Cone Tree
A 3D tree visualization where each parent node sits at the apex of a cone and its children ring the base — stacking nested cones from root to leaves.
// 01 — The chart
What it looks like
A cone tree showing a hierarchy in 3D perspective. The root sits at the apex; children ring the base of each cone. Sub-cones nest beneath their parents.
// 02 — Definition
What is a cone tree?
A cone tree is a 3D hierarchy visualization invented by Robertson, Mackinlay, and Card at Xerox PARC in 1991. Each parent node sits at the apex of a cone, and its children are arranged around the circular base. The children themselves become apexes of smaller cones beneath, creating a recursive, nested structure.
The 3D projection allows the tree to use depth as a visual channel, fitting more nodes into the display than a flat 2D tree. Users rotate the cones to bring different children into focus — when one child is clicked, the cone smoothly rotates to bring it to the front.
Cone trees are primarily a research visualization rather than a production tool. They pioneered the idea of using animated 3D to make hierarchies navigable, and influenced later techniques like hyperbolic trees and treemaps.
Historical note: Cone trees were part of the “Information Visualizer” project at Xerox PARC in the early 1990s — the same lab that produced the concept of the desktop metaphor, Ethernet, and laser printing. They remain an important milestone in the history of information visualization.
// 03 — Anatomy
Parts of a cone tree
// 04 — Usage
When to use it — and when not to
- Exploring a hierarchy in a 3D environment with rotation and zooming
- The aesthetic impact of 3D rendering adds value to the presentation
- Each level has many children that benefit from circular arrangement
- Interactive demonstrations of hierarchy concepts in education
- The platform supports WebGL or 3D rendering natively
- Users are familiar with 3D navigation paradigms
- You need a static, printable diagram — 3D doesn't work on paper
- Accessibility matters — 3D navigation is challenging for many users
- The hierarchy is small enough for a 2D tree, treemap, or sunburst
- Occlusion hides important nodes — back-facing cones are invisible
- Performance is constrained — 3D rendering is compute-intensive
- The audience expects conventional charts — cone trees are unfamiliar to most people
// 05 — Reading guide
How to read a cone tree
Follow these steps whenever you encounter a cone tree.
Locate the root at the top
The topmost node is the root of the hierarchy. Everything below it is a descendant.
Read each level as a cone
Each parent node sits at the apex of a cone. Its children are arranged around the circular base below. The cone geometry makes it clear which nodes belong to which parent.
Rotate to see hidden children
Children at the back of the cone may be hidden by occlusion. Click or drag to rotate the cone and bring hidden nodes into view.
Look for nested sub-cones
Children that have their own children produce sub-cones beneath the main cone. The deeper you go, the smaller the cones become.
Gauge breadth by cone base size
A wide cone base means many children. A narrow base means few. This gives a quick visual estimate of how branchy each part of the hierarchy is.
// 06 — Pitfalls
Common mistakes
Occlusion hiding important nodes
Implement transparency, semi-transparent cones, or automatic rotation so that no important node is permanently hidden.
No rotation controls
Without the ability to rotate cones, users can only see the front half. Drag-to-rotate or click-to-front interaction is essential.
Disorienting camera angles
Provide fixed camera presets (top, front, 45°) so users can reset to a comfortable viewpoint if they get lost.
Ignoring performance
Large trees with thousands of 3D objects need level-of-detail rendering. Distant cones should simplify or disappear.
Using 3D for its own sake
Cone trees are rarely the best choice for actual data analysis. Use them when the 3D metaphor genuinely adds value, not just for novelty.
// 07 — In the wild
Real-world examples
Xerox PARC Information Visualizer
The original cone tree implementation — used to browse file system hierarchies in 3D, demonstrating that animated 3D transitions help users maintain context.
3D organizational charts
Some enterprise tools render org charts as cone trees in WebGL, allowing executives to fly through the organizational hierarchy.
Taxonomy visualization in museums
Interactive kiosks have used cone trees to let visitors explore biological classification hierarchies in an engaging 3D format.
// 08 — Quick reference
Key facts
Also known as
3D cone tree, Information Visualizer tree
Data type
Hierarchical (tree-structured)
Best for
Interactive 3D hierarchy exploration
Audience level
Advanced — requires 3D navigation skills
Invented
1991, Robertson, Mackinlay & Card (Xerox PARC)
Related to
Tree diagram, radial tree, hyperbolic tree, treemap
// 09 — Variations
Variations and extensions
Cam tree
A variation where only a single path from root to focus is expanded — sibling cones collapse flat to reduce visual clutter.
Cylinder tree
Uses cylinders instead of cones. Children wrap around the cylinder surface, avoiding the sharp convergence at the apex.
Semi-transparent cones
Drawing cones as transparent surfaces lets users see occluded children without rotating, solving the biggest usability issue.
// 10 — FAQs
Frequently asked questions
What is a cone tree?+
A cone tree is a 3D hierarchy visualization invented by Robertson, Mackinlay, and Card at Xerox PARC in 1991. Each parent node sits at the apex of a cone, and its children are arranged around the circular base. The children themselves become apexes of smaller cones beneath, creating a recursive, nested structure.
When should you use a cone tree?+
Use a cone tree when exploring a hierarchy in a 3D environment with rotation and zooming. It also works well when the aesthetic impact of 3D rendering adds value to the presentation, and when each level has many children that benefit from circular arrangement.
When should you avoid a cone tree?+
Avoid a cone tree when you need a static, printable diagram — 3D doesn't work on paper. It is also a poor fit when accessibility matters — 3D navigation is challenging for many users, or when the hierarchy is small enough for a 2D tree, treemap, or sunburst.
What is another name for a cone tree?+
Cone Tree is also known as 3D cone tree, Information Visualizer tree. The name varies between fields, but the visualisation technique is the same.
What size of dataset works best for a cone tree?+
Cone Tree works best for Interactive 3D hierarchy exploration. Outside that range the chart either looks empty or becomes too cluttered to read clearly.
Is a cone tree suitable for dashboards?+
Yes — a cone tree can work well in dashboards as long as the panel is large enough for readers to perceive the encoded values, has a clear title, and includes the legend or axis labels needed to interpret it.