Use a treemap instead of a node-link tree when a weighted hierarchy must fit in one view
For overview of large weighted hierarchies, use a space-filling treemap instead of a node-link tree on hierarchical data to improve insight and mitigate screen-space overload for users scanning many levels at once.
- purpose:select
- basis:empirical
- chart:treemap:use
- chart:network:avoid
- data:hierarchical
- quality:insight
- lever:chart-family
- reading-mode:overview
- density:dense
advice
Choose a space-filling tree layout
Replace a full node-link tree with a treemap when the hierarchy is large and the display must show the whole structure at once. For example, use one screen-filling set of rectangles sized by leaf or subtree size instead of a rooted tree with lines when large items can appear at any level and the full node-link view no longer fits.
reason
Why the space-filling choice works
A space-filling tree layout turns nearly all available display area into data-carrying area. That lets readers scan the whole hierarchy at once and see large leaves wherever they occur, instead of spending space on links and indentation.
Mechanism: Area-proportional rectangles preserve relative size while using the full screen, so overview reading and large-item detection stay possible even when the tree has many levels and many leaves.
Evidence: The paper states that conventional rooted tree displays quickly overwhelm available display space so users cannot grasp the entire picture, and presents tree-maps as a 2-d space-filling representation that lets users rapidly recognize larger files anywhere in a large multi-level hierarchy (Shneiderman, 1992).
context
Use when the whole weighted tree must fit
- User Goal: See the entire hierarchy and quickly spot the largest leaves.
- Task: Compare relative sizes across many leaves and levels in one view.
- Data: Hierarchical data with a meaningful size or weight on leaves or subtrees.
- Chart Setting: Limited screen space and a need to show the whole tree simultaneously.
- Audience: People inspecting large hierarchical collections.
- Success Criterion: The full hierarchy fits in one view and the largest leaves stand out immediately.
exceptions
Do not use when area cannot carry the tree information
- Break it when: The hierarchy has no meaningful size attribute to map to rectangle area. Why: The method depends on each node rectangle having area proportional to an attribute such as size.
- Break it when: Many tiny or zero-size leaves must all remain individually visible at the current display resolution. Why: The paper notes that very small or zero-byte items become too small to represent and may be eliminated.
costs
Costs of switching to a treemap
Sacrifice: You give up the explicit node-and-edge drawing of a rooted tree. Risk: Very small leaves can disappear when the size range is large. Mitigation: Add zooming or restrict the view to a selected subtree when small items cannot be shown clearly.
mistakes
Common failure mode in this choice
Mistake: Keeping a full rooted node-link tree on one screen for a very large weighted hierarchy. Why it fails: The links and indentation consume the display before the reader can see the whole hierarchy or quickly spot the largest leaves.
check
How to test the chart-family decision
Failure Sign: The full node-link tree does not fit comfortably on one screen or the largest leaves are hard to identify across levels. Quick Check: Render the same hierarchy once as a node-link tree and once as a treemap; if only the treemap shows the whole hierarchy while keeping large leaves obvious, choose the treemap. Stronger Test: Ask a reviewer to point out the largest leaves anywhere in the hierarchy from a single screen; prefer the design that supports that overview without scrolling or page turning.
fix
What to change
- Replace the full node-link tree view with a treemap.
- Size each treemap rectangle by the stored leaf or subtree total.
- Use the full display area for the treemap instead of reserving space for links and indentation.
- Add zooming or a subtree-focused view if the smallest leaves still disappear.