Hive Plot
A perceptually uniform network layout using radial axes to position nodes by attribute — making network structure readable, reproducible, and comparable.
// 01 — The chart
What it looks like
A hive plot with three axes showing software modules positioned by connectivity. The highlighted edge traces a key dependency from a source hub to a utility node.
// 02 — Definition
What is a hive plot?
A hive plot is a network visualization method invented by Martin Krzywinski in 2012 to address the fundamental problem of force-directed layouts: they look different every time you run them. In a hive plot, nodes are placed along linear axes radiating from a central point, with their position on each axis determined by a measurable node attribute (such as degree, age, or cluster membership).
Typically a hive plot uses 2–4 axes. Nodes are assigned to an axis based on one categorical attribute (like node type or community), then positioned along that axis by a quantitative attribute (like connectivity or centrality). Edges are drawn as curved lines connecting nodes on different axes.
The key advantage is perceptual uniformity: the same data always produces the same layout, making hive plots ideal for comparing networks, spotting structural differences, and communicating reproducible findings. Unlike force-directed graphs, there’s no randomness — the layout is entirely determined by data.
Origin: The hive plot was introduced by Martin Krzywinski at the BC Cancer Research Centre in 2012, published in Briefings in Bioinformatics. It was designed specifically to create rational, reproducible network visualizations for genomics and systems biology research.
// 03 — Anatomy
Parts of a hive plot
// 04 — Usage
When to use it — and when not to
- You need a reproducible, deterministic network layout that doesn’t change between runs
- Nodes have meaningful categorical attributes that map naturally to 2–4 axes
- You want to compare two networks side by side using the same layout rules
- Your network has known node types (e.g., proteins, genes, metabolites) that define axes
- You need to present network findings in a publication or report where reproducibility matters
- You want to reveal connection patterns between specific node categories
- Your nodes lack meaningful categorical attributes to define axes
- You want an exploratory, free-form layout — use a force-directed graph instead
- The audience is unfamiliar with network visualization — hive plots have a learning curve
- You have more than 4 node categories — too many axes become confusing
- Your network is very small (<10 nodes) — a simple node-link diagram is clearer
- You need to show geographic or spatial positions of nodes
// 05 — Reading guide
How to read a hive plot
Follow these steps whenever you encounter a hive plot.
Identify the axes and their meaning
Each axis represents a category of nodes. Read the axis labels to understand what grouping is being used — node type, community, functional role, or another categorical attribute.
Understand the position metric
Nodes are placed along their axis based on a quantitative attribute. Nodes closer to the center typically have lower values (e.g., fewer connections), while those further out have higher values. Check the legend or caption for the metric.
Read the edges between axes
Curved lines connecting nodes on different axes represent relationships. Dense bundles of edges between two axes indicate strong inter-group connectivity.
Look for edge density patterns
Where edges cluster between axes reveals the dominant connection patterns. Sparse zones between axes mean those groups interact less frequently.
Compare with other hive plots
Because hive plots are deterministic, you can place two side by side and spot structural differences — a missing bundle, a shift in density, or new high-degree hubs.
// 06 — Common mistakes
Mistakes to watch out for
Using meaningless axis assignments
The power of a hive plot comes from intentional axis assignment. Randomly assigning nodes to axes produces no insight. Each axis should represent a meaningful category that you want to compare.
Too many axes
More than 4 axes makes the plot confusing and hard to interpret. The angles between axes become too narrow, edges overlap unpredictably, and the visual advantage is lost. Stick to 2–4 axes.
Ignoring the position metric
If node position along the axis is arbitrary, you lose half the information the plot can convey. Always map position to a meaningful quantitative variable (degree, centrality, expression level).
Not labeling axes clearly
Without clear axis labels and position metric descriptions, readers can’t interpret the plot. Always annotate what each axis represents and how nodes are ordered along it.
Overloading with too many edges
Dense networks create edge bundles that obscure all structure. Filter edges by weight threshold, or show only the top connections to keep the visualization readable.
// 07 — Real-world examples
Where you’ll see hive plots used
Genomics: Protein interaction networks
Researchers use hive plots to visualize protein–protein interaction networks in model organisms. Axes represent functional categories (kinases, receptors, transcription factors), and position encodes connectivity. Side-by-side hive plots compare healthy vs. diseased interaction networks.
BioinformaticsSoftware engineering: Module dependencies
Software architects map codebase dependencies using hive plots. Source modules, libraries, and utilities each get an axis, with position determined by import count. This reveals which modules are over-coupled and where dependency refactoring is needed.
Software AnalysisSocial networks: Community interactions
Network scientists use hive plots to compare social network structures across platforms or time periods. Axes represent detected communities, and the deterministic layout allows precise before/after comparisons when studying structural changes.
Network Science// 08 — At a glance
Quick reference
// 09 — Variations
Types of hive plots
The basic hive plot can be adapted with different axis configurations and edge encodings.
Two-axis hive plot
The simplest form with just two axes, useful for visualizing connections between two distinct node types.
Four-axis hive plot
Uses four axes for networks with four node categories. More expressive but requires careful labeling to remain readable.
Paired hive plot (comparison)
Two hive plots side by side using identical axis rules, designed for comparing network structure across conditions or time.
Weighted hive plot
Edge thickness encodes relationship strength or frequency, adding a quantitative dimension to the connection pattern.
// 10 — FAQs
Frequently asked questions
What is a hive plot?+
A hive plot is a network visualization method invented by Martin Krzywinski in 2012 to address the fundamental problem of force-directed layouts: they look different every time you run them. In a hive plot, nodes are placed along linear axes radiating from a central point, with their position on each axis determined by a measurable node attribute (such as degree, age, or cluster membership).
When should you use a hive plot?+
Use a hive plot when you need a reproducible, deterministic network layout that doesn’t change between runs. It also works well when nodes have meaningful categorical attributes that map naturally to 2–4 axes, and when you want to compare two networks side by side using the same layout rules.
When should you avoid a hive plot?+
Avoid a hive plot when your nodes lack meaningful categorical attributes to define axes. It is also a poor fit when you want an exploratory, free-form layout — use a force-directed graph instead, or when the audience is unfamiliar with network visualization — hive plots have a learning curve.
Is a hive plot suitable for dashboards?+
Yes — a hive plot 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.
What category of chart is a hive plot?+
Hive Plot belongs to the Network family of charts. Charts in that family are designed to answer the same kind of question, so they often work as alternatives when one doesn't quite fit your data.
How do you read a hive plot?+
Start with the axis labels and legend, then look at the overall shape before zooming into individual marks. Compare prominent features against the rest of the data, and verify any conclusion against the underlying numbers when precision matters.