supports four methods for displaying multivariate data: each of them has a flat approach
and a hierarchical approach:
data can be defined as a set of entities E, where the ith
element ei consists of a vector with n varibles,
( xi1, xi2, ..., x in ).
Each variable (dimension) may be independent of or interdependent
with one or more of the other variable. Variables may be discrete
or continuous in nature, or take on symbolic (nominal) values.
Variables also have a scale associated with them, where scales
are defined according to the existence or lack of an ordering
relationship, a distance (interval) metric, and an absolute zero
multivariate data, each variable may map to some graphical entity
or attribute. In doing so, the type (discrete, continuous,
nominal) or scale may be changed to facilitate display.
In such situations, care must be taken, as a graphical variable
with a perceived characteristic (type or scale) which is mapped
to a data variable with a different characteristic can lead to
multivariate visualization techniques generally do not scale well
with respect to the size of the data sets. To deal with the clustering
and overlapping of large data sets, we extended the flat approaches
to hierarchical approaches. In hierarchical approaches, multi-resolutional
views of the data via hierarchical clustering substitute static
views in flat approaches. Variable-width opacity bands are
used to represent the information of clusters. Proximity-based
coloring highlights the relationships among clusters. Details
of the techniques used in the hierarchical approaches can
be obtained here.