A two-axis scatter chart answers one question well: how do two properties relate. Formulation and QC work rarely stops at two. A resin's viscosity, cure time, and tensile strength move together, and a chart that only shows two of the three at a time forces you to guess at the third or flip between plots to reconstruct it.
Within Uncountable, Explore Data's 3D scatter chart type puts all three on screen together, plus a fourth property as point color. Build one from the Build Graph interface (Visualize > Explore Data) by selecting the 3D chart type and assigning any plottable values to the X axis, Y axis, Z axis, and Color fields. The chart is interactive: rotate and zoom to change the angle, hover over a point for its exact values, and click a point to open its Compare view for a closer look at that formulation.
How does 3D scatter fit alongside Explore Data's other charts?
It's the newest addition to a chart library that already covers scatter, line, bar, box, and violin plots, and it draws on the same data Explore Data always has: ingredient amounts (as parts or percentages), process parameters, measurements, output conditions, experiment groupings, and derived custom values. Any of those can go on the X, Y, or Z axis as long as it's numeric, and any of them can drive the color field, numeric or not. At launch, 3D scatter doesn't yet support lasso filtering or Bodie-driven creation, both of which the other five chart types offer. Point count hasn't been an issue in testing either: charts with 150+ points still display and rotate cleanly, well above a typical formulation or QC dataset.
What does a 3D scatter chart show that a 2D chart doesn't?
A 3D scatter chart shows how three numeric properties relate at once, with a fourth property layered in as point color, so relationships and outliers that only appear across all three dimensions together become visible instead of hidden across separate two-axis charts.
Where this fits in formulation work
Multi-property analysis. When properties move together, real trade-offs only show up when you can see them together. Plotting three related properties at once surfaces the shape of that trade-off directly, rather than reconstructing it from several 2D charts.
Specification compliance. Coloring points by a spec property turns the chart into a compliance view: which formulations sit inside spec across all three dimensions, and which drift out, is visible at a glance rather than checked row by row.
Pattern recognition. Groupings and outliers that a 2D projection can flatten or hide often separate cleanly once a third axis is added and the view is rotated. A cluster that looks like noise from one angle can resolve into two distinct groups from another.
Do I need to set anything up first?
No. The 3D chart type is available now in Build Graph for every user, with no configuration or opt-in required. The only requirement is numeric data on the three axes you plot; the color field accepts any plottable value.
Getting there from a data point you're already looking at
Because a 3D scatter chart lives in Explore Data, any dataset you're already working with can go straight into one. Selecting a point opens its Compare view, so a chart built to spot an outlier or an out-of-spec cluster leads directly into the individual formulation record behind it. And like other Explore Data charts, a 3D scatter view can be saved to a report, notebook, or template and reloaded later, so a view built to catch a pattern doesn't have to be rebuilt each time.
“One of the things I already love about Uncountable is its data structure. It makes visualization much easier without having to think about chart formatting like in Excel. The new 3D charts take this to the next level. In some R&D projects, we need to analyze three or four properties simultaneously, and capturing insights in 2D charts was often challenging. The new 3D view opens exciting possibilities! There's still room for enhancement, but I'm really looking forward to seeing the next features!”
Benoit, Innovation Digitalization Specialist, Ahlstrom
.png)
.png)
