PlotSet LogoPlotSet
Todos os tipos de gráficos

Stream Graph Maker

A stacked area chart with no flat baseline, flowing around a centre line. Upload your data and publish an interactive stream graph you can embed anywhere — free.

Criar um Stream Graph — grátisVer a galeria

O que é um Stream Graph?

A stream graph is a stacked area chart that has been freed from the axis. Instead of sitting on a flat baseline at zero, the bands are arranged symmetrically around a central line, so the whole shape swells and narrows with the total while each band keeps its own thickness. Giving up the baseline costs precision and buys legibility: the middle bands stop being distorted by everything below them, and a dozen categories can coexist in one picture. It became well known through the New York Times box-office graphic, and it remains a chart for showing the shape of a shift rather than the size of a number.

Quando usar um Stream Graph

  • You have many categories over time

    Ten or twenty series would make a stacked area chart unreadable. A stream graph is one of the few forms that stays legible at that count.

  • The shape of change is the story

    Which genre took over, when one topic displaced another, how attention moved. Rise, fall and displacement read immediately.

  • The total varies a lot

    The symmetric layout makes growth and contraction of the whole visible as the width of the stream, without any single series being blamed for it.

  • It is a headline graphic

    Stream graphs are unusually good at making people stop and look. For an editorial piece where engagement matters, that is a real property.

Quando não usar um Stream Graph

  • Readers need to read values

    There is no baseline to measure from and no axis a band can be read against. Nobody will get a number off this chart without a tooltip.

  • You have only a few series

    Three categories do not need this. A stacked area or a set of lines gives the same information with a usable axis.

  • Precise comparison between categories matters

    Comparing the thickness of two curved bands at different points is guesswork. Use small multiples of line charts.

  • The audience is unfamiliar and unhurried explanation is impossible

    The floating baseline is genuinely unusual. In a report where every chart must be self-explanatory, it will cost more than it gives.

Como criar um Stream Graph

De uma planilha a um gráfico publicado, passo a passo.

01

Start from the stream graph template

Open the PlotSet stream graph template. It loads with sample data bound so you can see how a time column and several series become bands.

02

Shape your data with time in the first column

One row per time point and one column per series. Dense, evenly spaced time points are what make the curves smooth rather than jagged.

03

Upload or connect your data

Import a CSV or Excel file, paste from a spreadsheet, or connect a Google Sheet so the graph can be refreshed later.

04

Order the bands so the story reads

Ordering by when each series peaks — earliest at the edges, latest in the middle, or the inside-out variants — is what turns a tangle into a visible sequence.

05

Publish, embed or export

Publish for a live interactive embed where hovering isolates a band and shows its value, or export a PNG, SVG or PDF.

Comece pelo modelo de Stream Graph, que já vem com dados de exemplo, e substitua-os pelos seus.

Abrir o modelo

As partes de um Stream Graph

Band
One series. Its thickness at any point is its value; its vertical position means nothing at all, which is the hardest thing for new readers to accept.
Centre line
The floating baseline the bands are arranged around. It is a layout device, not zero, and it is why values cannot be read off an axis.
Stream width
The total across all series at that moment. The swelling and narrowing of the whole shape is the clearest quantitative reading available.
Band ordering
The sequence of series from the outside in. The single biggest determinant of whether the graph is legible.
Colour
The only way to identify a band. With many series, an ordered or grouped palette beats a categorical one.
Smoothing
The curve interpolation between time points. Too much makes shapes that the data does not contain, and invents peaks between real observations.

Erros comuns em um Stream Graph

Expecting readers to get values from it

There is no axis. If numbers matter, put a table or a line chart beside it rather than pretending the stream can carry them.

Arbitrary band order

Alphabetical or source order produces a knot. Order by peak time or by size, and the same data becomes a story.

Over-smoothing

Aggressive curve fitting creates bulges between data points that never happened. Keep the interpolation modest and dense in time.

Too many colours

Twenty distinct hues cannot be told apart or legended. Group the small series, or colour by family and let the rest be grey.

Stream Graph: perguntas frequentes

Showing how many categories change over time when the overall shape of the shift matters more than exact values — genre popularity, topic attention, category mix.

Crie o seu Stream Graph agora

Comece grátis com mais de 300 modelos. Sem cartão de crédito.

Criar um Stream Graph — grátisTodos os tipos de gráficos