Ridgeline Plot¶
A graphic that places multiple comparable profile curves on vertically offset baselines over a shared horizontal scale so their shapes and positions can be compared.
Core Idea¶
A ridgeline plot places several comparable profile curves on vertically offset baselines over a shared horizontal scale. Each group gets a row; its curve height represents a supplied or estimated quantity at each horizontal position. Readers can compare the shapes and locations of groups without giving every curve a separate panel.[ref-ada2f95813a9][ref-a51eb12d9264]
Kernel-density estimation and positive overlap are common choices, not defining requirements. The ggridges documentation supports directly supplied heights and shows separated, touching, and overlapping ridges. A ridgeline depicts the selected data relationships; it does not explain their causes or provide exact density values when no separate density axis is present.[ref-ada2f95813a9][ref-a51eb12d9264]
Scope of Application¶
The layout fits a question about how distributions or other profiles vary across groups. Wilke uses Lincoln, Nebraska, daily mean temperatures grouped by month and movie lengths grouped by release year. Those subjects differ, while the repeated-row, common-axis encoding remains the same. Groups need not be chronological: the ggridges vignette also uses iris species.[ref-ada2f95813a9][ref-a51eb12d9264]
A single filled curve lacks the repeated-group structure. Several unrelated panels with independent horizontal scales lack meaningful horizontal alignment. A violin mirrors a density around its centerline, while a ridgeline stacks profiles at distinct baselines. An area chart can fill a curve without having this multiple-row grammar. Positive overlap, equal spacing, a caption, and a particular smoothing method are optional.[^ref-a51eb12d9264]
Clarity¶
Read the shared horizontal variable, then each group's baseline and curve. A peak at the same x position is comparable across rows because the horizontal scale is common. Curve height must be interpreted according to its actual construction; a smooth ridge is not automatically a probability density or a measure of sample size. Wilke cautions that ridgeline densities lack a separate axis for reading exact density values.[ref-ada2f95813a9][ref-a51eb12d9264]
In Wilke's November temperature ridge, two visible clusters lie near roughly 35 and 50 degrees Fahrenheit. That is a description of the graphic, not evidence of why the weather took those values. The movie-length figure likewise does not explain production decisions.[^ref-ada2f95813a9]
Manages Complexity¶
The common axis and offset rows make many groups scannable in one frame. Wilke's movie example displays nearly one hundred yearly distributions from 1913 to 2005. Compactness can reduce legibility: as rows crowd or curves overlap, peaks, tails, and group boundaries may become harder to assign. The spacing and overlap should preserve the comparisons the reader needs.[ref-ada2f95813a9][ref-a51eb12d9264]
Abstract Reasoning¶
Identify the grouping variable, the common horizontal measurement, the rule that generates curve height, and the offset for each group. Check whether a claimed shift or mode survives the smoothing and scaling choices. If exact density readings matter, provide a quantitative density scale or choose another display; if boundaries disappear, reduce overlap or increase space. A visible difference remains descriptive until other evidence supports a causal claim.[ref-ada2f95813a9][ref-a51eb12d9264]
Knowledge Transfer¶
The same group → row, x value → common horizontal position, profile value → height mapping works for temperatures by month and movie lengths by year. It can also display supplied-height curves, but a supplied profile must not be mislabeled as an estimated density. Prime Representation is the broader target-to-medium mapping; this named plot is a domain-specific graphical convention, not a substrate-neutral Prime.[ref-ada2f95813a9][ref-a51eb12d9264]
Example¶
Monthly temperatures in Lincoln. Wilke's Figure 9.9 assigns a ridge to each month in 2016. Groups → months; horizontal scale → daily mean Fahrenheit temperature; offset baselines → monthly rows; curve height → estimated temperature density; reading → compare distribution shapes and positions, not exact density values or causes.[^ref-ada2f95813a9]
Movie lengths by release year. Wilke's Figure 9.11 assigns rows to years from 1913 to 2005. Groups → release years; horizontal scale → movie length; offset baselines → year rows; curve height → within-year distribution profile; reading → compare descriptive shapes and locations without claiming why movie lengths changed.[^ref-ada2f95813a9]
Relationships to Other Abstractions¶
Current abstraction Ridgeline Plot Domain-specific
Parents (1) — more general patterns this builds on
-
Ridgeline Plot is a kind of Representation Prime
A ridgeline plot is a representation that maps comparable group profiles to curves on vertically offset baselines and a common horizontal scale.
Hierarchy path (1) — routes to 1 parentless root
- Ridgeline Plot → Representation → Abstraction
Neighborhood in Abstraction Space¶
Ridgeline Plot sits in a sparse region of the domain-specific corpus (96th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Unclustered & Miscellaneous (2551 abstractions)
Nearest neighbors
- Parallel coordinates — 0.78
- Bubble Chart — 0.78
- Siegel–Tukey Test — 0.78
- Gower's Distance — 0.77
- Quadrant Count Ratio — 0.77
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
Violin Plot: a mirrored density form rather than a stack of upward profiles. Area Chart: a filled series may lack offset comparable rows. Small multiples: separate panels rather than one aligned ridgeline frame. Exact density readout: ridgelines commonly omit a density axis. A causal explanation: the plotted difference cannot establish its own cause. The historical pulsar image in the frozen seed was not independently verified and is not used as an evidenced case.[ref-ada2f95813a9][ref-a51eb12d9264]
References¶
[^ref-ada2f95813a9]: Claus O. Wilke, “Fundamentals of Data Visualization A Primer on Making Informative and Compelling Figures”, O’Reilly Media, 2019, chapter 9, especially §9.2 and Figures 9.9 and 9.11. The original book title uses a colon after “Visualization”; this linked transcription preserves the full title words for reference binding. Author-hosted complete manuscript; the publisher records 2019 as the first-edition year.
[^ref-a51eb12d9264]: Claus O. Wilke, “Introduction to ggridges”, ggridges package vignette, updated August 26, 2025, “Geoms,” “Ridgelines,” and “Density ridgeline plots.” First-party implementation documentation; supplied heights, density estimates and variable overlap are documented as distinct conventions.