ABSTRACT
VECTOR data is as frequently encountered, and as important, as scalardata. Strictly put, a vector is a tuple of n scalar components v = (v1, · · · , vn), vi ∈ R. An n-dimensional vector describes, for example, a position, direction, rate of change, or force in Rn. However, the majority of visualization applications deal with data that describes physical phenomena in 2D or 3D space. As a consequence, most visualization software defines all vectors to have three components. 2D vectors are modeled as 3D vectors with the third (z) component equal to null. Although one could provide separate implementation-level support for 2D vectors, this would massively complicate the structure of visualization software, lead to code replication, and ultimately reduce performance.
