More Information

Citrin Ternary Scatter Graphs

Creating a ternary chart involves transforming three variables to ratios of three dependent proportions that are plotted in a 2-D simplex.


Ternary Scatter Graph Types

A 2-D projection of a ternary plot displaying the relative proportions (as percentages) of three variables on three axes. In this graph type, you can display the data points by group using object markers (see the left-hand side image below), or display the data as value-labels representing values of a 4th variable (see the right-hand side image below).

A Ternary Scatter Chart Displaying Three Data Groups

A Ternary Scatter Chart Displaying Value Labels of a 4th Variable

Ternary Error Regions

For calculating ternary error regions, three possible values are considered for each variable:

  • The raw value
  • The raw value plus the positive error bar value
  • The raw value plus the negative error bar value

    • Accordingly, Citrin calculates 27 possible combinations for each ternary data point and uses the convex hull to delineate the resulting hexagonal error region:

Applying Ternary Error Regions to Ternary Scatter Graphs

Ternary Zooming Support

Precise zooming into a sub-triangle region of data within a ternary chart can not be achieved using the standard zoom tool. In Citrin, controls are provided in the ternary axis dialog to perform ternary zooming.

The Significance of Ternary Zooming