1. Visualization Object Model

Every visualization created using Atlas can be described using the following object model:

At the top level, we have a scene, acting as a container for all the visualization objects. In a scene, we typically have axes, legends, gridlines, and collections of items. A collection can be nested, e.g., the children of a collection are also collections. A collection consists of multiple marks or glyphs; where each glyph is essentially a group of marks. Finally, a mark is typically represented as a set of vertices, connected by line or curve segments. For example, a rectangle mark is composed of four vertices connected by four line segments.

In the diverging bar chart in Figure 1, we have three main components in the scene: a legend, an axis without ticks or path, and a collection of rectangle marks which forms the main chart area. Inside this collection nest four collections of bars, each collection representing an age group and consisting of four bars.

In the box plot in Figure 2, we have two main components in the scene: an axis without ticks or path, and a collection of four box and whiskers glyphs. Inside each glyph, there are multiple primitive marks in the form of rectangle and lines.