Trellis charts are grouped sets of visuals that represent different partitions (segments) of the dataset in the same chart type, scale, and axes. This makes the comparison of sub-groups very intuitive. Trellis charts are sometimes called lattice charts, grid charts, panel charts, or small multiples.
Any chart type that uses the X and Y axes explicitly may be trellisable, depending on the dataset. What is required is an extra entry on either the X or Y shelf; the first field on the shelf represents the segment definition of the trellis, and the second field is the actual dimension. Some charts types have optional X and Y shelves so that trellises may be enabled.
A dataset may be double-trellisable — you could declare a 'partition' on both X and Y shelves at the same time.
There are two approaches for trellises:
- Commonly, we use Trellising on dimensions to show the sections of data in separate but related charts.
- We can also show different data aggregates, side by side, by placing extra measurements on the Measures shelf, to Trellis on measures.
You can find specific examples of trellised charts in the following articles:
- Word clouds
Some visual types (bars, lines, areas, and grouped bars) support trellising on measures.