We describe the design space for visualizations with attributes spanning several orders of magnitude, termed Orders of Magnitude Values (OMVs), and present OMVis, a tool for the interactive exploration of this design space. We divide OMVs into mantissa and exponent for separate visual encoding, similar to scientific notation. We create visualizations combining an OMV with another attribute-nominal, ordinal, time, or quantitative-using various marks and visual channels following the rules of the Grammar of Graphics. We refine this space by enforcing integrity constraints from visualization literature, aiming to enhance the effectiveness of the generated visualizations.

• Human-centered computing → Visualization systems and tools; Visualization theory, concepts and paradigms;