A Pack for Superhuman Docs

BI-grade charts, straight from your tables.

Type =BarChart(Sales) and get a crisp, carefully designed chart of your table: bars, lines, donuts, funnels, waterfalls, treemaps, sankeys, heatmaps and more. For statistics, =DataChart(Survey, "raincloud") draws with ggplot2: distributions, fitted trends and small multiples. Both redraw whenever the data changes.

Try the playground See every chart

Real data: Our World in Data / Ember. The formula above draws it from a 208-row table.

How it works

1
Name a table
=BarChart(Sales) or =Chart(Sales, "line"). The first text column becomes the labels, number columns become series, and a second text column splits them. Raw rows are fine: repeats are summed.
2
The Pack renders it
The Pack reads the table, takes currency and percent formats from its columns, and a rendering service draws a 2× PNG with a colorblind-checked palette: Apache ECharts for Chart(), R and ggplot2 for DataChart().
3
A chart lands in your doc
The image is attached to the doc like any other. Change a cell and the formula recalculates and redraws it.

One formula per chart type, each just (Table, "title"), plus Chart(Table, type).
Extra settings in plain text: options: "sort: desc; group: Segment". Repeats are summed, or averaged or counted if you ask.
Two engines: Apache ECharts for business charts, R and ggplot2 for statistical ones. Same table, same Pack.
Three themes: paper, dark and navy, with number formats for currency, percent and compact.
Private by design: the chart service only accepts requests from Superhuman's Pack servers. It reads your table through the Superhuman Docs API (read-only) to draw the chart; table data isn't stored or logged. Rendered images are cached for 7 days.

Gallery

Two engines, one Pack. Pick a side to see what each one draws, then click any chart for its table and formula.

More from DataChart

Real public datasets and realistic business tables. Name a chart, like "raincloud", and the columns are picked for you, or describe it in layers: "point(Days, Size, color: Segment) + smooth(lm) + log(y)".

The plot text

Charts by name: raincloud ridgeline beeswarm violin trend slope dotplot multiples hex auto
Layers: point line col smooth(lm) boxplot halfeye interval(95%) histogram density tile text hline(100, 'Target')
Settings: facet(Region) sort(desc) log(x, y) flip() legend(none) labs(x: 'Days')
Columns: color: Segment or by: Team to group; Revenue & Cost to plot several number columns together.

Real-world data

Public datasets loaded as ordinary tables, charted with one formula each. Click a chart for its table, formula and source.

Playground

A live table, like one in your doc. Add rows and columns, change the numbers, rename things, or paste cells from a spreadsheet. Then edit the formula or pick a chart type. The chart is drawn with the same code the Pack runs. Switch to DataChart() and R with ggplot2 starts right in your browser (webR). Nothing you type leaves your browser.

Getting it

Superdocs Charts is in private beta and not yet listed in the Pack gallery. When it is, you'll add it from Insert → Packs in any doc and start typing =ColumnChart(.