Graphics Reference
In-Depth Information
Figure 5-19:
Colors and size of labels reinforce the color and size of the
underlying nodes.
Ingeneral, theapproachofusingonedataattributetosettwoormorevisual
attributes is called
redundant encoding
. Some redundant encoding can
be considered to be good because, as viewers learn the visualization, they
see these different cues representing the same thing, and it helps them
understand it. In
Figure 5-19
, it seems logical to have smaller labels on
smaller items and larger labels on larger items—the redundant encoding
here acts as an aid to the viewers.
Key Edge Attributes
Edges, like nodes, may have many attributes associated with them. Edges
represent relationships. There can be different types of relationships,
including directions, quantities, measurements, and statistics. Weight,
color, and line style are key visual attributes for revealing additional data
about edges.
Edge Weight
Edge weight
is common in many data sets and is important for many
layout algorithms. Visually, edge weight is simply the thickness of the edge
between two nodes. In the e-mail data set, edge weight has already been
included in all the previous examples because the particular force-directed
layout used edge weight as part of its calculations. However, the edges were
all drawn fairly thin to de-emphasize edge weight. As shown in
Figure 5-20
,
edge weight can be increased, revealing connections where there have been
many e-mails.
