Ridgeline plot maker
Compare the spread of measurements across groups.
Enter moves down.
Chart style
Look & reading
Ties use competition ranks: 1, 2, 2, 4. Blank cells break the line.
1,234.5
More settings · source, target and labels
Reader view
See it as your reader will.
The same live chart, without the editing fields. Hover or focus a mark to inspect its value.
Live preview
Start with the example or add your own data.
The same numbers, a different view. Change the style in Design. Your values stay with you.
Make a free ridgeline plot to compare distributions across groups. Paste raw observations, adjust the smoothing and export the chart with each group's median.
See the waits that a median hides
The sample contains fictional support response times. Each weekday has its own ridge on a shared minutes scale. A long right tail means some replies took much longer than the typical wait. The thin mark shows the median.
How to make a ridgeline plot
- Enter Group and Value columns, one measurement per row, or paste one column per group.
- Choose a shared bandwidth and height scale. Keep input order or sort the groups by median.
- Download a PNG or SVG for a slide or document, copy the embed code for a web page, or copy a link that reopens the chart with your data.
Keep every measurement
Use two columns, Group and Value, with one observation per row. Repeat the group name for every reading. Paste a spreadsheet with one column per group and the maker converts it to this long format.
| Group | Value |
|---|---|
| Monday | 12 |
| Monday | 18 |
| Tuesday | 9 |
Use up to 12 groups and 2,000 observations. Negative measurements are allowed. No data or a blank value means missing, never zero. Fewer than five known readings and constant groups appear as dots.
Choose how much detail to show
Auto uses Silverman's bandwidth on the within-group deviations from each median. Every group shares that bandwidth, so differences in smoothness come from the data. A smaller manual bandwidth reveals finer detail; a larger one blends nearby peaks.
The Gaussian density estimate stops four bandwidths beyond the overall data range. Heights share one density scale by default. Scaling each group to its own peak adds a notice because heights then mean different densities.
When to choose another chart
Use a dot plot for a few readings, a histogram for exact bin counts, or a bar chart of calculated medians when the center alone answers the question. Those makers take different data, so the shape picker here stays on Ridgeline.
Is the ridgeline plot maker free?
Yes. You can enter observations, adjust the chart and download PNG or SVG without signing up. Free exports and live embeds carry the Meiku Charts mark; Pro removes it.
What does Auto bandwidth do?
Auto applies Silverman's rule to the within-group deviations from each median and uses that bandwidth for every group. This gives all ridges the same smoothing. Move the bandwidth slider to inspect finer detail or blend nearby peaks.
How many readings does a group need?
A density ridge needs at least five known readings. Smaller samples and constant groups appear as dots, while empty groups say No data. Missing readings remain in the data table and never become zero.
Can I compare ridge heights?
Yes, when Shared density scale is selected. With Scale each group to its own peak, every maximum has the same height. The chart then prints a notice, and only the shapes remain comparable.