Vega-Altair Tutorial for SciPy 2024

This repository contains resources for the Vega-Altair tutorial for SciPy 2024.

See the tutorial description for more information.

Dependencies

The only Python dependency of this tutorial is the altair package with the all extras enabled.

To install with pip from PyPI:

pip install altair[all]

Or, to install from conda-forge:

conda install -c conda-forge altair-all vega_datasets anywidget

During the live tutorial at SciPy 2024, participants are encouraged to use the Nebari JupyterHub distribution (https://scipy.quansight.dev) with the vega-altair-tutorial conda environment, which has these dependencies pre-installed.

Outline

This tutorial is divided into four parts, each focusing on different aspects of data visualization using Vega-Altair. Below is an outline of the tutorial along with links to each notebook.

Part 1: Effective Visual Data Communication

• Objective: Understand the principles of designing effective and perceptually efficient visualizations, including chart anatomy, the grammar of graphics, and how design choices impact viewer comprehension.
• Key Topics:
  • Grammar of Graphics: marks and encoding channels
  • Human perception and graphical effectiveness
  • Visual affordances and Gestalt principles
  • Using color, annotations, and layout to guide attention
  • Common visualization pitfalls and how to avoid them

Part 2: Data Types, Graphical Marks, and Visual Encoding Channels

• Objective: Learn the basic syntax of Altair, understand how to encode data into visual properties, and customize chart appearances.
• Key Topics:
  • Basic Altair syntax
  • Encoding data dimensions
  • Customizing chart appearance
• Notebook: Part 2 - Data Types, Graphical Marks, and Visual Encoding Channels
• Exercises: Part 2 - Exercises

Part 3: Data Transformation and Interactivity

• Objective: Learn about integrating data transformations and interactivity into a charts.
• Key Topics:
  • Binning and aggregation
  • Basic interactive features (tooltips and pan/zoom)
  • Selection parameters
  • Conditional encodings and filtering
• Notebook: Part 3 - Data Transformation
• Exercises: Part 3 - Exercises

Part 4: Workflows with Altair

• Objective: Learn how to export and share Altair visualizations in various formats, integrate with dashboarding systems, and create custom theming for consistent visual design across projects.
• Key Topics:
  • Exporting visualizations (HTML, PNG, SVG, PDF)
  • Creating self-contained shareable URLs
  • Integration with dashboarding systems (PowerBI, Panel, Streamlit)
  • Building custom charting libraries with Altair
  • Applying themes and consistent styling across charts
• Notebook: Part 4 - Workflows with Altair
• Exercises: Part 4 - Exercises