominal (categorical) and quantitative (numerical) data types, respectively.
N(Nominal): Nominal data represents categories or groups without any inherent order or magnitude. Examples include names, labels, or categorical variables like colors, types, or species.Q(Quantitative): Quantitative data represents numerical values with an inherent order and magnitude. Examples include integers, floats, or any numerical values that can be measured or compared.
Line 23: We save the chart using
chart.save('chart.html'). It exports the chart to an HTML file namedchart.html.Line 24: We display the chart on the console.
Free Resources
