Patterns

Lesson 12: Patterns

Briefly exploring visualization patterns and user tasks.

Objective

Introduce common visualization patterns organized by variable count, establish core data terminology, and explore how user domains and tasks may guide pattern selection.

Outline

This lecture provides a tour through common chart types used for different situations as well as common terminology. It also offers a brief introductory mention of domains and tasks as a way to contextualize visualization patterns by understanding users.

Data terminology

Understanding the language used to describe data structures. Often:

• Variables are the column names in a dataset.
• Values are seen inside individual cells.
• Observations are rows in a table.
• Dimensions are variables by which observations are divided or segmented.
• Measures are typically numeric values which are encoded in visual channels.

These terms can be found in a number of software packages.

Visualization patterns by variable count

The lecture systematically explores visualization options as data complexity increases from one variable to seven variables.

• 1 variable:
  • Histogram
  • Pie chart
  • Donut chart
  • Circular graph
• 2 variables:
  • Scatter plot
  • Bar chart
  • Treemap
  • Sunburst chart
  • Chord diagram
  • Arc diagram
• 3 variables:
  • Bubble chart
  • Grouped bar chart
  • Stacked bar chart
  • Marimekko chart
  • Sankey diagram
• 4 variables:
  • Parallel coordinates plot
• Beyond 4 variables:
  • Shared axes
  • Dual axes
  • Novel representations
  • Interactivity
  • Multiple coordinated views
  • Small multiples

The grammar is limited

For higher density plots, traditional chart types become insufficient.

• Standard visualization patterns struggle with 5 or more variables.
• Alternative approaches include shared axes, novel (bespoke) representations, and interactivity.
• Multiple coordinated views and small multiples help manage complexity.

Re-centering the user

Moving from pattern selection to user-centered design through domains and tasks.

• Domains define who the users are and what concepts matter in their problem area.
• Tasks identify what questions users are trying to answer.
• The same data can be visualized differently depending on user needs.
• This approach connects to Tamara Munzner's "What/Why/How" framework explored in Lecture 14.

Take Aways

Visualization pattern selection should be driven by both data complexity and user needs, not just variable count alone.

• Learn core data terminology: variables, values, observations, dimensions, and measures.
• Understand common visualization patterns and their appropriate use cases.
• Recognize that traditional patterns struggle beyond 4 variables.
• Center design around user domains (who they are and their concepts) and tasks (questions they need to answer).
• The same dataset may require different visualizations depending on user goals.

Citations

[1] Tableau, "Tableau Logo," Salesforce, 2024. Available: https://www.salesforce.com/news/tableau-from-salesforce-logo-color-1/

[2] H. Wickham, "Tidy Data," Journal of Statistical Software, 2014. Available: https://www.jstatsoft.org/article/view/v059i10

[3] Tableau, "Dimensions and Measures, Blue and Green," Salesforce. Available: https://help.tableau.com/current/pro/desktop/en-us/datafields_typesandroles.htm

[4] S. Carmody, "Mosaic Big," Wikimedia Foundation, 2009. Available: https://en.wikipedia.org/wiki/Mosaic_plot#/media/File:Mosaic-big.png

[5] G. Silvermanaz, "2012 Singapore Products Export Treemap," Wikimedia Foundation, 2014. Available: https://en.wikipedia.org/wiki/Treemapping#/media/File:2012_Singapore_Products_Export_Treemap.png

[6] S. Maskey, "Zoomable Sunburst with Labels Issue," StackOverflow, 2014. Available: https://stackoverflow.com/questions/24547620/zoomable-sunburst-with-labels-issue

[7] T. Munzner, "A Nested Model for Visualization Design and Validation," IEEE Transactions on Visualization and Computer Graphics, vol. 15, no. 6, pp. 921-928, Nov.-Dec. 2009, doi: 10.1109/TVCG.2009.111

[8] S. Zhang, "Canyon," OpenProcessing. Available: https://openprocessing.org/sketch/2552991

[9] Y. Holtz and C. Healy, "From Data to Viz," 2018. Available: https://www.data-to-viz.com

License

This lesson is part of Interactive Data Science and Visualization and is released under a CC-BY-NC 4.0 license.