PlotSet LogoPlotSet
全部图表类型

Violin Plot Maker

See the whole shape of a distribution, not just its five-number summary. Upload a spreadsheet, pick a group column and a value column, and publish an interactive violin plot you can embed anywhere — free.

免费创建 Violin Plot查看图库

什么是 Violin Plot？

A violin plot is a distribution drawn as a silhouette. For each group, the chart estimates how densely the values cluster at every point on the scale and draws that density as a width — wide where many observations pile up, narrow where few do — then mirrors it so the shape reads symmetrically around a central axis. The result is a smooth outline, vaguely instrument-shaped, that carries everything a histogram would show while taking up the same narrow strip a box plot needs. That is the whole argument for it: a box plot compresses a group into five numbers and loses any structure those five numbers do not capture, and the most important thing it loses is a second peak. A violin keeps the shape, so two clusters stay two clusters, and you can still stand a dozen groups side by side.

什么时候使用 Violin Plot

  • You suspect the distribution is not one simple hump

    Bimodality is the classic case — two populations quietly mixed into one group. A violin shows two bulges where a box plot shows one wide, featureless box.

  • You are comparing several groups and shape matters

    Response times by server, salaries by role, test scores by school. One violin per group on a shared axis, and the differences in spread and skew are visible without reading a single number.

  • You have enough data to justify a density estimate

    Violins earn their keep from a few hundred observations per group upward, where a histogram's bin edges start to feel arbitrary and a smooth curve is genuinely the more honest summary.

  • Your audience will read a shape faster than a statistic

    Nobody needs the term 'kernel density estimate' explained to see that one group is fat in the middle and another is spread thin. The outline does the teaching.

什么时候不该用 Violin Plot

  • You have only a handful of observations per group

    Below roughly fifty values, the density curve is mostly the smoothing parameter talking. Plot the points — a strip plot or a jittered dot plot is honest at that size and a violin is decoration.

  • Your readers need exact quartiles

    A violin communicates shape, not precise summary statistics. If the median and the interquartile range are the numbers people will quote, use a box plot, or overlay one inside the violin.

  • Your values are bounded and pile up at the edge

    Percentages that bunch at 0 or 100, counts that cannot go below zero. The smoothing spills past the boundary and draws density where no value can exist, which is a lie the chart tells confidently.

  • The distribution genuinely is one plain hump

    If every group is unimodal and roughly symmetric, the violin adds curves and a harder read for no new information. A box plot or a simple ranking of medians is the better chart.

如何制作 Violin Plot

从电子表格到已发布的图表，一步一步来。

01

Start from the violin chart template

Open the PlotSet violin template. Sample data comes bound, so you can see the shape of spreadsheet it expects before you replace it with your own.

02

Bring raw values, one row per observation

Two columns is the usual arrangement: a group label and a numeric value. Do not pre-aggregate — the density is computed from the individual observations, and a table of averages has nothing left to estimate a shape from.

03

Upload or connect your data

Import a CSV or Excel file, paste straight from a spreadsheet, or connect a Google Sheet so the chart can be refreshed when the data moves.

04

Check the smoothing before you trust the shape

The bandwidth controls how much the curve is smoothed. Too high and two peaks merge into one; too low and random noise sprouts bumps. Nudge it in both directions and keep only the features that survive.

05

Order the groups deliberately

Alphabetical order is almost never the useful order. Sort by median and the chart becomes a ranking with shape attached rather than a lookup table.

06

Publish, embed or export

Publish for a live interactive embed with a tooltip on every group, or export a PNG, SVG or PDF for a report or a deck.

从已加载示例数据的 Violin Plot 模板开始，再替换成你自己的数据。

打开模板

Violin Plot 的组成部分

Density curve
The outline itself. Its width at any height is an estimate of how many observations sit near that value — wide means common, narrow means rare. It is an estimate, not a count.
Mirroring
The curve is drawn on both sides of a central axis. The two halves are identical and carry no extra information; the symmetry exists because a shape reads more clearly than a half-shape.
Bandwidth
The smoothing parameter behind the curve. It is a choice, not a property of the data, and it is the single setting that most changes what the chart appears to say.
Inner summary
Many violins carry a box plot, a median marker or the raw points inside them. That inner layer is what restores the exact quartiles the smooth outline blurs away.
Extent of the curve
Where the violin stops. Some implementations trim it to the observed minimum and maximum, others let the smoothing run past them — which is why a violin can appear to show impossible values.
Category axis
One slot per group. Its order is a design decision: sorting by median turns a set of shapes into a comparison the reader can follow left to right.

Violin Plot 的常见错误

Accepting the default bandwidth without looking

The same data can be drawn as one broad hump or three crisp peaks depending on a number nobody chose deliberately. Try two or three settings and report a shape that is stable across them.

Reading width as a count

A wide violin means values are concentrated there, not that the group is large. Groups are usually scaled to equal area, so a group of 50 and a group of 5,000 can look the same size. Label the counts.

Letting the curve run past what is possible

Smoothing a bounded variable draws density below zero or above 100 per cent. Trim the violin to the observed range, or say in the caption that the tails are an artefact.

Using a violin on tiny groups

With fifteen observations the outline describes the smoother, not the data. Plot the points instead; it is more honest and, at that size, just as readable.

Dropping the summary statistics entirely

Shape without a median is hard to quote. Keep a box, a bar or a dot inside the violin so a reader who wants the number can find it.

Violin Plot：常见问题

Comparing the full distribution of a numeric variable across several groups at once. Unlike a box plot, it shows the shape of each distribution — including multiple peaks — while still fitting many groups on one axis.

立即制作你的 Violin Plot

免费开始，300 多个模板，无需信用卡。

免费创建 Violin Plot全部图表类型